CN119231657B

CN119231657B - Intelligent dispatching method and system based on multi-source microgrid fluctuation prediction

Info

Publication number: CN119231657B
Application number: CN202411746138.6A
Authority: CN
Inventors: 赵荣; 周元; 梁辅雄; 张仁; 唐正平
Original assignee: Hunan Xilaike Energy Storage Technology Co ltd
Current assignee: Hunan Xilaike Energy Storage Technology Co ltd
Priority date: 2024-12-02
Filing date: 2024-12-02
Publication date: 2025-03-28
Anticipated expiration: 2044-12-02
Also published as: CN119231657A

Abstract

The application discloses an intelligent scheduling method and system based on fluctuation prediction of a multi-source micro-grid, and relates to the technical field of intelligent grids. The method comprises the steps of obtaining physical data of various renewable energy sources in real time, determining fluctuation modes of the various energy sources based on the physical data, respectively constructing a neural network model for the various energy sources according to the fluctuation modes, and dynamically adjusting energy storage systems and power output of the sub-power grids based on the fluctuation data and real-time loads of the sub-power grids. In addition, the method also realizes the optimal distribution of surplus power by calculating the power supply state of each sub-power grid, and determines the power supply priority of the sub-power grid by adopting a weighted calculation mode based on the fluctuation score and the physical characteristic score. The application improves the stability and the energy utilization efficiency of the multi-source micro-grid, and is suitable for intelligent management and scheduling of a large-scale multi-source micro-grid.

Description

Intelligent scheduling method and system based on multisource micro-grid fluctuation prediction

Technical Field

The application relates to the technical field of smart grids, in particular to an intelligent scheduling method and system based on multi-source micro-grid fluctuation prediction.

Background

The multi-source micro-grid is particularly important in grid-connected control technology because of integration of various renewable energy sources (such as solar energy, wind energy, water energy and the like). However, due to the different power generation characteristics of different types of renewable energy sources, how to realize stable and efficient grid-connected control is still a technical problem to be solved.

For example, chinese patent with the publication number of CN115021318B discloses a synchronization control method and system for grid connection of a multi-support source micro-grid. The method comprises the steps of obtaining voltage signals of a micro-grid island main switch power grid side and a micro-grid side with multiple support sources, calculating according to the voltage signals of the two sides to obtain amplitude deviation, frequency deviation and phase deviation of power grid side and micro-grid side voltage information, performing PI adjustment according to the frequency deviation and the phase deviation to obtain a frequency adjustment instruction, judging whether the frequency deviation and the phase deviation meet a first judgment condition, adjusting the amplitude deviation of the voltage if the frequency deviation and the phase deviation meet a first judgment condition, recalculating the deviations if the frequency deviation and the phase deviation do not meet the first judgment condition, obtaining an amplitude adjustment instruction according to the amplitude deviation adjustment, judging whether the amplitude deviation, the frequency deviation and the phase deviation meet a second judgment condition, issuing a grid-connected closing control instruction if the amplitude deviation, setting the frequency adjustment instruction and the amplitude adjustment instruction to be zero, and recalculating the deviations if the frequency deviation and the phase deviation do not meet the first judgment condition. The method can realize the seamless switching of the multi-support source micro-grid from the island state to the grid-connected state, and is suitable for application in the micro-grid field.

However, the above method has some disadvantages. The method is mainly controlled by depending on the deviation of voltage, frequency and phase, and lacks a real-time prediction and intelligent scheduling mechanism for power generation fluctuation of various renewable energy sources. This results in that when there is a large fluctuation in the multi-source renewable energy generation amount, the response speed and the adjustment accuracy of the system may be insufficient, and it is difficult to achieve optimal power resource allocation. Secondly, in the existing method, when facing a complex multi-source micro-grid environment, the PI regulation strategy may not fully consider the interaction among various energy sources, so that the overall stability and the operation efficiency of the micro-grid are affected. Therefore, a multi-source micro-grid-connected method based on real-time data acquisition and intelligent prediction is needed to solve the defects in the prior art and improve the stability and energy utilization efficiency of the micro-grid.

Disclosure of Invention

Aiming at the defects of the prior art, the application discloses an intelligent scheduling method and system based on multi-source micro-grid fluctuation prediction. The method aims at realizing the efficient and stable grid-connected control of the multi-source micro-grid by acquiring physical data of various renewable energy sources in real time and dynamically adjusting the energy storage system and the power output of each sub-grid by combining an intelligent prediction model.

In a first aspect, the application provides an intelligent scheduling method based on multi-source micro-grid fluctuation prediction, wherein the multi-source micro-grid comprises at least three sub-grids, each sub-grid corresponds to a type of renewable energy source, and the method comprises the following steps:

acquiring physical data of various renewable energy sources in real time; determining a fluctuation mode of each type of renewable energy source based on the physical data of each type of renewable energy source;

Based on the fluctuation modes of the renewable energy sources, respectively constructing a neural network model for predicting the renewable energy source power generation fluctuation data of each renewable energy source;

Performing fluctuation prediction on physical data of various renewable energy sources by utilizing neural network models corresponding to the various renewable energy sources to generate fluctuation data;

And dynamically adjusting the energy storage system and the power output of each sub-grid based on the fluctuation data and the real-time load of each sub-grid.

