JP6274166B2 - Battery temperature estimation device - Google Patents

Description

  The present invention relates to a battery temperature estimation device.

  Conventionally, as seen in, for example, Patent Document 1 below, a battery that sequentially identifies resistance values and the like of resistance components in an equivalent circuit model of the battery by an observer, and sequentially estimates the internal temperature of the battery based on the identified resistance value and the like A temperature estimation device is known.

JP 2010-135075 A

  Some battery temperature estimation devices include an observer that sequentially estimates the battery temperature in the equation of state derived from the thermal network model. The thermal network model includes, as input variables, the current value flowing through the battery and the temperature at a predetermined position of the transmission path through which the heat generated in the battery is transmitted, and the battery temperature as a state variable. This is a model of movement. The thermal network model includes the internal resistance of the battery whose resistance value increases as the battery temperature decreases.

  In the observer, an observer gain is designed to estimate the battery temperature. The observer gain is defined as the time until the estimated battery temperature converges to the true battery temperature within the fluctuation range of the internal resistance value corresponding to the fluctuation range of the battery temperature assumed at the time of design. It is designed to satisfy the robustness required of the observer, such as guaranteeing the performance.

  Here, if the variation range of the battery temperature assumed at the time of design is wide, the variation range of the internal resistance value of the battery is also widened. If the variation range of the internal resistance value is wide, the robustness required for the observer may not be satisfied in the entire variation range of the battery temperature. For this reason, for example, it is conceivable to design the observer gain so as to satisfy the robustness required for the observer in a part of the battery temperature fluctuation range. However, in this case, if the battery temperature deviates from the partial temperature range described above, the time until the estimated battery temperature converges to the true battery temperature cannot be reduced to a predetermined time or less, or the observer's stability is reduced. There is concern that it cannot be guaranteed. In this case, there is a concern that the battery temperature estimation accuracy by the observer is reduced.

  The main object of the present invention is to provide a battery temperature estimation device that can increase the estimation accuracy of the battery temperature even when the variation range of the battery temperature is wide.

  Hereinafter, means for solving the above-described problems and the operation and effects thereof will be described.

  The first invention includes a battery (21), a current detector (53) for detecting a current flowing through the battery, and a heat transfer path (K1a, Kb, Kp) through which heat generated in the battery is transmitted. And a temperature detection unit (50-52) provided in the battery unit (10), the battery unit (10) including a current value flowing through the battery and a temperature at a predetermined position of the heat transfer path as input variables, and the battery. The temperature of the battery in a state equation derived from a thermal network model that models the movement of heat in the heat transfer path is included in the state variable, the current detection value of the current detection unit, and the temperature detection unit The thermal circuit network model includes an internal resistance of the battery that has a resistance value that increases as the temperature of the battery decreases. The battery corresponding to the boundary of the selected temperature range is selected from a plurality of temperature ranges set by dividing a variation range that can be taken by the temperature detection value of the degree detection unit, and the temperature range including the temperature detection value is selected. A setting unit (62) for setting upper and lower limit values of the internal resistance value, and a gain calculation unit for calculating an observer gain used by the observer based on the upper and lower limit values of the internal resistance value set by the setting unit (61c).

  In the above invention, the observer sequentially estimates the battery temperature in the state equation derived from the thermal network model based on the current detection value of the current detection unit and the temperature detection value of the temperature detection unit. Here, the resistance value of the internal resistance of the battery included in the thermal network model is increased as the temperature of the battery is lower. For this reason, if the fluctuation range that the battery temperature can take is wide, the fluctuation range of the internal resistance value becomes wide, and the robustness required for the observer may not be satisfied in all the fluctuation ranges that the battery temperature can take.

  Therefore, in the above invention, the temperature detection value that is correlated with the battery temperature is used, and the temperature detection value is included from the temperature ranges that are set by dividing the variation range that the temperature detection value can take. Select the temperature range. After selecting the temperature range, the upper and lower limit values of the internal resistance value corresponding to the boundary of the selected temperature range are set, and the observer gain used by the observer is calculated based on the set upper and lower limit values.

  According to the above invention, in each of the temperature ranges set by dividing the variation range that the temperature detection value can take, the robustness required for the observer in the variation range of the internal resistance value corresponding to the temperature range is achieved. Observer gain can be calculated so as to satisfy. For this reason, even when the variation range that the battery temperature can take is wide, the estimation accuracy of the battery temperature can be increased, for example, the divergence of the estimated value of the battery temperature can be avoided.

  In the second invention, in each of the temperature ranges set by dividing the fluctuation range, a convergence rate for individually converging the battery temperature estimated by the observer to its true value is set. It is characterized by that.

  In the above-described invention, an observer gain with an optimized convergence rate can be calculated in each temperature range so that the estimated time for the estimated battery temperature to converge to its true value can be shortened. For this reason, even if there is an estimation error between the initial estimated value of the battery temperature and its true value, the estimated error can be quickly eliminated.

  According to a third aspect of the present invention, the internal resistance value has a larger increase amount per unit temperature decrease amount of the battery as the temperature of the battery is lower. Each of the ranges is set to be narrower as the temperature detection value is lower.

  In the above invention, the increase in the internal resistance value per unit temperature decrease amount of the battery is increased as the battery temperature is lower. Therefore, for example, when the fluctuation range that the temperature detection value can take is divided into a plurality of temperature ranges having the same temperature range, the lower the temperature detection value, the higher the internal resistance value corresponding to the boundary of the temperature range. The difference between the lower limit values increases. As a result, there is a concern that the observer gain cannot be calculated so as to satisfy the robustness required for the observer in the temperature range where the temperature detection value is low.

  Therefore, in the above invention, each of the temperature ranges set by dividing the variation range that the temperature detection value of the temperature detection unit can take is set narrower as the temperature detection value is lower. For this reason, it can suppress that the difference of the upper and lower limit of an internal resistance value corresponding to the boundary of the temperature range becomes large, so that a temperature range with a low temperature detection value. Accordingly, the observer gain can be calculated so as to satisfy the robustness required for the observer in each temperature range.

  According to a fourth aspect of the present invention, the gain calculation unit sets the observer gain that stabilizes the closed loop of the observer with respect to fluctuations in the internal resistance value in each of a plurality of temperature ranges set by dividing the fluctuation range. It is characterized by calculating.

  In the above invention, even when the internal resistance value varies according to the detected temperature value, the stability of the closed loop of the observer can be guaranteed in each temperature range. Thereby, the estimation accuracy of the battery temperature can be further increased, for example, the divergence of the estimated value of the battery temperature can be avoided.

  In a fifth aspect, the gain calculation unit calculates the observer gain that stabilizes the closed loop of the observer with respect to process noise and observation noise mixed in a signal in the thermal network model to be controlled. Features.

  In the above invention, even when process noise and observation noise are mixed in the signal, the stability of the closed loop including the observer can be guaranteed. Thereby, the estimation precision of battery temperature can be raised more.

  According to a sixth aspect of the present invention, there is provided another estimation for estimating the temperature of the battery with a processing load smaller than the processing load required for the temperature estimation of the battery by the observer and different from the battery temperature estimation method by the observer The temperature estimation accuracy by the observer is higher than the temperature estimation accuracy by the separate estimation unit, a temperature range with high reliability of the battery is defined as a high reliability temperature range, and adjacent to the high reliability temperature range In addition, a temperature range in which the reliability of the battery is lower than the high reliability temperature range is set as a low reliability temperature range, and a temperature range including the temperature estimated by the separate estimation unit is the high reliability temperature range. In this case, the temperature estimation by the separate estimation unit is continued, and the temperature range including the temperature estimated by the separate estimation unit is changed from the high reliability temperature range to the low reliability temperature range. If Tsu, characterized in that it comprises a switching unit for switching the temperature estimation by the observer.

