CN108616977B - UWB (ultra wide band) multipoint time delay correction method based on least square adjustment - Google Patents

Abstract

The invention relates to a UWB (ultra-wide band) multipoint time delay correction method based on least square adjustment, which comprises the steps of firstly, selecting a plurality of uniformly distributed control points on the ground, collecting static data on the control points, and solving the relative signal propagation time delay quantity among UWB base stations by adopting the least square adjustment idea according to the obtained static observation data, the control points and the coordinates of the UWB system base stations by combining a GNSS data processing method so as to realize the time synchronization among the UWB base stations. The invention is designed aiming at the problem of asynchronous clocks of all base stations in UWB positioning, and is suitable for a UWB positioning system which adopts a synchronous control device to carry out time synchronous control. When the delay of the signal propagation time among the base stations is calculated, the observation values of the control points are selected to be jointly solved, compared with the solving of a single control point, the influence of various observation errors on time delay correction is reduced, the calculation result is more accurate, and the reliability is higher.

Description

A UWB Multi-point Delay Correction Method Based on Least Squares Adjustment

technical field

The invention relates to the field of ultra-wide band (UWB: ultra-wide band) indoor navigation and positioning technology data processing, in particular to a time delay correction method for realizing clock synchronization between UWB base stations.

Background technique

In the past few decades, high-precision outdoor navigation and positioning technologies such as Global Navigation Satellite System (GNSS) have become more and more mature, and the accuracy has been improved from the initial meter level to the centimeter level. With the development of precision positioning technology, GNSS technology has been widely used in both military and civilian fields. At the same time, high-precision navigation and positioning technology in indoor environment has also become a research hotspot of major research institutions and enterprises. Compared with outdoor positioning, the structure of indoor environment is more complex, people and obstacles are denser, and there is no stable and universal sensor system. Currently commonly used indoor positioning technologies include Ultra-wideband (Ultra-wideband), Radio-Frequency Identification (RFID), Infrared-Ray (IR), Ultrasound-Wave (Ultrasound-Wave), Bluetooth and Wireless Local Area Networks, WLAN) and other technologies.

Among them, UWB technology is a wireless communication technology developed by the US military in the mid-20th century, and gradually began to turn to civilian use in the early 21st century. UWB technology has the characteristics of high positioning accuracy (up to within 10cm), high transmission rate (up to 1Gbit/s), large space capacity, low power consumption, good concealment, and strong anti-interference ability. It can meet the needs of some high-precision indoor positioning scenarios, such as industrial measurement, military training, personnel supervision, large venue navigation and other fields.

Similar to the principle of satellite positioning, the UWB system also performs positioning by measuring the propagation time of the signal from the tag to each base station. Since each base station has its own independent clock source, and different crystal oscillators used in the clock source of each base station have different frequency deviations, the frequency deviation is not a constant. Therefore, the time of arrival (TOA) of the tag signal measured by each base station does not have the same time reference. If the time synchronization problem is not dealt with, the observation data will be invalid or the control error will be caused, and the accurate positioning will not be possible.

Domestic and foreign scholars have done a lot of research on the problem of UWB clock synchronization. Most of them are for UWB ranging protocols, such as the classic two-way ranging method, asymmetric bilateral two-way ranging, symmetric bilateral two-way ranging, and some improved ranging protocols on this basis. Decawave has also studied the introduction of a synchronization control device in addition to the clocks of each base station. The data is sent from the tag and received by the base station and then sent to the synchronization control device to realize the synchronization of each clock. However, most of these methods are from the perspective of communication technology, lack of data analysis in the calibration process, and do not take into account the effects of various random errors and accidental errors on time synchronization calibration.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a UWB multi-point time delay correction method, which can deal with the problem of time synchronization of UWB base stations in applications such as UWB indoor navigation and positioning.

Referring to the idea of GNSS satellite clock error calculation and combining with GNSS data processing theory, the present invention proposes a UWB multi-control point time delay correction method based on a synchronous control device. The basic idea is to select a number of discretely distributed known points as control points, and by observing static data for a period of time at each control point, combined with GNSS least squares adjustment and gross error detection methods, to obtain each base station relative to the reference base station. The signal propagation time, that is, the amount of time delay. Compared with the traditional time delay correction with a single control point, this method weakens the influence of observation errors in different observation environments, making the time delay calculation results more accurate and more reliable.