As an alternative embodiment, the dynamically adjusting the energy storage system and the power output of each sub-grid includes:

determining the power supply state of each sub-grid based on the fluctuation data and the real-time load of each sub-grid;

the power supply state comprises a surplus state, a sufficient state and an insufficient state;

and in response to the fact that the power supply state of the sub-power grid is the insufficient state, utilizing the sub-power grid with the power supply state being the surplus state to supply power to the sub-power grid in the insufficient state.

As an alternative embodiment, the determining the power supply status of each sub-grid includes:

Determining the generated energy and the generated electricity demand of each sub-grid based on the fluctuation data and the real-time load of each sub-grid;

in response to the presence of the sub-grid having the power generation amount exceeding the power generation demand, and the exceeding value being greater than or equal to a first threshold, marking the sub-grid as being in a surplus state;

In response to the generation amount of the sub-grid exceeding the generation demand amount and the exceeding value being smaller than a first threshold, marking the sub-grid as being in a sufficient state;

In response to the power generation demand of the sub-grid exceeding the power generation amount, the sub-grid is marked as being in an insufficient state.

As an alternative embodiment, the supplying power to the sub-grid in the insufficient state includes:

Determining the power-available priority of each surplus state sub-grid based on fluctuation data corresponding to each surplus state sub-grid;

and supplying power by using the sub-power grid with the highest priority among the power-available priorities.

As an alternative embodiment, said determining the energizable priority of the sub-grid of each of said surplus states comprises:

determining a volatility score of each sub-grid based on volatility data of each surplus state sub-grid;

Determining physical characteristic scores of all the sub-grids based on physical data of renewable energy sources corresponding to all the surplus state sub-grids;

and weighting and calculating the powerable priority of each power-rich state sub-grid based on the volatility score and the physical characteristic score of each sub-grid.

As an alternative embodiment, the volatility score is calculated based on a dynamic time period, and the determining the volatility score of each sub-grid comprises:

And adjusting the calculation time period of the fluctuation score based on the real-time load of each sub-power grid, wherein the calculation time period of the fluctuation score is shorter as the real-time load of the sub-power grid is closer to a preset load threshold of the sub-power grid.

As an alternative embodiment, determining the physical characteristic score of each of the sub-grids comprises:

The renewable energy sources comprise solar energy, wind energy and water energy;

Responding to the renewable energy source corresponding to the sub-power grid as solar energy, wherein the physical characteristic score is based on the current illumination intensity and the sunlight duration, and the physical characteristic score is higher as the illumination intensity is higher;

The method comprises the steps of responding to renewable energy sources corresponding to a sub-grid to be wind energy, wherein the physical characteristic score is based on the current wind speed, and the physical characteristic score is highest when the wind speed is in a preset power generation interval;

And responding to the renewable energy source corresponding to the sub-power grid as water energy, wherein the physical characteristic score is based on the current water flow and water storage quantity, and the physical characteristic score is higher as the water flow and water storage quantity are closer to the preset target value.

In a second aspect, the application also provides an intelligent scheduling system based on the fluctuation prediction of the multi-source micro-grid, which comprises an acquisition unit, a construction unit, a prediction unit and a scheduling unit, wherein:

The acquisition unit is used for acquiring physical data of various renewable energy sources in real time, and determining fluctuation modes of the various renewable energy sources based on the physical data of the various renewable energy sources;

The construction unit is used for respectively constructing a neural network model for predicting the power generation fluctuation data of each type of renewable energy sources for each type of renewable energy sources based on the fluctuation modes of each type of renewable energy sources;

the prediction unit is used for predicting fluctuation according to physical data of various renewable energy sources by utilizing neural network models corresponding to the various renewable energy sources to generate fluctuation data;

The scheduling unit is used for dynamically adjusting the energy storage system and the power output of each sub-power grid based on the fluctuation data and the real-time load of each sub-power grid.

Compared with the prior art, the method has the beneficial effects that by acquiring physical data of various renewable energy sources in real time and combining an intelligent prediction model, the power generation fluctuation can be accurately predicted, and the energy storage system and the power output can be dynamically adjusted, so that the response speed and the energy utilization efficiency of the micro-grid are remarkably improved. In addition, a weighting calculation method based on fluctuation score and physical characteristic score is introduced, so that more intelligent power supply distribution is realized, the sub-power grid with high power generation stability and superior physical conditions is ensured to provide power support preferentially, the complex fluctuation characteristics of various renewable energy sources in the multi-source micro-power grid are effectively treated, and the overall stability and efficient operation of the system are maintained.