  In the above invention, the temperature estimation accuracy by the observer is higher than the temperature estimation accuracy by the separate estimation unit, and the processing load required for the temperature estimation by the separate estimation unit is smaller than the processing load required for the temperature estimation by the observer. . Therefore, in the above invention, when the temperature range including the temperature of the battery estimated by the separate estimation unit is the high reliability temperature range, the temperature estimation by the separate estimation unit is continued. For this reason, the processing load of the battery temperature estimation device can be reduced until the temperature range including the battery temperature estimated by the separate estimation unit changes from the high reliability temperature range to the low reliability temperature range.

  On the other hand, in the above invention, when the temperature range including the battery temperature estimated by the separate estimation unit changes from the high reliability temperature range to the low reliability temperature range, the temperature estimation by the separate estimation unit is switched to the temperature estimation by the observer. For this reason, in the situation where the estimated temperature of the battery is included in the low reliability temperature range, the temperature estimation accuracy can be increased. Thereby, it can avoid that a battery is used in an overheated state or used in a low temperature state.

  According to a seventh aspect of the invention, the temperature detection unit includes a short-distance temperature detection unit (50, 51) provided on a position where the path length of the heat transfer from the battery is small on the heat transfer path, and the heat transfer path. A long-distance temperature detection unit (52) provided on a position on the heat path where the heat transfer path length from the battery is large, and the heat transfer path includes the battery and the short-distance temperature detection unit. The first heat transfer resistance (R1hc, R2hc, Rb) is present in the first path portion (K1a, K2a, Kb) between the first distance portion and the far distance temperature detection portion. The second heat transfer resistance (Rp) exists in the two path portions (Kp), and the separate estimation portion is a detected temperature that is a difference between the temperature detection values of the short distance temperature detection portion and the long distance temperature detection portion. Based on the difference and the first heat transfer resistance and the second heat transfer resistance, the short-range temperature detection is performed. A temperature difference calculation unit that calculates an estimated temperature difference that is a temperature difference between the installation position of the unit and the battery, and adds the estimated temperature difference calculated by the temperature difference calculation unit to the temperature detection value of the short-range temperature detection unit And a battery temperature calculation unit for calculating the temperature of the battery.

  According to the above invention, the short-distance temperature detection unit and the long-distance temperature detection unit are provided on the heat transfer path through which the heat generated in the battery is transmitted, and the battery and the short-distance temperature detection unit are provided in the heat transfer path. It is assumed that the first heat transfer resistance exists and the second heat transfer resistance exists between the short-distance temperature detection unit and the long-distance temperature detection unit. Under such a configuration, if the difference between the temperature detection values of the short-distance temperature detection unit and the long-distance temperature detection unit is obtained, the heat transmitted through the second path unit based on the temperature difference and the second heat transfer resistance. The amount is determined. If the amount of heat transfer in the second path portion and the amount of heat transfer in the first path portion are associated in advance, the amount of heat transfer in the first path portion is also determined, whereby the amount of heat transfer in the first path portion and the first amount of heat transfer are determined. Based on the heat transfer resistance, the temperature difference between the battery and the short-distance temperature detection unit can be obtained.

  Therefore, in the above invention, based on the detected temperature difference that is the difference between the temperature detection values of the short-distance temperature detection unit and the long-distance temperature detection unit, and the first heat transfer resistance and the second heat transfer resistance, the short-distance temperature detection unit And the estimated temperature difference which is a temperature difference of a battery is calculated. And the temperature of a battery is computable by adding the computed estimated temperature difference to the temperature detection value of a short distance temperature detection part.

Sectional drawing of the battery unit which concerns on 1st Embodiment. The top view of a control board. The figure which shows the thermal network model for designing an observer. The figure which shows the structure of an observer. The figure which shows the temperature characteristic of the internal resistance value of a battery cell. The figure which shows the thermal circuit network model which concerns on 2nd Embodiment. The functional block diagram of a 2nd temperature estimation process. The flowchart which shows the procedure of a battery temperature estimation process.

(First embodiment)
Hereinafter, a first embodiment in which a battery temperature estimation device according to the present invention is applied to an in-vehicle battery unit will be described with reference to the drawings.

  First, the whole structure of the battery unit 10 is demonstrated using FIG. In the following description, for the sake of convenience, the vertical direction of the battery unit 10 is defined with reference to FIG. 1 in which the battery unit 10 is installed on a horizontal plane.

  The battery unit 10 includes an assembled battery 20, a control board 30 that controls charge / discharge of the assembled battery 20, and a housing case 40 that houses the assembled battery 20 and the control board 30.

  The housing case 40 includes a bottom plate portion 41 that is fixed to a place where the battery unit 10 is mounted, a peripheral wall portion 42, and a cover 43. The bottom plate portion 41 is formed in a rectangular shape and is made of a metal material such as aluminum. The peripheral wall portion 42 is provided along the peripheral edge portion of the bottom plate portion 41. The cover 43 is attached to the end of the peripheral wall 42 in the wall height direction opposite to the bottom plate 41.

  The bottom plate portion 41 has a placement portion 44 on which the assembled battery 20 is placed. The assembled battery 20 placed on the placing portion 44 and the control board 30 are arranged to face each other vertically so that the assembled battery 20 is on the bottom and the control board 30 is on the top. The assembled battery 20 and the control board 30 are disposed so as to be surrounded by the peripheral wall portion 42.

  The cover 43 is formed in a rectangular shape like the bottom plate portion 41 and is formed of a metal material such as aluminum. The cover 43 has substantially the same size as the bottom plate portion 41 in plan view.

  Next, the assembled battery 20 will be described. The assembled battery 20 has a plurality of battery cells 21 as single cells. Each battery cell 21 is a laminated battery having a plate shape, and the battery cells 21 are joined in a state of being stacked one above the other. Specifically, a double-sided adhesive type adhesive tape is interposed between the battery cells 21, and the battery cells 21 are integrated by adhesion of the adhesive tape.

  In the present embodiment, the assembled battery 20 is composed of four battery cells 21. Each of these battery cells 21 is a first battery cell 21a, a second battery cell 21b, a third battery cell 21c, and a fourth battery cell 21d in order from the upper one to the lower one.

  The battery cell 21 has a square plate-shaped battery main body 22 and a pair of plate-shaped electrode tabs 23 and 24 as electrode terminals connected to the battery main body 22. The battery body 22 is accommodated in a flat container 25 made of a laminate film, and the peripheral edge of the flat container 25 is sealed to be sealed in the container 25. The battery body 22 of each battery cell 21 is provided in a state where it is stacked one above the other.

  The pair of electrode tabs 23 and 24 are respectively provided on the two opposite sides of the battery body 22. Each of the electrode tabs 23 and 24 is drawn out from the battery body 22 toward the opposite side, and more specifically, is drawn out in a direction perpendicular to the stacking direction of the battery cells 21. One of the electrode tabs 23, 24 is a positive electrode tab 23, and the other is a negative electrode tab 24. In the present embodiment, the positive electrode tab 23 is made of aluminum, and the negative electrode tab 24 is made of copper. Hereinafter, the direction orthogonal to the stacking direction of the battery cells 21 is also referred to as a tab pull-out direction.

  The battery cells 21 stacked one above the other are arranged such that the positive electrode tabs 23 and the negative electrode tabs 24 are alternately arranged between the battery cells 21 adjacent to each other in the vertical direction. In this case, in each of the battery cells 21 adjacent to each other in the vertical direction, the positive electrode tab 23 of one battery cell 21 and the negative electrode tab 24 of the other battery cell 21 face each other vertically and overlap each other, and are joined to each other at the overlapping portion. ing. Thereby, each battery cell 21 is connected in series.