In order to achieve the above-mentioned purpose, the method technical scheme that the present invention provides is:

A UWB multi-point delay correction method based on least squares adjustment, comprising:

Step L1: select several uniformly distributed discrete control points within the coverage of the UWB system, and obtain the coordinates of the control points and the coordinates of each base station of the UWB system;

Step L2: place labels on the laid control points, and collect static data on each control point;

Step L3: Preprocess the original observation data in combination with the prior environment information in the UWB system deployment scene, select a reference base station, and construct a differential observation value;

Step L4: performing systematic error correction on the static observation data of each control point described in Step L2 in combination with the existing error correction model;

Step L5: Combined with the GNSS data processing method, according to the observation values of each control point obtained in step L4 , obtain the delay value corresponding to each control point, and then obtain the relative delay amount between each UWB base station.

Optionally, in the above UWB multi-point delay correction method, the step L1 includes:

Select several discrete control points with better observation conditions and uniform distribution in the UWB system deployment environment;

All base stations can be effectively observed by at least one control point;

It is ensured that every two control points have at least one common base station that can be effectively observed.

Optionally, in the above UWB multi-point delay correction method, the data preprocessing in step L3 includes:

Considering the occlusion of the indoor environment, remove the coordinates of the base station that cannot be seen from each control point;

The reference base station corresponding to each control point is selected as the reference for the difference between stations.

Optionally, in the above-mentioned UWB multi-point delay correction method, the selection of the reference base station in the step L3 includes:

Ensure that the reference base station and the control point can communicate with each other;

Ensure that the reference base station can be effectively observed by at least two control points including the control point at the same time;

Select the base station near the control point as the reference base station as much as possible;

Each control point selects the same reference base station as much as possible.

Optionally, in the above UWB multi-point delay correction method, the step L4 includes:

Correction of antenna phase center deviation and other system errors of UWB base station antenna.

Optionally, in the above UWB multi-point delay correction method, the step L5 includes:

According to the joint solution of the observed values of each control point, the method of least squares adjustment is used to obtain the relative time delay between each UWB base station;

In the solution process, the gross error detection theory is used to eliminate the observations with large deviation, and the final delay value is iteratively solved.

To sum up, based on the UWB synchronous control device, the present invention introduces redundant observations through multi-control point least-squares adjustment and joint time-delay calculation, reduces the influence of various observation errors on time-delay correction, and improves Accuracy and Stability of Delay Correction

Specifically, compared with the prior art, the present invention has the following advantages:

The existing UWB clock synchronization technology is mostly aimed at the field of electronic communication, and does not consider the data processing method in the time delay correction process. Based on the clock synchronization control device, the present invention proposes a multi-control point time delay correction method. Compared with the traditional single-point delay correction method, this method refers to the GNSS satellite clock error calculation method, and calibrates by arranging multiple discretely distributed control points, and introduces redundant observations; and the least squares adjustment method is used to calculate the final It can weaken the influence of various observation errors caused by different observation environments to a certain extent, and improve the accuracy of time delay correction; in the calculation process, the influence of indoor environmental object occlusion is considered, and gross error detection is adopted. The method eliminates the observations with large errors, which ensures the stability and availability of the calculation results.

Description of drawings

1 is a schematic flowchart of a UWB multi-point delay calibration method in a preferred embodiment of the present invention;

Fig. 2 is the specific flow chart of step S3 in Fig. 1;

FIG. 3 is a schematic diagram of a specific flow of step S5 in FIG. 1 .

Detailed ways

The specific embodiments of the present invention will be described in more detail below with reference to the schematic diagrams. The advantages and features of the present invention will become apparent from the following description and claims. It should be noted that, the accompanying drawings are all in a very simplified form and in inaccurate scales, and are only used to facilitate and clearly assist the purpose of explaining the embodiments of the present invention.

Referring to FIG. 1, in a preferred embodiment of the present invention, a multi-point extension correction method based on least squares adjustment includes:

Step S1: select several uniformly distributed discrete control points within the coverage of the base station of the UWB system, and obtain the coordinates of the control points and the coordinates of each base station of the UWB system;

Specifically, pay attention to the following when selecting control points:

Select several discrete control points with better observation conditions and uniform distribution in the UWB system deployment environment;

All base stations can be effectively observed by at least one control point;

It is ensured that every two control points have at least one common base station that can be effectively observed.