By deeply analyzing the fluctuation modes of various energy sources and adopting a special neural network model for prediction and scheduling, the problem that the response speed and the adjustment precision are insufficient when the traditional method is used for coping with the fluctuation of a large-scale multi-source micro-grid is solved, the reliability and the operation efficiency of the micro-grid are obviously improved, and the optimal distribution and utilization of power resources are ensured.

Drawings

Fig. 1 is a flowchart of an intelligent scheduling method based on multi-source micro-grid fluctuation prediction provided by an embodiment of the application;

FIG. 2 is a flowchart of a method for determining a power supply status of each sub-grid according to an embodiment of the present application;

Fig. 3 is a schematic diagram of an intelligent scheduling system based on multi-source micro-grid fluctuation prediction according to an embodiment of the present application.

Detailed Description

The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments.

In this disclosure Grid-Connected refers to the process of connecting a power generation system (e.g., solar, wind, or other distributed energy source) to a utility Grid. Grid-tied systems are capable of delivering power generated by power generation directly into the grid for sharing with other power sources or to obtain power from the grid when power generation is insufficient.

Referring to fig. 1, a flowchart of an intelligent scheduling method based on multi-source micro-grid fluctuation prediction according to an embodiment of the present application is shown, where the multi-source micro-grid includes at least three sub-grids, each sub-grid corresponds to a type of renewable energy source, and the method includes steps S101 to S104, where:

S101, acquiring physical data of various renewable energy sources in real time, and determining fluctuation modes of the various renewable energy sources based on the physical data of the various renewable energy sources;

s102, respectively constructing a neural network model for predicting power generation fluctuation data of various renewable energy sources for the various renewable energy sources based on the fluctuation modes of the various renewable energy sources;

s103, carrying out fluctuation prediction on physical data of various renewable energy sources by utilizing a neural network model corresponding to the various renewable energy sources to generate fluctuation data;

And S104, dynamically adjusting the energy storage system and the power output of each sub-grid based on the fluctuation data and the real-time load demand of each sub-grid.

The multi-source micro-grid comprises at least three sub-grids, wherein each sub-grid is respectively associated with different types of renewable energy sources. The type of renewable energy source may be solar, wind, water or other types of clean energy sources. The sub-grids can realize cooperative power supply through intelligent scheduling, so that the stability of the whole micro-grid system is ensured.

For S101 described above:

In a specific implementation, corresponding physical parameter sensors and data acquisition devices can be deployed in each sub-grid to enable comprehensive monitoring of different types of renewable energy sources.

For a solar energy sub-grid, an illumination intensity sensor, an ambient temperature sensor, a solar radiometer and other key parts are arranged on a photovoltaic panel array, a solar heat collector and the like. The sensors can acquire physical data such as illumination intensity, ambient temperature, solar radiation intensity, sunlight duration and the like in real time. The collected data is transmitted to a local controller or a central control system through wired (such as Ethernet, optical fiber) or wireless (such as Wi-Fi, zigBee, loRa) communication modes.

In the wind energy sub-grid, sensors such as an ultrasonic anemometer, a mechanical cup anemometer, a wind vane and the like are arranged at the positions of a wind generating set, a wind tower and the like. The sensors monitor key parameters such as wind speed, wind direction, air pressure, temperature and humidity in real time.

For the water energy power grid, a water level gauge, a flowmeter, a pressure sensor and a temperature sensor are arranged at key positions such as a water inlet, a water outlet and a dam body of the hydroelectric power station. The sensors can acquire physical data such as water level height, water flow, pressure, water temperature and the like in real time, and transmit the physical data to a control room or a remote monitoring center in a high-speed communication mode such as industrial Ethernet, optical fiber and the like.

In the data acquisition process, reasonable sampling frequency needs to be set to meet the real-time requirements of different energy types. For example, the data sampling frequency of the solar energy sub-grid can be set to be once per minute, the wind energy sub-grid can be set to be multiple times per second, and the water energy sub-grid can dynamically adjust the sampling frequency according to the water flow change condition. In addition, the sensor node can be further provided with a data caching module to prevent data loss when communication is interrupted, and real-time transmission of data is guaranteed by adopting a real-time communication protocol.

After the data is transmitted to the central control system, the received physical data is preprocessed, for example, including data cleansing, calibration and time synchronization. The data cleaning removes abnormal values and noise through filtering algorithms (such as mean filtering and Kalman filtering), and ensures the smoothness and accuracy of the data. And correcting the data according to the calibration curve of the sensor and the environmental factors by the data correction rule, and eliminating the system error. And the time synchronization aligns the data of different sources in time through a time stamp or a synchronous clock, so that the synchronism and consistency of the data are ensured.