  Among the battery cells 21 stacked one above the other, in the first battery cell 21 a, the positive electrode tab 23 </ b> A is not connected to the negative electrode tabs 24 of the other battery cells 21. Further, in the fourth battery cell 21 d arranged at the lowermost portion, the negative electrode tab 24 </ b> A is not connected to the positive electrode tab 23 of the other battery cell 21. The positive electrode tab 23 </ b> A and the negative electrode tab 24 </ b> A constitute the positive electrode terminal and the negative electrode terminal of the series connection body of each battery cell 21, respectively, and the tab drawing directions are the same.

  The negative electrode tab 24 of the first battery cell 21 a and the positive electrode tab 23 of the second battery cell 21 b are electrically connected to the control board 30 via the first bus bar 31. The negative electrode tab 24 of the third battery cell 21 c and the positive electrode tab 23 of the fourth battery cell 21 d are electrically connected to the control board 30 via the second bus bar 32. The positive electrode tab 23 </ b> A of the first battery cell 21 a is electrically connected to the control board 30 via the third bus bar 33. The negative electrode tab 24 of the second battery cell 21 b and the positive electrode tab 23 of the third battery cell 21 c are electrically connected to the control board 30 via the fourth bus bar 34. The negative electrode tab 24A of the fourth battery cell 21d is electrically connected to the control board 30 via the fifth bus bar 35. Each of the bus bars 31 to 35 is provided so as to extend upward. Thereby, the terminal voltage can be detected for each of the battery cells 21a to 21d via the bus bars 31 to 35 connected to each other.

  Next, the structure regarding the control board 30 is demonstrated using FIG.1 and FIG.2.

  As shown in the drawing, the control board 30 is formed of a rectangular board (rectangular board) printed board having a circuit pattern formed on the board surface. As described above, the control board 30 is arranged above the assembled battery 20, and the board long side direction is arranged in the tab drawing direction of the electrode tabs 23 and 24.

  Various electronic components are mounted on the board surface of the control board 30. These electronic components include a control unit 60 (CPU) that executes charge / discharge control processing of the assembled battery 20, a switching element 36, and the like. In FIG. 2, for convenience of illustration, the control unit 60 is shown separated from the control board 30.

  The control board 30 is formed with a through hole 30a penetrating in the thickness direction at a substantially central portion thereof. The switching element 36 is disposed on one side of both sides of the through-hole 30a in the short side direction of the substrate on the control substrate 30, and is not disposed on the other side. In this case, the switching element 36 is not present on the other side of the control board 30 from the through hole 30a. Further, from the viewpoint that the switching element 36 is a heat generating element that generates heat, the region can also be referred to as an unheated region 30b in which no heat generating element exists.

  First and second bus bars 31 and 32 and third to fourth bus bars 33 to 35 are connected to the control board 30 at positions opposite to each other in the board longitudinal direction. Each of these bus bars 31 to 35 is connected to the control board 30 while being inserted into a hole formed in the control board 30. Further, each of the bus bars 31 to 35 is connected to the non-heated region 30 b in the control board 30.

  On the control board 30, a first temperature sensor 50 that detects the temperature of the connection part between the control board 30 and the first bus bar 31 and a second temperature that detects the temperature of the connection part between the control board 30 and the second bus bar 32. A temperature sensor 51 is mounted. The first temperature sensor 50 is disposed in the vicinity of the first bus bar 31 on the control board 30, and the second temperature sensor 51 is disposed in the vicinity of the second bus bar 31 on the control board 30. In the present embodiment, the thermistors are used as the first and second temperature sensors 50 and 51. In the present embodiment, the first and second temperature sensors 50 and 51 correspond to a short-range temperature detection unit.

  The arrangement of the first and second temperature sensors 50 and 51 will be described in detail. The first and second bus bars 31 and 32 are arranged side by side in the board longitudinal direction on the control board 30, and the arrangement direction of the bus bars 31 and 32 is arranged. The temperature sensors 50 and 51 are arranged along the line. In this case, the temperature sensors 50 and 51 are arranged on the end side (the same side) in the short-side direction of the substrate with respect to the bus bars 31 and 32.

  In addition to the first and second temperature sensors 50 and 51, a third temperature sensor 52 that detects the temperature of the control board 30 is mounted on the control board 30. In the present embodiment, a thermistor is used as the third temperature sensor 52. Unlike the first and second temperature sensors 50 and 51, the third temperature sensor 52 is disposed on the control board 30 at a position away from the first and second bus bars 31 and 32. Therefore, the third temperature sensor 52 is less affected by the temperature from the second bus bars 31 and 32 than the first and second temperature sensors 50 and 51, and detects the temperature of the control board 30. . In the present embodiment, the third temperature sensor 52 corresponds to a long-distance temperature detection unit.

  The third temperature sensor 52 is disposed on the same side as the first and second temperature sensors 50 and 51 with respect to the first and second bus bars 31 and 32. The second temperature sensors 50 and 51 are located on the opposite side of the first and second bus bars 31 and 32. Further, the third temperature sensor 52 is disposed in the non-heated region 30 b on the control board 30. Therefore, the third temperature sensor 52 is less affected by heat from the switching element 36.

  Each of the temperature sensors 50 to 52 is connected to the control unit 60. The temperature detection values are input to the control unit 60 from the temperature sensors 50 to 52, respectively. In addition, a current detection value is input to the control unit 60 from a current sensor 53 that detects a charge / discharge current flowing through each battery cell 21. The controller 60 estimates the temperature of the battery body 22 of each battery cell 21 (hereinafter referred to as “internal temperature”) based on the temperature detection values of the temperature sensors 50 to 52 and the current detection value of the current sensor 53. Perform the temperature estimation process. According to this process, the temperature of each battery cell 21 can be acquired without directly attaching a temperature sensor to each battery cell 21. Hereinafter, the first and second battery cells 21a and 21b are taken as an example to describe the internal temperature estimation method, and then the temperature estimation process will be described.

<Internal temperature estimation method>
In the battery unit 10, the electrode tabs 23 and 24, the first and second bus bars 31 and 32, and the control board 30 form a heat transfer path through which heat generated in each battery body 22 is transmitted. In the present embodiment, a thermal circuit network model that models the movement of heat in the heat transfer path is created, and the internal temperature of each battery cell 21 is estimated using the thermal circuit network model.

  FIG. 3 shows a thermal network model. FIG. 3 shows a thermal circuit network model by extracting the first and second battery cells 21 a and 21 b from the assembled battery 20.

  As shown in FIG. 3, the battery main bodies 22 of the first and second battery cells 21a and 21b serve as heat sources in the thermal circuit network model. The heat transfer path from the first and second battery cells 21a, 21b to the control board 30 will be described. The first tab path portion K1a is constituted by a negative electrode tab 24 connected to the first battery cell 21a. The 2nd tab path | route part K2a is comprised by the positive electrode tab 23 connected to the 2nd battery cell 21b. The bus bar path portion Kb includes a first bus bar 31 and a substrate portion from the connection portion of the control board 30 to the first bus bar 31 to the mounting position of the first temperature sensor 50. The board path portion Kp is configured by a board portion from the mounting position of the first temperature sensor 50 to the mounting position of the third temperature sensor 52 on the control board 30.

  In the thermal circuit network model, an inter-cell path portion Kt that is a heat transfer path formed by the joint surfaces of the adjacent first and second battery cells 21a and 21b is shown. The thermal network model further shows a first space path portion K1b and a second space path portion K2b. The first space path portion K1b is a heat transfer path from the first battery cell 21a to the mounting position of the third temperature sensor 52 on the control board 30 through the space in the housing case 40. The second space path portion K2b is a heat transfer path from the second battery cell 21b to the mounting position of the third temperature sensor 52 on the control board 30 through the space in the housing case 40.