Step S2: place labels on the laid control points to obtain static data on each control point (step S2 is a conventional technique in the art):

Its basic observation equation is:

表示标签k收到的基站i的观测值。此处应注意，有同步控制装置的情况下，观测值实际为信号由标签经基站到达同步控制装置的时间。

为标签和基站之间的距离，

表示基站i的天线相位偏差，dt_k为标签的钟差，δt为同步控制装置的钟差，rⁱ为第i个基站到同步控制装置之间的时延量，

为观测噪声，c表示光速。Among them, the subscript k represents the kth label, the superscript i represents the ith base station,

represents the observation value of base station i received by tag k. It should be noted here that in the case of a synchronization control device, the observed value is actually the time when the signal arrives at the synchronization control device from the tag via the base station.

is the distance between the tag and the base station,

Represents the antenna phase deviation of base station i, dt _k is the clock difference of the label, δt is the clock difference of the synchronization control device, ri is the delay amount between the ⁱ -th base station and the synchronization control device,

is the observation noise, and c is the speed of light.

Since the original observation value may increase in number, the method of the difference of observation values between each base station is usually used for calculation of time delay correction and positioning solution.

Step S3: Data preprocessing to construct differential observations.

Specifically, referring to FIG. 2 , step S3 includes:

Step S31: Delete the observation values corresponding to the base stations that cannot be seen from each control point.

Specifically, the influence of occlusion by obstacles in the environment should be considered.

Optionally, the vertex coordinates of the obstacle can be obtained first, and the method of judging whether the line segments intersect by the outer product can be used to determine whether the base station signal is blocked. Let the coordinates of the two adjacent vertices of the obstacle be n, n+1, the label coordinate be k, and the base station coordinate be i, if it satisfies:

(nk×ni)*((n+1)k×(n+1)i)<0 and (2)

(kn×k(n+1))*(in×i(n+1))<0

Among them, nk represents the vector formed by connecting the two points n and k, ni represents the vector formed by connecting the two points n and i, (n+1)k represents the vector formed by connecting the two points n+1 and k, (n +1)i represents the vector formed by connecting the two points of n+1 and i. The meaning of the symbol “×” is a×b=a _x b _y -a _y b _x , and a and b represent any 2 cross-products. vector, the base station signal is considered to be blocked, and the base station observation value is deleted. There are various methods for judging the occlusion situation, which is not required in the present invention.

Step S32: Select the reference base station of each control point.

When the reference base station is selected, the line of sight between the base station and the control point should be ensured first. The base station closest to the control point is selected as the reference base station to minimize the influence of various errors in complex environments. At the same time, it should be ensured that the base station can be effectively observed by at least two control points including the control point at the same time, so as to unify the benchmark for multi-control point delay calculation. If the base station can only be observed by the current control point, the base station is removed from the candidate subset, and the base station closest to the control point is selected as the reference base station in the candidate subset, and so on until the reference base station of the control point is determined.

Step S33: Construct the difference observation value of each control point.

Specifically, if there is any base station i, reference base station j, and label k, the differential observation equation is:

可以看出，通过基站间观测值做差，标签和同步控制装置的钟差均被消掉，但此时还残留有天线相位偏差等系统误差以及观测噪声。Among them, the difference operator

It can be seen that the clock difference between the tag and the synchronization control device is eliminated by the difference between the observed values between the base stations, but there are still system errors such as antenna phase deviation and observation noise.

Step S4: Perform systematic error correction in combination with the existing error correction model.

Specifically, the antenna phase deviation can be corrected in combination with the existing error correction model, and the corrected observation equation is:

At this time, the unknown in (4) is only r ^ij , and by substituting label coordinates (control point coordinates) and base station coordinates, the relative time delay between i and j base stations can be obtained.

Step S5: Obtain the delay value of each base station by using the least squares adjustment method.

Specifically, referring to FIG. 3 , step S5 in a preferred embodiment of the present invention includes:

Step S51: Adjustment to obtain the relative time delay value of each base station on each control point.