After finishing data preprocessing, firstly extracting features based on physical data acquired in real time, for solar energy, extracting features such as sunlight peak value, slope, fluctuation frequency and the like by analyzing the change trend of illumination intensity, for wind energy, calculating the mean value, variance and frequency spectrum characteristics of wind speed, identifying the periodicity and randomness of wind speed change, for water energy, evaluating the change rate of water flow and water level fluctuation, and identifying seasonal and sudden change features. Then, the system adopts a machine learning algorithm (such as cluster analysis and support vector machine) or a signal processing method (such as wavelet analysis and Fourier transformation) to analyze the extracted characteristics, and identifies the fluctuation modes of various renewable energy sources.

For S102 described above:

In a specific implementation, for a solar sub-grid, a Long Short-Term Memory (LSTM) model may be constructed using the identified wave pattern. Due to the advantages of the LSTM model in the aspect of processing time sequence data, the solar periodic variation and short-term fluctuation characteristics of the solar energy power generation amount can be effectively captured, and therefore accurate prediction of the future power generation amount is achieved. The wind energy sub-grid has higher fluctuation because the generated energy is greatly influenced by wind speed and wind direction, so a mixed model combining a convolutional neural network (Convolutional Neural Network, CNN) and LSTM can be adopted. The CNN is used for extracting spatial features in wind speed data, and the LSTM is used for capturing dynamic changes in time, so that the accuracy of wind power generation fluctuation prediction is improved. For the water energy grid, a recurrent neural network (Recurrent Neural Network, RNN) may be selected as a predictive model, taking into account that the changes in water flow and water level are relatively stable but there are seasonal and sudden fluctuations. The RNN can effectively treat medium-and-long-term fluctuation trend of the water energy generation capacity and adapt to seasonal changes and influence of occasional events.

By taking an LSTM model of a solar energy sub-grid as an example in the training process of the model, the preprocessed time series data can be input into the model, and model parameters are optimized through multiple rounds of iterative training, so that the periodic variation and short-term fluctuation of the solar energy power generation amount can be accurately captured. The CNN-LSTM hybrid model of the wind energy sub-grid is trained in stages, spatial features of wind speed data are extracted by the CNN layer, and the extracted features are input into the LSTM layer for time sequence prediction. The RNN model of the water energy power grid is used for processing the time dependence of water flow and water level through a recursion structure and adapting to the change of the fluctuation mode. In the training process, mean Square Error (MSE) can be adopted as a loss function, and Adam optimization algorithm is used for accelerating convergence, so that training efficiency is improved.

In addition, after training is completed, the neural network model can be deployed to a central control system or an edge computing node to realize real-time power generation fluctuation prediction. And generating power generation fluctuation prediction data in a future period of time by continuously inputting physical data acquired in real time and utilizing a trained model. These predictive data will be used to guide the charging and discharging strategy of the energy storage system and the dynamic adjustment of the power output, ensuring that the microgrid can remain stable and efficient in handling renewable energy fluctuations.

For S103 described above:

in specific implementation, physical data of various renewable energy sources are analyzed by using the constructed neural network model, and predicted power generation fluctuation data are generated. The power generation fluctuation data can be used for evaluating the power generation capacity of each sub-grid and providing basis for subsequent energy storage and output adjustment.

For S104 described above:

And dynamically adjusting the charge and discharge states and the power output of the energy storage system based on the fluctuation data and the real-time load condition of each sub-power grid. The flexible adjustment of the energy storage system can cope with the fluctuation of various renewable energy sources, and the energy utilization efficiency is improved. By matching the power output with the real-time demand, energy waste or shortfall can be avoided, thereby maintaining the overall stability and efficient operation of the micro-grid.

As an optional implementation manner, in the intelligent scheduling method based on the fluctuation prediction of the multi-source micro-grid, the process of dynamically adjusting the energy storage system and the power output of each sub-grid further comprises judging and scheduling the power supply state of each sub-grid.

In a specific implementation, the power supply state of each sub-grid is determined based on the fluctuation data and the real-time load of the respective sub-grid. The power supply state can be divided into three states, a rich state, a sufficient state, and a deficient state, wherein:

and in a surplus state, the generated energy of one sub-power grid exceeds the current load demand, and the surplus generated energy can be supplied to other sub-power grids.

And in a sufficient state, the generated energy of a certain sub-power grid is basically matched with the current load demand, and the power supply demand can be balanced autonomously without redundancy and shortage.

And in an insufficient state, the generated energy of one sub-power grid is insufficient to meet the current load demand, and additional power support needs to be obtained from other sub-power grids.

And when detecting that a certain sub-power grid is in an insufficient state, automatically searching the sub-power grid in a surplus state. The redundant sub-grids can supply power to the sub-grids in the insufficient state through the intelligent dispatching system so as to keep the stable operation of the whole micro-grid.

Specifically, when a certain sub-grid has insufficient power supply, a power supply request is sent to other sub-grids through a communication network. In response, the redundant sub-grids can provide power support according to the power generation capacity and the energy storage condition of the sub-grids, so that the sub-grids in the insufficient state can recover the power supply capacity, and the risk of load outage or overload is avoided.