  The first tab path K1a has a heat transfer resistance R1hc, the second tab path K2a has a heat transfer resistance R2hc, and the inter-cell path Kt has a heat transfer resistance R12. Is present. A heat transfer resistance Rb exists in the bus bar path portion Kb, and a heat transfer resistance Rp exists in the substrate path portion Kp. The first space path portion K1b has a heat transfer resistance R1ht, and the second space path portion K2b has a heat transfer resistance R2ht. A heat capacity C1 exists between the first battery cell 21a and the mounting part 44, and a heat capacity C2 exists between the second battery cell 21b and the mounting part 44.

  The amount of heat generated by the battery body 22 of the first battery cell 21a is indicated by Q1j, and the amount of heat transfer from the battery body 22 of the first battery cell 21a to the first space path portion K1b is indicated by Q1ht. The amount of heat generated by the battery body 22 of the second battery cell 21b is indicated by Q2j, and the amount of heat transfer from the battery body 22 of the second battery cell 21b to the second space path portion K2b is indicated by Q2ht. The amount of heat transfer in the inter-cell path portion Kt is indicated by Q12.

  The amount of heat transfer from the battery body 22 of the first battery cell 21a to the first tab path portion K1a is indicated by Q1hc, and the amount of heat transfer from the battery body 22 of the second battery cell 21b to the second tab path portion K2a is indicated by Q2hc. Show. For this reason, the amount of heat transfer in each of the bus bar path portion Kb and the substrate path portion Kp is “Q1hc + Q2hc”.

  From the thermal circuit network model shown in FIG. 3, the internal temperature T1 of the first battery cell 21a is expressed by the following equation (eq1).

上式（ｅｑ１）において、各伝熱量Ｑ１ｊ、Ｑ１２，Ｑ１ｈｃ，Ｑ１ｈｔは下式（ｅｑ２）で表される。

In the above equation (eq1), each heat transfer amount Q1j, Q12, Q1hc, Q1ht is represented by the following equation (eq2).

上式（ｅｑ２）において、Ｒ１ｊは第１電池セル２１ａの内部抵抗値を示し、Ｉ（ｔ）は第１，第２電池セル２１ａ，２１ｂに流れる充放電電流を示す。また、上式（ｅｑ２）において、Ｔｓｅｎｓ（ｔ）は第１温度センサ５０の温度検出値を示し、Ｔａｉｒ（ｔ）は第３温度センサ５２の温度検出値を示す。上式（ｅｑ１）を微分した式に上式（ｅｑ２）を代入することにより、下式（ｅｑ３）が導かれる。

In the above equation (eq2), R1j represents the internal resistance value of the first battery cell 21a, and I (t) represents the charge / discharge current flowing through the first and second battery cells 21a, 21b. In the above equation (eq2), Tsens (t) represents the temperature detection value of the first temperature sensor 50, and Tair (t) represents the temperature detection value of the third temperature sensor 52. By substituting the above equation (eq2) into the equation obtained by differentiating the above equation (eq1), the following equation (eq3) is derived.

第２電池セル２１ｂの内部温度Ｔ２は下式（ｅｑ４）で表される。

The internal temperature T2 of the second battery cell 21b is represented by the following formula (eq4).

上式（ｅｑ４）において、各伝熱量Ｑ２ｊ、Ｑ２ｈｃ，Ｑ２ｈｔは下式（ｅｑ５）で表される。

In the above equation (eq4), each heat transfer amount Q2j, Q2hc, Q2ht is represented by the following equation (eq5).

上式（ｅｑ５）において、Ｒ２ｊは第２電池セル２１ｂの内部抵抗値を示す。上式（ｅｑ４）を微分した式に上式（ｅｑ５）を代入することにより、下式（ｅｑ６）が導かれる。

In the above equation (eq5), R2j represents the internal resistance value of the second battery cell 21b. By substituting the above equation (eq5) into an equation obtained by differentiating the above equation (eq4), the following equation (eq6) is derived.

上式（ｅｑ３），（ｅｑ６）から、下式（ｅｑ７）に示す状態方程式が導かれる。

From the above equations (eq3) and (eq6), the equation of state shown in the following equation (eq7) is derived.

上式（ｅｑ７）において、状態変数ｘｒ等は下式（ｅｑ８）で表される。

In the above equation (eq7), the state variable xr and the like are represented by the following equation (eq8).

ここで、下式（ｅｑ９）のように状態変数ｘｅ（ｔ）を新たに定義する。

Here, a state variable xe (t) is newly defined as in the following equation (eq9).

上式（ｅｑ７），（ｅｑ９）に基づいて、下式（ｅｑ１０）に示すディスクリプタ表現における状態方程式が導かれる。

Based on the above equations (eq7) and (eq9), a state equation in the descriptor expression shown in the following equation (eq10) is derived.

上式（ｅｑ１０）において、Ｅｐをディスクリプタ行列、Ａｐをシステム行列、Ｂｐを制御行列ということとする。また、上式（ｅｑ１０）では、第１，第３温度センサ５０，５２の温度検出値が入力変数とされている。本実施形態において、出力行列をＣｐとし、伝達行列をＤｐとすると、出力方程式が下式（ｅｑ１１）のように表される。

In the above equation (eq10), Ep is a descriptor matrix, Ap is a system matrix, and Bp is a control matrix. In the above equation (eq10), the temperature detection values of the first and third temperature sensors 50 and 52 are input variables. In this embodiment, when the output matrix is Cp and the transfer matrix is Dp, the output equation is expressed as the following equation (eq11).

上式（ｅｑ１１）において、出力変数ｙ（ｔ）は状態変数ｘｅ（ｔ）そのものである。上式（ｅｑ１０），（ｅｑ１１）のそれぞれに、観測ノイズｖ（ｔ）及びプロセスノイズｗ（ｔ）の影響を反映させたものを下式（ｅｑ１２），（ｅｑ１３）に示す。

In the above equation (eq11), the output variable y (t) is the state variable xe (t) itself. The following equations (eq12) and (eq13) reflect the effects of the observation noise v (t) and the process noise w (t) in each of the above equations (eq10) and (eq11).

上式（ｅｑ１２），（ｅｑ１３）において、行列Ｇ，Ｈは、プロセスノイズｗ（ｔ）を、制御対象の入力に混入するノイズと制御対象の出力に混入するノイズとのそれぞれに振り分ける重み付けを行うための重み付け行列である。上式（ｅｑ１２），（ｅｑ１３）において、観測ノイズｖ（ｔ）及びプロセスノイズｗ（ｔ）は、白色雑音とする。このため、観測ノイズｖ（ｔ）及びプロセスノイズｗ（ｔ）について下式（ｅｑ１４）が成立するものとする。

In the above equations (eq12) and (eq13), the matrices G and H are weighted to distribute the process noise w (t) to the noise mixed into the control target input and the noise mixed into the control target output, respectively. Is a weighting matrix. In the above equations (eq12) and (eq13), the observation noise v (t) and the process noise w (t) are white noise. Therefore, it is assumed that the following equation (eq14) is established for the observation noise v (t) and the process noise w (t).

また、観測ノイズｖ（ｔ）及びプロセスノイズｗ（ｔ）について、下式（ｅｑ１５）が成立するものとする。

In addition, the following equation (eq15) is established for the observation noise v (t) and the process noise w (t).

上式（ｅｑ１５）において、Ｒｒ，Ｑｒは各ノイズｖ（ｔ），ｗ（ｔ）の共分散行列を示し、Ｎは観測ノイズｖ（ｔ）及びプロセスノイズｗ（ｔ）の相関に係る行列を示す。上式（ｅｑ１５）において、添え字のＴは転置行列であることを示す。図４に、上式（ｅｑ１２），（ｅｑ１３）に基づく制御対象ＣＴＬの状態変数線図を示す。

In the above equation (eq15), Rr and Qr indicate the covariance matrix of each noise v (t) and w (t), and N indicates the matrix related to the correlation between the observed noise v (t) and the process noise w (t). Show. In the above equation (eq15), the subscript T indicates a transposed matrix. FIG. 4 shows a state variable diagram of the control object CTL based on the above equations (eq12) and (eq13).