Suppose there are n base stations in total, and (4) is written in matrix form, then the observation equation is:

L=Ax+∈(5)

x＝[r^1j r^2j ... r^(j-1)jr^(j+1)j... r^nj]，A为n-1维单位阵，∈为残差向量，则其法方程为：in

x=[r ^1j r ^2j ... r ^(j-1)j r ^(j+1)j ... r ^nj ], A is the n-1 dimensional unit matrix, ∈ is the residual vector, then its normal equation for:

N=A ^T PA, U=A ^T PL (6)

Optionally, parameters such as the distance between the tag and the base station, the signal-to-noise ratio of the observation value, etc. may be used to determine the weight matrix P, which is not limited in the present invention.

Assuming that the observation duration is m epochs, the delay value at the control point is finally obtained as:

where the subscript t represents the t-th epoch.

Step S52: Return each group of time delays to the same reference base station.

In the final solution, a reference base station needs to be selected, and a base station with good observation conditions is often selected for each control point that can see through. However, since the base stations that can be seen by each control point are not the same, when there is no base station that can be seen by each control point, it is necessary to convert the reference base station at the time of difference.

最后一行并乘以转换矩阵：If the original reference base station is j, and now it is converted to j', r ^jj' can be placed in

Last line and multiply by the transformation matrix:

表示以j′为基准基站的时延向量，

为其对应的方差协方差阵，设计矩阵

in,

represents the delay vector with j' as the reference base station,

For its corresponding variance-covariance matrix, design the matrix

Step S53: Adjustment to solve the final delay value of each base station.

Assuming there are K control points and s observations in total, the error equation is:

v表示平差后的残差向量，则in,

v represents the residual vector after adjustment, then

Next, you can perform a gross error test on the adjustment results to remove observations with large errors. Optionally, use 3δ as the tolerance for gross error detection, which is not limited by Taifang:

其中，v_i为第i个观测值对应的残差值，δ为单位权中误差，则认为第i个观测值中含有粗差，其中，v_i一般选取为残差中的最大值。删掉含有粗差的观测值后重新进行平差，直到没有探测出粗差为止，即可得到最终各基站间的时延值。like

Among them, v _i is the residual value corresponding to the i-th observation value, and δ is the error in the unit weight, then it is considered that the i-th observation value contains gross errors, among which, v _i is generally selected as the maximum value in the residual error. After deleting the observation values with gross errors, the adjustment is performed again until no gross errors are detected, and then the final delay value between the base stations can be obtained.

To sum up, the present invention is based on the UWB synchronous control device, and the multi-control point least squares adjustment jointly calculates the time delay, introduces redundant observations, reduces the influence of various observation errors on the time delay correction, and improves the accuracy of the time delay correction. Accuracy and Stability of Delay Correction

Specifically, compared with the prior art, the present invention has the following advantages:

The existing UWB clock synchronization technology is mostly aimed at electronic communication technology, and does not consider the data processing method in the process of time delay correction. Based on the clock synchronization control device, the present invention proposes a multi-control point time delay correction method. Compared with the traditional single-point delay correction method, this method refers to the GNSS satellite clock error calculation method, and calibrates by arranging multiple discretely distributed control points, and introduces redundant observations; and the least squares adjustment method is used to calculate the final It can weaken the influence of various observation errors caused by different observation environments to a certain extent, and improve the accuracy of time delay correction; in the calculation process, the influence of indoor environmental object occlusion is considered, and gross error detection is adopted. The method eliminates the observations with large errors, which ensures the stability and availability of the calculation results.

The above are only preferred embodiments of the present invention, and do not have any limiting effect on the present invention. Any person skilled in the art, within the scope of not departing from the technical solution of the present invention, makes any form of equivalent replacement or modification to the technical solution and technical content disclosed in the present invention, all belong to the technical solution of the present invention. content still falls within the protection scope of the present invention.