Referring to fig. 2, a flowchart of a method for determining a power supply state of each sub-grid according to an embodiment of the present application is shown, where the method includes steps S201 to S204, where:

S201, determining the generated energy and the generated electricity demand of each sub-grid based on the fluctuation data and the real-time load of each sub-grid;

s202, when the generated energy of one of the sub-grids exceeds the generated energy demand and the exceeding value is larger than or equal to a first threshold value, marking that the sub-grid is in a surplus state;

S203, in response to the generated energy of a certain sub-grid exceeding the generated energy demand, and when the exceeding value is smaller than a first threshold, marking that the sub-grid is in a sufficient state;

And S204, in response to the power generation demand of one of the sub-grids exceeding the power generation amount, marking that the sub-grid is in an insufficient state.

In a specific implementation, the power generation amount and the power generation demand amount of each sub-grid are firstly collected in real time. For example, the power generation amount can be measured by a power metering device and represents the actual power generation output of renewable energy sources (such as solar energy, wind energy and water energy) in the sub-power grid, and the power generation demand is obtained by a load detection device and represents the power required in the current sub-power grid to maintain the normal operation of all loads.

For a certain solar sub-grid, the current generated power can be collected in real time through a power sensor connected to a photovoltaic inverter, and meanwhile, the power consumption requirements of all electrical equipment in the sub-grid are detected through a load controller.

And determining a power supply state according to the difference value of the generated energy and the generated demand of each sub-power grid. The difference reflects the power supply and demand balance condition of the sub-power grid, and the difference is taken as a judgment basis.

When the power generation amount is larger than the power generation demand amount, the difference is compared with a preset 'first threshold'. The threshold is dynamically set by historical data and system requirements. For example, the threshold may be determined based on a power safety margin of the system or a charging capability of the energy storage device.

In particular implementations, when the amount of power generated by a particular sub-grid exceeds the power generation demand by a value greater than or equal to a first threshold, the sub-grid is marked as being in a "surplus state". The surplus state means that the sub-power grid can not only meet the load demand of the sub-power grid, but also provide power support for other sub-power grids with insufficient loads.

For example, at higher wind speeds, one wind power sub-grid may generate power that greatly exceeds the load demand, and the excess may be stored or transferred to other sub-grids. At this time, the sub-grid can be marked as a surplus state, so that the surplus power can be quickly identified and utilized in the subsequent power supply scheduling.

When the generated energy of a certain sub-grid exceeds the generated demand, but the exceeding value is smaller than a first threshold value, the sub-grid is marked to be in a sufficient state. The sufficiency state indicates that the power generation of the sub-grid is substantially balanced with its load demand, but that its power generation is insufficient to provide additional power to the other sub-grids.

For example, when the water flow rate of one hydraulic sub-grid is moderate, the generated energy of the hydraulic sub-grid can maintain the normal operation of the sub-grid, but no redundant electric power is stored or output to other sub-grids.

When the power generation demand of a certain sub-power grid exceeds the power generation capacity, the sub-power grid is marked to be in an 'insufficient state'. The insufficient state means that the generated energy of the sub-grid cannot meet the current load demand, which may cause power failure of the load or unstable system, and power support needs to be obtained from the sub-grid in other surplus states.

For example, when solar intensity is reduced, the power production of a solar sub-grid may be insufficient to support all load operations, at which time the sub-grid may be marked as an insufficient condition and other surplus condition sub-grids automatically sought to provide power replenishment.

The first threshold is dynamically adjusted based on historical power generation and load data of the multi-source micro-grid so as to ensure that surplus power resources are utilized to the maximum extent on the premise of not wasting energy. For example, the threshold may be set based on factors such as the capacity of the energy storage system, the load characteristics of the sub-grid, and the power generation volatility.

For example, if the capacity of the energy storage system of a certain sub-grid is smaller, the first threshold may be set lower, so as to ensure that the sub-grid does not waste energy due to excess power exceeding the energy storage capacity.

Therefore, the power supply state of each sub-power grid can be rapidly and accurately judged through detailed analysis of the generated energy and the required amount, and reasonable distribution of energy resources is ensured. The method can effectively cope with the generation uncertainty caused by renewable energy fluctuation, avoid the occurrence of power shortage or waste and further improve the reliability and the operation efficiency of the micro-grid.

As an optional implementation mode, the power supply of the sub-power grid in the insufficient state comprises the steps of determining the power supply priority of each sub-power grid in the excessive state based on fluctuation data corresponding to the sub-power grid in the excessive state, and utilizing the sub-power grid with the highest priority in the power supply priority to supply power.

In a specific implementation, when a certain sub-grid is in an insufficient state, all sub-grids in a surplus state are first identified, and fluctuation data of the sub-grids are acquired. The fluctuation data represents the variation trend of the power generation amount of each sub-grid in a specific period. And (3) judging the power generation stability and the sustainability of each surplus sub-power grid in a period of time in the future by analyzing the data.