  The observer equation is represented by the following equation (eq16) for the control object CTL represented by the above equations (eq12) and (eq13).

上式（ｅｑ１６）において、ｘｈ（ｔ）は状態変数ｘｅ（ｔ）の推定値を示し、Ｌはオブザーバゲインを示し、ゲイン行列ともいう。本実施形態において、オブザーバゲインＬは下式（ｅｑ１７）で表される。

In the above equation (eq16), xh (t) represents an estimated value of the state variable xe (t), L represents an observer gain, and is also referred to as a gain matrix. In the present embodiment, the observer gain L is expressed by the following equation (eq17).

上式（ｅｑ１７）において、行列Ｐは代数リカッチ方程式の解を示し、正定行列である。オブザーバゲインＬは、行列Ｐ、出力行列Ｃｐ、重み付け行列Ｇ，Ｈ、各ノイズｖ（ｔ），ｗ（ｔ）の共分散行列Ｒｒ，Ｑｒ、及び各ノイズｖ（ｔ），ｗ（ｔ）の相関に係る行列Ｎｒに基づいて算出される。

In the above equation (eq17), the matrix P represents the solution of the algebraic Riccati equation and is a positive definite matrix. The observer gain L includes the matrix P, the output matrix Cp, the weighting matrices G and H, the covariance matrices Rr and Qr of the noises v (t) and w (t), and the noises v (t) and w (t). It is calculated based on the matrix Nr related to the correlation.

  Here, the internal resistance value R1j of the first battery cell 21a constituting the system matrix Ap changes according to the temperature of the first battery cell 21a, and the internal resistance value R2j of the second battery cell 21b is the second battery cell. It changes according to the temperature of 21b. Therefore, in order to satisfy the robustness required for the observer, the observer gain L that satisfies the secondary stability with respect to the fluctuations of the internal resistance values R1j and R2j is calculated. Also, an observer gain L that satisfies the secondary stability is calculated for each noise v (t), w (t).

  Specifically, for the sake of simplification, the influence of the observation noise v (t) and the process noise w (t) is ignored in the observer state equation. Further, fluctuations of the internal resistance values R1j and R2j constituting the system matrix Ap are shown in a polytope format. Here, the upper and lower limits of the internal resistance value R1j of the first battery cell 21a are R1U and R1L, and the upper and lower limits of the internal resistance value R2j of the second battery cell 21b are R2U and R2L.

  Since the combinations of the internal resistance values R1j and R2j are 2 2 (four ways), the parameter box has four vertices. Using the upper and lower limits of the internal resistance values R1j and R2j that are the vertices of the parameter box, the matrices Ap, Bp, Cp, and Dp shown in the above equation (eq16) are converted into the matrices An, Bn, Cn, and Dn (n = 1, 2, 3, 4). Here, each matrix An, Bn, Cn, and Dn is expressed as the following equation (eq18).

上式（ｅｑ１２）における状態変数ｘｅ（ｔ）と上式（ｅｑ１６）における推定値ｘｈ（ｔ）の間の推定誤差ｅ（ｔ）を下式（ｅｑ１９）で表す。

An estimation error e (t) between the state variable xe (t) in the above equation (eq12) and the estimated value xh (t) in the above equation (eq16) is represented by the following equation (eq19).

上式（ｅｑ１０），（ｅｑ１６）をポリトープ形式で表現したものと、上式（ｅｑ１９）とに基づいて、誤差ダイナミクスとして下式（ｅｑ２０）が導かれる。

Based on the above expressions (eq10) and (eq16) expressed in a polytope form and the above expression (eq19), the following expression (eq20) is derived as the error dynamics.

上式（ｅｑ２０）で表される誤差ダイナミクスの２次安定性を示すため、下式（ｅｑ２１）で表されるリアプノフ関数を用いる。

In order to show the secondary stability of the error dynamics represented by the above equation (eq20), the Lyapunov function represented by the following equation (eq21) is used.

２次安定性を示すためには、上式（ｅｑ２１）の微分値が負定であればよい。上式（ｅｑ２１）を微分すると、下式（ｅｑ２２）が導かれる。

In order to show the secondary stability, the differential value of the above equation (eq21) may be negative. Differentiating the above equation (eq21) leads to the following equation (eq22).

上式（ｅｑ２２）において「Ｓ＝Ｐ＾２」とすると、下式（ｅｑ２３）で表される線形行列不等式（ＬＭＩ）が導かれる。

When “S = P ^ 2” in the above equation (eq22), a linear matrix inequality (LMI) represented by the following equation (eq23) is derived.

ここで、推定誤差ｅ（ｔ）の収束率をαとする。収束率αを考慮すると、上式（ｅｑ２３）から、収束率α及びディスクリプタ行列Ｅｐを含む下式（ｅｑ２４）が導かれる。

Here, the convergence rate of the estimation error e (t) is α. In consideration of the convergence rate α, the following equation (eq24) including the convergence rate α and the descriptor matrix Ep is derived from the above equation (eq23).

なお本実施形態において、収束率αは、下式（ｅｑ２５）に示すように、推定誤差ｅ（ｔ）の振幅の減衰特性として定義される。

In the present embodiment, the convergence rate α is defined as the attenuation characteristic of the amplitude of the estimation error e (t) as shown in the following equation (eq25).

上式（ｅｑ２４）で表されるＬＭＩをパラメータボックスの各頂点で解くことにより、行列Ｐが算出される。算出された行列Ｐを上式（ｅｑ１７）に入力することにより、各内部抵抗値Ｒ１ｊ，Ｒ２ｊの変動、及び各ノイズｖ（ｔ），ｗ（ｔ）に対して２次安定性を満たすオブザーバゲインＬが算出される。

The matrix P is calculated by solving the LMI expressed by the above equation (eq24) at each vertex of the parameter box. By inputting the calculated matrix P into the above equation (eq17), the observer gain satisfying the secondary stability with respect to the fluctuations of the internal resistance values R1j and R2j and the noises v (t) and w (t). L is calculated.

  Note that the internal temperatures of the third and fourth battery cells 21c and 21d can also be estimated by a method similar to the estimation method of the internal temperatures of the first and second battery cells 21a and 21b.

<About temperature estimation processing>
Then, the temperature estimation process which the control part 60 performs is demonstrated. FIG. 4 shows an observer 61 provided in the control unit 60. In the present embodiment, since the transfer matrix Dp is a zero matrix, the transfer matrix Dp is not shown in FIG.

  In the observer 61, the deviation calculation unit 61a subtracts the matrix Cpxh (t) output from the output multiplication unit 61b from the output variable y (t).

  The gain processing unit 61c calculates the observer gain L, and multiplies the calculated observer gain L by the matrix “y (t) −Cpxh (t)” output from the deviation calculation unit 61a. The adder 61d outputs the matrix “L (y (t) −Cpxh (t))” output from the gain processor 61c, the matrix Bnu (t) output from the control multiplier 61e, and the system multiplier 61f. The output matrix Anxh (t) is added. Here, the input variable u (t) includes the temperature detection values Tsens and Tair of the first and third temperature sensors 50 and 52.

  The coefficient multiplication unit 61g multiplies the descriptor matrix Ep by the matrix “Anxh (t) + Bnu (t) + L (y (t) −Cpxh (t))” output from the addition unit 61d. The integrator 61h integrates the matrix “Ep {Anxh (t) + Bnu (t) + L (y (t) −Cxh (t) −Du (t))}” output from the coefficient multiplication unit 61g. An estimated value xh (t) of the state variable is calculated.

  Incidentally, the parameters R12, R1hc, R1ht, R2hc, R2ht, C1, and C2 included in the system matrix An of the system multiplier 61f and the control matrix Bn of the control multiplier 61e are sequentially identified. However, since identification of these parameters is not a main part in the present embodiment, detailed description of the identification method is omitted.