Claims (6)

1. A UWB multipoint time delay correction method based on least square adjustment is characterized by comprising the following steps:

step L1: selecting a plurality of uniformly distributed discrete control points in the coverage area of the UWB system, and acquiring the coordinates of the control points and the coordinates of each base station of the UWB system;

step L2: placing labels on the distributed control points, and collecting static data on each control point;

the basic observation equation is as follows:

where the subscript k denotes the kth tag, the superscript i denotes the ith base station,

represents the observed value of base station i received by tag k,

is the distance between the tag and the base station,

indicating the antenna phase deviation, dt, of the base station i_kIs the clock difference of the label, δ t is the clock difference of the synchronous control device, rⁱThe delay amount from the ith base station to the synchronous control device,

to observe noise, c represents the speed of light;

step L3: preprocessing original observation data by combining prior environment information in a UWB system layout scene, selecting a reference base station, and constructing a differential observation value;

step S31: deleting the observed values corresponding to the base stations which cannot be seen through on each control point;

step S32: selecting a reference base station of each control point;

step S33: constructing a difference observed value of each control point;

specifically, if there is any base station i, reference base station j, and label k, the differential observation equation is:

wherein the difference operator

Step L4: in combination with the existing error correction model, pairStep L2Carrying out system error correction on the static observation data of each control point;

and correcting the antenna phase deviation by combining the existing error correction model, wherein the corrected observation equation is as follows:

step L5: in combination with a GNSS data processing method according toStep L4The observed value of each control point is obtained, the time delay value corresponding to each control point is obtained, and then each UWB base is obtainedRelative delay between stations;

step S51: the adjustment is used to obtain the relative time delay value of each base station at each control point

If a total of n base stations are set, and the (4) is written into a matrix form, the observation equation is as follows:

L＝Ax+∈ (5)

wherein

x＝[r^1jr^2j…r^(j-1)jr^(j+1)j…r^nj]If A is an n-1 dimensional unit matrix and epsilon is a residual vector, the normal equation is as follows:

N＝A^TPA，U＝A^TPL (6)

determining a weight matrix P by adopting the distance between the label and the base station and the signal-to-noise ratio parameter of the observed value;

and if the observation time length is m epochs, finally obtaining a time delay value on the control point as follows:

wherein the subscript t represents the tth epoch;

step S52: each group of time delays is reduced to the same reference base station;

if the original base station is j, it is now converted to j', r can be converted^jj′Is placed at

Last row and multiplication by the conversion matrix:

wherein,

representing the delay vector of the base station with j' as reference,

designing a matrix for its corresponding variance covariance matrix

Step S53: solving the final time delay value of each base station by adjustment;

and if a total of K control points and s observation values are set, the error equation is as follows:

wherein,

v represents the residual vector after adjustment, then

And then, performing gross error test on the adjustment result to remove the observed value with larger error, deleting the observed value containing gross error, and then performing adjustment again until no gross error is detected, thus obtaining the final time delay value among the base stations.

2. The least squares adjustment based UWB multi-point time delay correction method of claim 1, wherein said step L1 comprises:

selecting a plurality of discrete control points with good observation conditions and uniform distribution in a UWB system layout environment;

all base stations can be effectively observed by at least one control point;

and ensuring that every two control points have at least one common base station which can be effectively observed.

3. The least-squares adjustment based UWB multi-point time delay correction method of claim 1, wherein the data preprocessing of the step L3 comprises:

considering the shielding condition of the indoor environment, and eliminating the coordinates of the base station which cannot be seen through on each control point;

and selecting the reference base station corresponding to each control point as the reference of the difference between the stations.

4. The method of claim 3, wherein the UWB multi-point time delay correction method based on least square adjustment is characterized in that: the reference base station selection method comprises the following steps:

ensuring that the reference base station and the control point can be seen through each other;

ensuring that the reference base station can be effectively observed by at least two control points including the control point at the same time;

selecting base stations near the control point as reference base stations as far as possible;

each control point selects the same reference base station as much as possible.

5. The least squares adjustment based UWB multi-point time delay correction method of claim 1, wherein said step L4 comprises:

correction of antenna phase center deviation and correction of remaining system errors for UWB base station antennas.

6. The least squares adjustment based UWB multi-point time delay correction method of claim 1, wherein said step L5 comprises:

jointly solving according to the observed values of the control points, and solving the relative time delay quantity among the UWB base stations by adopting a least square adjustment method;

and eliminating the observed value with larger deviation by adopting a gross error detection theory in the solving process, and iteratively solving the final time delay amount.

Images