For example, a wind energy sub-grid is currently in a surplus state, but according to fluctuation data (such as wind speed fluctuation) of the wind energy sub-grid, large fluctuation of the power generation capacity of the sub-grid in the future can occur, and stable power is difficult to continuously provide. Therefore, the system may prefer a less fluctuating sub-grid with more stable power generation during power scheduling.

In a specific implementation, the priority of power supply of each surplus state sub-grid is calculated according to fluctuation data of the sub-grid. The determination of priority may be based on several factors:

generating stability-the sub-grids with stable fluctuation data usually have higher power supply priority, because the sub-grids can keep continuous power generation for a longer time in the power supply process, and the risk of load fluctuation is reduced.

The sub-power grid with larger current power generation capacity and energy storage capacity is also given higher priority, so that enough power is provided for the sub-power grid in the insufficient state.

In the energy storage state, if the energy storage device of one surplus sub-grid is in a full state, the priority of the energy storage device can be increased, because the continuous storage of the power at the moment can cause waste, and the power is not distributed to other sub-grids.

For example, two sub-grids in a surplus state, a wind energy sub-grid and a water energy sub-grid, are detected. The wind energy sub-grid has larger generating capacity but higher volatility, and the water energy sub-grid has smaller generating capacity but stable generating. Based on these data, the water energy grid may be given a higher power supply priority.

In a specific implementation, the sub-grid with the highest priority is selected to supply power to the sub-grid in the insufficient state. Through the priority ordering, more reasonable power dispatching in the micro-grid can be ensured, the requirements of insufficient sub-grids can be met, and the waste of power is avoided.

In this way, accurate and efficient power distribution can be achieved in the power supply scheduling process. And a mechanism for determining the priority based on the fluctuation data ensures that the sub-power grid in the insufficient state obtains the most stable power support preferentially, and the problem of insufficient secondary power supply caused by overlarge fluctuation is avoided.

In a specific implementation, the volatility score is used to evaluate the power generation stability of each surplus state sub-grid. The system calculates the score based on fluctuation data of each sub-grid, wherein the fluctuation data reflects the change condition of the power generation amount of various renewable energy sources (such as wind energy, solar energy and water energy) in a specific time period.

The calculation of the volatility score can be carried out by analyzing the historical power generation data of a certain sub-power grid, if the power generation capacity of the sub-power grid has larger fluctuation in a short time, the volatility score is lower, and if the power generation capacity is smoother, the volatility score is higher.

For example, when the wind speed of a wind energy sub-grid is relatively unstable, the generated energy of the wind energy sub-grid can be fluctuated, so that the fluctuation score can be lower, and the fluctuation score of the water energy sub-grid is higher due to the relatively stable water flow.

In a specific implementation, the physical characteristic score is used to evaluate the current physical condition of the renewable energy source of each surplus state sub-grid. The physical characteristic score is determined by analyzing physical data of the renewable energy source corresponding to the sub-grid, the data including illumination intensity of solar energy, wind speed of wind energy, water flow of water energy, etc., wherein:

The higher the illumination intensity of the solar energy sub-grid, the higher the physical characteristic score. The illumination intensity can be evaluated in real time based on factors such as sunlight duration, cloud cover and the like.

The wind energy sub-grid has the advantages that when the wind speed is in an ideal power generation interval, the physical characteristic score is highest, and if the wind speed deviates from the interval, the score is reduced.

The water energy power grid is that the closer the water flow and the water storage quantity are to the preset ideal state, the higher the physical characteristic score is.

For example, the physical characteristics of the wind energy sub-grid score is higher when the wind speed is in the optimal power generation interval (e.g. 7-15 m/s), whereas the score will decrease if the wind speed is too high or too low.

It is understood that the above-mentioned optimal power generation section is a preset power generation section.

In specific implementation, the powerable priority of each surplus state sub-grid is obtained by weighting and calculating the volatility score and the physical characteristic score. The weight calculation formula can be adjusted according to actual requirements. For example, if the volatility of a certain sub-grid is low, the power generation stability is good, the volatility score weight of the sub-grid can be increased, and in a specific case, if the physical characteristic score shows that the energy resources of the sub-grid are rich, the weight of the physical characteristic score can be increased.

According to the result of the weighting calculation, the sub-grid with higher priority will preferentially provide power for the sub-grid in the shortage state. The multi-dimensional priority evaluation method enables the system to select the most suitable sub-grid for power supply under the condition that the power generation capacity and the stability of different sub-grids are different.

In this way, when the power supply priority of the sub-grids in the surplus state is ordered, the power generation fluctuation and the physical characteristics are comprehensively considered. The method effectively improves the utilization efficiency of the power resources, and reduces the influence of renewable energy fluctuation on power supply to the greatest extent while maintaining the stability of the micro-grid.

In a specific implementation, first, real-time load data of each sub-grid is acquired. Based on these data, the calculation time period of the volatility score is dynamically adjusted.