  The gain processing unit 61c corresponds to a gain calculation unit that calculates the observer gain L based on the above equation (eq17). Specifically, the gain processing unit 61c first calculates the matrix P by solving the LMI expressed by the above equation (eq24). The gain processing unit 61c calculates the observer gain L based on the above equation (eq17) with the calculated matrix P as an input.

  Here, in the present embodiment, the gain processing unit 61c calculates the observer gain L so as to satisfy the robustness required for the observer in all the fluctuation ranges that the battery cell 21 can take. This can be realized by setting the fluctuation range of the internal resistance values R1j and R2j to be considered when calculating the matrix P by the setting unit 62. Hereinafter, the setting unit 62 will be described.

  The internal resistance values R1j and R2j of the first and second battery cells 21a and 21b increase as the temperature of the battery cell 21 decreases. For this reason, the fluctuation range of the internal resistance values R1j and R2j to be considered when calculating the matrix P is changed according to the temperature of the battery cell 21. Here, in the present embodiment, the battery unit 10 is not provided with a temperature sensor that directly detects the temperature of the battery cell 21a. For this reason, the fluctuation range of the internal resistance values R1j and R2j to be considered when calculating the matrix P is changed according to the temperature detection value Tair of the third temperature sensor 52. In the present embodiment, the fluctuation range can be changed based on the temperature detection value Tair of the third temperature sensor 52 because the temperature detection value Tair and the temperature of the battery cell 21 have a positive correlation during the operation of the battery unit 10. It is for having. Due to this correlation, the temperature detection value Tair and the internal resistance values R1j and R2j can be related.

  As shown in FIG. 5, in the fluctuation range TB that can be taken by the temperature detection value Tair of the third temperature sensor 52 from the temperature T0 to the temperature T7, the setting unit 62 sets each temperature set by dividing the fluctuation range TB. A temperature range including the temperature detection value Tair of the third temperature sensor 52 is selected from the ranges TA1 to TA7. The setting unit 62 sets the upper and lower limit values R1L, R1U, R2L, R2U of the internal resistance values R1j, R2j corresponding to the boundary of the selected temperature range. In FIG. 5, when the temperature detection value Tair is included in the temperature range TA4, the upper limit value R1U of the internal resistance value of the first battery cell 21a is set as RU, and the lower limit value R1L of the internal resistance value is set as RL. An example is shown.

  As shown in FIG. 5, the LMI is solved by setting the upper and lower limit values of the internal resistance values R1j and R2j in each of the temperature ranges TA1 to TA7 obtained by dividing the fluctuation range TB that the temperature detection value Tair can take. The fluctuation range of the internal resistance value to be considered when calculating the matrix P can be narrowed. In particular, in the present embodiment, each internal resistance value R1j, R2j has a larger increase amount per unit temperature decrease amount of the battery cell 21 as the temperature of the battery cell 21 is lower. Each is set narrower as the temperature detection value Tair is lower. Thereby, the fluctuation range of the internal resistance value to be considered can be narrowed. As a result, the observer gain L satisfying the robustness in each of the divided temperature ranges TA1 to TA7 can be calculated, and as a result, the observer robustness can be satisfied in the entire fluctuation range TB.

  In the present embodiment, the setting unit 62 stores a convergence rate α set in advance corresponding to each of the temperature ranges TA1 to TA7. The setting unit 62 selects a convergence rate α corresponding to a temperature range in which the temperature detection value Tair of the third temperature sensor 52 is included among the temperature ranges TA1 to TA7 and outputs the convergence rate α to the gain processing unit 61c. Thereby, in each temperature range TA1-TA7, the convergence rate (alpha) according to the temperature of the battery cell 21 can be used so that the time for the estimated internal temperature to converge to true internal temperature can be shortened.

  Incidentally, the convergence rate α may be the same value or a different value in each of the temperature ranges TA1 to TA7. The gain processing unit 61c calculates the observer gain L so that the eigenvalue of the matrix “An-LCp” is negative.

  Thus, according to the present embodiment, the robustness of the observer that estimates the internal temperature of the battery cell 21 can be satisfied in all the fluctuation range TB that the temperature detection value Tair can take. For this reason, it is not necessary to adapt the observer for each use environment of the vehicle such as a cold region specification or a warm region specification, and a common observer can be used.

  According to the embodiment described in detail above, the following effects can be obtained.

  A plurality of temperature ranges TA <b> 1 to TA <b> 7 are determined by dividing a fluctuation range TB that can be assumed as the temperature of the battery cell 21. In each of the temperature ranges TA1 to TA7, the observer gain L was calculated so as to satisfy the robustness with the upper and lower limit values of the internal resistance values R1j and R2j corresponding to the temperature range. In other words, the observer gain L is scheduled according to the temperature of the battery cell 21. Therefore, even when the variation range TB is wide, the LMI solution P satisfying the robustness can be calculated in each of the temperature ranges TA1 to TA7. Thereby, the observer gain L satisfying the robustness in each of the temperature ranges TA1 to TA7 can be calculated, and the estimation accuracy of the internal temperature of the battery cell 21 can be increased.

  In each of the temperature ranges TA1 to TA7, the convergence rate α of the estimated value xh (t) of the state variable was individually set. For this reason, the convergence rate α can be optimized in each of the temperature ranges TA1 to TA7. Thereby, even when there is an estimation error between the estimated internal temperature and the true internal temperature, the estimation error can be quickly converged to zero.

  Each of the temperature ranges TA1 to TA7 is set to be narrower as the temperature detection value Tair of the third temperature sensor 52 is lower. For this reason, in the temperature range where temperature detection value Tair is low, it can suppress that the difference of the upper and lower limit values of each internal resistance value R1j and R2j in the boundary of the temperature range becomes large. Thereby, in each temperature range TA1-TA7, the matrix P of LMI shown by the above formula (eq24) can be calculated, and the observer gain L satisfying the robustness can be calculated.

  The LMI represented by the above equation (eq24) includes matrices Rb and Nb related to noise. Thereby, even when there is observation noise and process noise, the stability of the closed loop including the gain processing unit 61c in the observer can be guaranteed. Therefore, the estimation accuracy of the internal temperature can be further increased, such as avoiding the divergence of the estimated value of the internal temperature of the battery cell 21.

(Second Embodiment)
Hereinafter, the second embodiment will be described with reference to the drawings with a focus on differences from the first embodiment. In the present embodiment, the temperature estimation process described in the first embodiment is referred to as a first temperature estimation process. In the present embodiment, the control unit 60 performs a second temperature estimation process in addition to the first temperature estimation process. And in this embodiment, according to the estimated temperature of the battery cell 21, the switching process which switches which temperature estimation is used among the 1st, 2nd temperature estimation processes is performed. Hereinafter, after describing the second temperature estimation process, the switching process will be described.

<About the second temperature estimation process>
In the present embodiment, the second temperature estimation process is performed using the thermal circuit network model shown in FIG. The thermal network model shown in FIG. 6 is a simplified version of the model shown in FIG. Specifically, the first and second spatial path portions K1b and K2b, the inter-cell path portion Kt, and the heat capacities C1 and C2 are deleted from the model shown in FIG. Hereinafter, the first and second battery cells 21a and 21b will be described as an example.

  The temperature difference at both ends of the first tab path portion K1a is represented by Q1j × R1hc, and the temperature difference at both ends of the second tab path portion K2a is represented by Q2j × R2hc. Further, the temperature difference between both ends of the substrate path portion Kp is represented by (Q1j + Q2j) × Rp. Therefore, the internal temperatures of the first and second battery cells 21a and 21b can be expressed by the following equation (eq26).