In a specific implementation, when the real-time load of the sub-grid is closer to the preset load threshold, the load pressure of the sub-grid is larger, and the power generation fluctuation needs to be monitored more frequently. Therefore, the calculation time period of the fluctuation score is shortened to capture finer fluctuation changes, and the power supply strategy is convenient to adjust in time.

When the real-time load of the sub-power grid is lower and the distance from the load threshold is longer, the operation of the sub-power grid is loose, the calculation time period can be properly prolonged, the long-term fluctuation trend is focused, and unnecessary frequent adjustment is reduced.

In particular embodiments, each sub-grid has a predetermined load threshold, which may be set according to the design capacity or safe operating criteria of the sub-grid.

And calculating the ratio of the real-time load to the load threshold value to obtain the load proximity. For example, if the real-time load is 90% of the threshold value, the load proximity is 90%. And setting a corresponding calculation time period according to the load proximity. The higher the load proximity, the shorter the calculation period. A mapping relationship may be preset or adjusted using a linear function.

And in the adjusted calculation time period, collecting the generated energy data of the sub-power grid, and calculating the volatility score by using a statistical method. For example, a standard deviation, variance, or other volatility index of the power generation amount over the period of time may be calculated.

By way of example, the load threshold for a wind energy sub-grid is 1000kW, the current real-time load is 950kW, and the load proximity is 95%. According to a preset rule, when the load proximity exceeds 90%, the calculation period is set to 10 minutes. The system collects power generation data in the last 10 minutes, calculates a volatility score to reflect the current power generation stability under high load.

By way of example, the load threshold for a water energy sub-grid is 800kW, the current real-time load is 400kW, and the load proximity is 50%. At this time, the calculation period may be set to 60 minutes, focusing on the power generation fluctuation situation over a long period of time.

The fluctuation score obtained after the calculation time period is dynamically adjusted can reflect the power generation stability of the sub-power grid under different load conditions. Under the condition of high load, the grading is more sensitive, the potential power supply risk can be found in time, and under the condition of low load, the grading is smoother, and unnecessary adjustment caused by short-term fluctuation is avoided.

For example, the relationship of the calculation period T and the load proximity D may be set as:

Wherein, For the maximum calculated time period,AndMinimum and maximum load proximity, respectively.

Collecting generating capacity data in a determined calculation time periodA volatility score S is calculated:

Where N is the total number of sampling points, representing the number of power generation data points collected during the calculation period, i is the index of the sampling points, and the value ranges are i=1, 2, 3. To calculate the average power generation amount over the period of time,Is thatGenerating capacity at moment.

Therefore, the calculation time period of the fluctuation score can be dynamically adjusted according to the real-time load of the sub-power grid, so that the scoring result can more accurately reflect the power generation fluctuation of the sub-power grid under different load levels. Under the high load condition, the calculation time period is shortened, the power generation fluctuation can be captured in time, the potential power supply problem can be responded quickly, under the low load condition, the calculation time period is prolonged, frequent adjustment is reduced, and the system operation efficiency is improved.

It will be appreciated by those skilled in the art that in the above-described method of the specific embodiments, the written order of steps is not meant to imply a strict order of execution but rather should be construed according to the function and possibly inherent logic of the steps.

Based on the same inventive concept, the embodiment of the application also provides an intelligent scheduling system based on the fluctuation prediction of the multi-source micro-grid, and because the principle of solving the problem by the system in the embodiment of the application is similar to that of the intelligent scheduling method based on the fluctuation prediction of the multi-source micro-grid, the implementation of the system can be referred to the implementation of the method, and the repetition is omitted.

Referring to fig. 3, a schematic diagram of an intelligent scheduling system based on multi-source micro-grid fluctuation prediction according to an embodiment of the present application is shown, where the multi-source micro-grid includes at least three sub-grids, each sub-grid corresponds to a type of renewable energy source, and the system includes an acquisition unit 10, a construction unit 20, a prediction unit 30, and a scheduling unit 40, where:

The acquisition unit 10 is used for acquiring physical data of various renewable energy sources in real time, and determining fluctuation modes of the renewable energy sources based on the physical data of the renewable energy sources;

The construction unit 20 is configured to construct a neural network model for predicting power generation fluctuation data of each type of renewable energy source for each type of renewable energy source, based on the fluctuation modes of each type of renewable energy source;

The prediction unit 30 is configured to perform fluctuation prediction on physical data of each type of renewable energy source by using a neural network model corresponding to each type of renewable energy source, so as to generate fluctuation data;

the scheduling unit 40 is configured to dynamically adjust the energy storage system and the power output of each sub-grid based on the fluctuation data and the real-time load of each sub-grid.

The process flow of each unit in the system and the interaction flow between each unit may be described with reference to the related description in the above method embodiment, which is not described in detail herein.