また、第１，第３温度センサ５０，５２の温度検出値の差は「Ｔｓｅｎｓ−Ｔａｉｒ」で表すことができる。この温度差「Ｔｓｅｎｓ−Ｔａｉｒ」は、基板経路部Ｋｐの両端における温度差に相当する。また、基板経路部Ｋｐにおける伝熱量は「Ｑ１ｊ＋Ｑ２ｊ」であるため、「Ｑ１ｊ＋Ｑ２ｊ」を下式（ｅｑ２７）で表すことができる。

The difference between the temperature detection values of the first and third temperature sensors 50 and 52 can be expressed as “Tsens-Tair”. This temperature difference “Tsens-Tair” corresponds to a temperature difference at both ends of the substrate path portion Kp. Further, since the heat transfer amount in the substrate path portion Kp is “Q1j + Q2j”, “Q1j + Q2j” can be expressed by the following equation (eq27).

上式（ｅｑ２７）を上式（ｅｑ２６）に代入すると、下式（ｅｑ２８）が導かれる。

Substituting the above equation (eq27) into the above equation (eq26) leads to the following equation (eq28).

これにより、伝熱経路上の各伝熱抵抗Ｒｂ，Ｒｐ，Ｒ１ｈｃ，Ｒ２ｈｃと、第１，第３温度センサ５０，５２の温度検出値Ｔｓｅｎｓ，Ｔａｉｒとがわかれば、それらに基づいて第１，第２電池セル２１ａ，２１ｂの内部温度Ｔ１，Ｔ２を算出することができる。

As a result, if the heat transfer resistances Rb, Rp, R1hc, R2hc on the heat transfer path and the temperature detection values Tsens, Tair of the first and third temperature sensors 50, 52 are known, the first, The internal temperatures T1, T2 of the second battery cells 21a, 21b can be calculated.

  Next, the second temperature estimation process based on the above-described thermal circuit network model will be described with reference to FIG. Each functional block in FIG. 7 is realized by the control unit 60. In the second temperature estimation process, a process for estimating the internal temperatures of the first and second battery cells 21a and 21b and a process for estimating the internal temperatures of the third and fourth battery cells 21c and 21d are individually performed. Done.

  As shown in FIG. 7, the first temperature difference calculation unit 71 subtracts the temperature detection value Tair of the third temperature sensor 52 from the temperature detection value Tsens of the first temperature sensor 50, so that the temperature difference between these temperature detection values. ΔT is calculated.

  Based on the temperature difference ΔT calculated by the first temperature difference calculation unit 71, the second temperature difference calculation unit 72 is a temperature between the internal temperature T1 of the first battery cell 21a and the temperature detection value Tsens of the first temperature sensor 50. The difference is calculated as y1. This temperature difference g1 is expressed by the following equation (eq29).

温度差ｇ１は、上式（ｅｑ２８）のＴ１についての第２項と第３項との和に相当する。また、上式（ｅｑ２９）の各伝熱抵抗Ｒｂ，Ｒｐ，Ｒ１ｈｃはいずれも適合値とされ、またＱ１ｈｃ×Ｒ１ｈｃについても適合値とされている。なお本実施形態において、温度差ｇ１が「推定温度差」に相当し、第２温度差算出部７２が温度差算出部に相当する。

The temperature difference g1 corresponds to the sum of the second term and the third term for T1 in the above equation (eq28). In addition, each of the heat transfer resistances Rb, Rp, R1hc in the above equation (eq29) is an appropriate value, and Q1hc × R1hc is also an appropriate value. In the present embodiment, the temperature difference g1 corresponds to an “estimated temperature difference”, and the second temperature difference calculation unit 72 corresponds to a temperature difference calculation unit.

  The adding unit 73 adds the temperature difference g1 calculated by the second temperature difference calculating unit 72 and the temperature detection value Tsens of the first temperature sensor 50. Thereby, the internal temperature T1 of the first battery cell 21a is calculated. In the present embodiment, the adding unit 73 corresponds to a battery temperature calculating unit.

  Based on the temperature difference ΔT calculated by the first temperature difference calculation unit 71, the third temperature difference calculation unit 74 is a temperature between the internal temperature T2 of the second battery cell 21b and the temperature detection value Tsens of the first temperature sensor 50. The difference is calculated as g2. This temperature difference g2 is expressed by the following equation (eq30).

温度差ｇ２は、上式（ｅｑ２８）のＴ２についての第２項と第３項との和に相当する。また、上式（ｅｑ３０）の伝熱抵抗Ｒｂ，Ｒｐ，Ｒ２ｈｃはいずれも適合値とされ、またＱ２ｈｃ×Ｒ２ｈｃについても適合値とされている。なお本実施形態において、温度差ｙ２が「推定温度差」に相当し、第３温度差算出部７４が温度差算出部に相当する。

The temperature difference g2 corresponds to the sum of the second term and the third term for T2 in the above equation (eq28). In addition, the heat transfer resistances Rb, Rp, R2hc in the above equation (eq30) are all appropriate values, and Q2hc × R2hc is also an appropriate value. In the present embodiment, the temperature difference y2 corresponds to an “estimated temperature difference”, and the third temperature difference calculation unit 74 corresponds to a temperature difference calculation unit.

  The adding unit 75 adds the temperature difference g2 calculated by the third temperature difference calculating unit 74 and the temperature detection value Tsens of the first temperature sensor 50. Thereby, the internal temperature T2 of the second battery cell 21b is calculated. In the present embodiment, the adding unit 75 corresponds to a battery temperature calculating unit. Further, for the third and fourth battery cells 21c and 21d, the internal temperature can be estimated by the same method as that shown in FIG. In this case, the heat transfer resistance and heat transfer amount in the above equations (eq29) and (eq30) are changed to the heat transfer resistance and heat transfer amount corresponding to the third and fourth battery cells 21c and 21d, and the first temperature sensor 50 Instead of the detected temperature value, the detected temperature value of the second temperature sensor 51 may be Tsens.

<About switching processing>
FIG. 8 shows the procedure of the switching process for switching the temperature estimation between the first and second temperature estimation processes. This process is repeatedly executed by the control unit 60 at a predetermined cycle, for example. In the present embodiment, in the process shown in FIG. 8, the internal temperature Te used in the first step S10 or the first steps S10 and S11 is estimated by a predetermined one of the first and second temperature estimation processes. Values are used. Specifically, for example, the internal temperature estimated by the first temperature estimation process with high estimation accuracy is used.

  In this series of processes, first, in step S10, it is determined whether or not the internal temperature Te estimated by the currently executed estimation process out of the first and second temperature estimation processes exceeds the upper limit temperature Tmax. The upper limit temperature Tmax is set to a value smaller than the upper limit value of the temperature at which the reliability of the battery cell 21 can be maintained (hereinafter referred to as “allowable upper limit value TUlim”).

  If a negative determination is made in step S10, the process proceeds to step S11, and whether or not the internal temperature Te estimated by the currently executed estimation process of the first and second temperature estimation processes is less than the lower limit temperature Tmin. Determine. The lower limit temperature Tmin is set to a value smaller than the upper limit temperature Tmax and larger than the lower limit value of the temperature at which the reliability of the battery cell 21 can be maintained (hereinafter referred to as “allowable lower limit value TLlim”). ing.

  If a negative determination is made in step S11, the process proceeds to step S12 to perform a second temperature estimation process. On the other hand, when an affirmative determination is made in steps S10 and S11, the process proceeds to step S13 to perform a first temperature estimation process.

  Incidentally, in this embodiment, the temperature range from the lower limit temperature Tmin to the upper limit temperature Tmax corresponds to the high reliability temperature range. Further, the temperature range from the lower limit temperature Tmin to the allowable lower limit value TLlim and the temperature range from the upper limit temperature Tmax to the allowable upper limit value TUlim correspond to the low reliability temperature range. In the present embodiment, the reliability of the battery cell 21 in the low reliability temperature range is lower than the high reliability temperature range. For example, the reliability temperature is lower than the maximum value of the discharge capacity of the battery cell 21 in the high reliability temperature range. It means that the maximum value of the discharge capacity of the battery cell 21 in the range is smaller.

  In the present embodiment, the two estimation processes are switched according to the estimated temperature for the reason described below.

  In the present embodiment, the estimation accuracy of the internal temperature by the first temperature estimation process is higher than the estimation accuracy of the internal temperature by the second temperature estimation process. Further, the processing load required for the estimation of the internal temperature by the second temperature estimation process is smaller than the processing load required for the estimation of the internal temperature by the first temperature estimation process. Therefore, the estimation of the internal temperature Te by the second temperature estimation process is continued until the internal temperature Te estimated by the second temperature estimation process exceeds the upper limit temperature Tmax or falls below the lower limit temperature Tmin. As a result, the processing load on the control unit 60 can be reduced until the internal temperature Te estimated by the second temperature estimation process exceeds the upper limit temperature Tmax or falls below the lower limit temperature Tmin.

  On the other hand, when the internal temperature Te estimated by the second temperature estimation process exceeds the upper limit temperature Tmax or falls below the lower limit temperature Tmin, the second temperature estimation process is switched to the first temperature estimation process. Therefore, when the internal temperature Te increases and approaches the allowable upper limit value TUlimit, or when the internal temperature Te decreases and approaches the allowable lower limit value TLlimit, the estimation accuracy of the internal temperature Te can be increased. Thereby, it can avoid using the battery cell 21 in the overheated state or the low temperature state. Therefore, deterioration of the battery cell 21 can be avoided.

(Other embodiments)
Each of the above embodiments may be modified as follows.

  In the first embodiment, in each of the temperature ranges TA1 to TA7, the temperature ranges TA1 to TA7 are set so that the difference between the upper and lower limits of the internal resistance values R1j and R2j corresponding to the boundary of the temperature range is equal to each other. Each temperature range of TA7 may be set.

  In the second embodiment, the first temperature estimation process may be switched only when the internal temperature Te estimated by the second temperature estimation process exceeds the upper limit temperature Tmax. Alternatively, the first temperature estimation process may be switched only when the internal temperature Te estimated by the second temperature estimation process falls below the lower limit temperature Tmin.

  The second temperature estimation process is not limited to that exemplified in the second embodiment. If the processing load is lower than the first temperature estimation process, the process may be different from the process exemplified in the second embodiment.

  In the first embodiment, the upper and lower limit values of the internal resistance values of the first and second battery cells 21a and 21b are set based on the temperature detection value Tair of the third temperature sensor 52, but the present invention is not limited to this. For example, when there is a positive correlation between the internal resistance value and the temperature at the installation position of the first temperature sensor 50, the upper and lower limit values of the internal resistance value are set based on the temperature detection value Tsens of the first temperature sensor 50. May be.

  In the first embodiment, the thermal network model is not limited to that shown in FIG. For example, a model in which any one of the inter-cell route portion Kt and the space route portions K1b and K2b is omitted may be used.

  The arrangement method of the control board 30 in the housing case 40 is not limited to that shown in FIG. For example, the control substrate 30 may be disposed between the assembled battery 20 and the peripheral wall portion 42 with the substrate surface facing the inner surface of the peripheral wall portion 42.

  Further, the arrangement method of the battery cells 21 in the housing case 40 is not limited to that shown in FIG. For example, a vertical arrangement in which the battery cells are arranged in the housing case 40 with the plate surface of the battery cells 21 facing the inner surface of the peripheral wall portion 42 may be adopted.

  -In 1st Embodiment, although the internal temperature of the battery cell was estimated by the observer, it is not restricted to this. For example, the surface temperature of the battery cell (the surface temperature of the flat container 25) may be estimated.

  DESCRIPTION OF SYMBOLS 10 ... Battery unit, 21 ... Battery cell, 60 ... Control part, 61 ... Observer, 62 ... Setting part.

Claims (7)

Temperature detection provided on a battery (21), a current detector (53) for detecting a current flowing through the battery, and a heat transfer path (K1a, Kb, Kp) through which heat generated in the battery is transmitted Part (50-52), and a battery unit (10) comprising:
From a thermal circuit network model that includes the current value flowing through the battery and the temperature at a predetermined position of the heat transfer path as input variables and the temperature of the battery as a state variable, and models heat transfer in the heat transfer path. An observer (61) for sequentially estimating the temperature of the battery in the derived state equation based on each of the current detection value of the current detection unit and the temperature detection value of the temperature detection unit;
The thermal network model includes the internal resistance of the battery having a larger resistance value as the temperature of the battery is lower,
The variation range that the temperature detection value of the temperature detection unit can take is divided, and a temperature range including the temperature detection value is selected from a plurality of set temperature ranges, and the temperature range corresponding to the boundary of the selected temperature range is selected. A setting unit (62) for setting upper and lower limits of the internal resistance value of the battery;
A battery temperature estimation device comprising: a gain calculation unit (61c) that calculates an observer gain used by the observer based on the upper and lower limits of the internal resistance value set by the setting unit.   The convergence ratio for converging the battery temperature estimated by the observer to the true value is individually set in each of the temperature ranges set by dividing the fluctuation range. Item 2. The battery temperature estimation device according to Item 1. The internal resistance value, the lower the temperature of the battery, the greater the increase amount per unit temperature decrease amount of the battery,
3. The battery temperature estimation device according to claim 1, wherein each of the temperature ranges set by dividing the fluctuation range is set to be narrower as the temperature detection value is lower. 4.   The gain calculation unit calculates the observer gain that stabilizes the closed loop of the observer with respect to fluctuations in the internal resistance value in each of a plurality of temperature ranges set by dividing the fluctuation range. The battery temperature estimation apparatus according to any one of claims 1 to 3.   The gain calculation unit calculates the observer gain that stabilizes the closed loop of the observer with respect to process noise and observation noise mixed in a signal in the thermal network model to be controlled. The battery temperature estimation apparatus of any one of -4. A processing load that is smaller than the processing load required for the temperature estimation of the battery by the observer, and a different estimation unit that estimates the battery temperature by a method different from the battery temperature estimation method by the observer;
The temperature estimation accuracy by the observer is higher than the temperature estimation accuracy by the separate estimation unit,
The temperature range with high reliability of the battery is a high reliability temperature range,
Adjacent to the high reliability temperature range and a temperature range in which the reliability of the battery is lower than the high reliability temperature range is a low reliability temperature range,
When the temperature range including the temperature estimated by the separate estimation unit is the high reliability temperature range, the temperature estimation by the separate estimation unit is continued, and the temperature range including the temperature estimated by the separate estimation unit is The battery temperature estimation device according to any one of claims 1 to 5, further comprising a switching unit that switches to temperature estimation by the observer when the high reliability temperature range is changed to the low reliability temperature range. The temperature detection unit is on the heat transfer path, the short distance temperature detection unit (50, 51) provided at a position where the path length of heat transfer from the battery is small, and on the heat transfer path. A long-distance temperature detector (52) provided at a position where the path length of heat transfer from the battery is large,
In the heat transfer path, a first heat transfer resistance (R1hc, R2hc, Rb) exists in a first path part (K1a, K2a, Kb) between the battery and the short-range temperature detection unit, and the near A second heat transfer resistance (Rp) exists in the second path part (Kp) between the distance temperature detection part and the long distance temperature detection part,
The separate estimation unit is:
Based on the detected temperature difference, which is the difference between the temperature detection values of the short distance temperature detection unit and the long distance temperature detection unit, and the first heat transfer resistance and the second heat transfer resistance, the short distance temperature A temperature difference calculator that calculates an estimated temperature difference that is a temperature difference between the installation position of the detector and the battery; and
A battery temperature calculation unit that calculates the temperature of the battery by adding the estimated temperature difference calculated by the temperature difference calculation unit to the temperature detection value of the short-range temperature detection unit. The battery temperature estimation apparatus according to claim 6.

Images