It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described system and apparatus may refer to corresponding procedures in the foregoing method embodiments, which are not described herein again. In the several embodiments provided in the present disclosure, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners. The above-described apparatus embodiments are merely illustrative, for example, the division of the units is merely a logical function division, and there may be other manners of division in actual implementation, and for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some communication interface, device or unit indirect coupling or communication connection, which may be in electrical, mechanical or other form.

The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a non-volatile computer readable storage medium executable by a processor. Based on such understanding, the technical solution of the present disclosure may be embodied in essence or a part contributing to the prior art or a part of the technical solution, or in the form of a software product stored in a storage medium, including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in the embodiments of the present disclosure. The storage medium includes a U disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, an optical disk, or other various media capable of storing program codes.

It should be noted that the foregoing embodiments are merely specific implementations of the disclosure, and are not intended to limit the scope of the disclosure, and although the disclosure has been described in detail with reference to the foregoing embodiments, it should be understood by those skilled in the art that any modification, variation or substitution of some of the technical features described in the foregoing embodiments may be made or equivalents may be substituted for those within the scope of the disclosure without departing from the spirit and scope of the technical solutions of the embodiments. Therefore, the protection scope of the present disclosure shall be subject to the protection scope of the claims.

Claims

1. An intelligent dispatching method based on multi-source microgrid fluctuation prediction, characterized in that the multi-source microgrid includes at least three subgrids; each of the subgrids corresponds to a type of renewable energy, and the method includes:

Acquire physical data of each type of renewable energy in real time; determine fluctuation patterns of each type of renewable energy based on the physical data of each type of renewable energy;

Based on the fluctuation patterns of the various types of renewable energy, a neural network model for predicting the fluctuation data of power generation of the various types of renewable energy is constructed for each type of renewable energy;

For the physical data of each type of renewable energy, a neural network model corresponding to each type of renewable energy is used to perform fluctuation prediction to generate fluctuation data;

Based on the fluctuation data and the real-time load of each subgrid, the energy storage system and power output of each subgrid are dynamically adjusted; the dynamic adjustment of the energy storage system and power output of each subgrid also includes: determining the volatility score of each subgrid based on the fluctuation data of each subgrid; determining the physical characteristic score of each subgrid based on the change information of the physical data of the renewable energy corresponding to each subgrid; wherein the change information includes: light change information, wind speed change information, and water flow change information; based on the volatility score and physical characteristic score of each subgrid, weighted calculation of the power supply priority of each subgrid; wherein the volatility score is calculated based on a dynamic time period; the determination of the volatility score of each subgrid includes: adjusting the calculation time period of the volatility score based on the real-time load of each subgrid; wherein, the closer the real-time load of the subgrid is to the preset load threshold of the subgrid, the shorter the calculation time period of the volatility score;

The dynamic adjustment of the energy storage system and power output of each sub-grid includes:

Determine the power supply status of each subgrid based on the fluctuation data and the real-time load of each subgrid;

The power supply status includes: surplus status, sufficient status, and insufficient status;

In response to the subgrid being in the insufficient power supply state, using the subgrid in the surplus power supply state to supply power to the insufficient power supply subgrid;

Determining the power supply status of each sub-grid includes:

Determine the power generation and power generation demand of each subgrid based on the fluctuation data and the real-time load of each subgrid;

In response to the existence of the power generation of the subgrid exceeding the power generation demand, and the excess value is greater than or equal to a first threshold, marking the subgrid as being in a surplus state;

In response to the existence of the power generation of the subgrid exceeding the power generation demand, and the excess value is less than a first threshold, marking the subgrid as being in a sufficient state;

In response to the generation demand of the subgrid exceeding the generation capacity, the subgrid is marked as being in a deficit state.

2. The method according to claim 1, characterized in that supplying power to the sub-grid in insufficient state comprises:

Determining the power supply priority of each sub-grid in a surplus state based on the fluctuation data corresponding to each sub-grid in a surplus state;

The subgrid with the highest priority among the available power supply priorities is used for power supply.

3. The method according to claim 2, characterized in that the step of determining the power supply priority of each of the sub-grids in the surplus state comprises:

Determining a volatility score of each of the subgrids based on the volatility data of each of the subgrids in a surplus state;

Determining a physical characteristic score of each of the subgrids based on the physical data of the renewable energy corresponding to each of the surplus state subgrids;

Based on the volatility score and the physical characteristic score of each of the subgrids, the power supply priority of each of the subgrids in the surplus state is weightedly calculated.

4. The method according to claim 3, characterized in that determining the physical characteristic score of each of the sub-grids comprises:

The renewable energy includes: solar energy, wind energy, and water energy;

In response to the renewable energy corresponding to the subgrid being solar energy, the physical property score is based on current light intensity and sunshine duration, and the higher the light intensity, the higher the physical property score;

In response to the renewable energy corresponding to the subgrid being wind energy, the physical characteristic score is based on the current wind speed; wherein, when the wind speed is within a preset power generation interval, the physical characteristic score is the highest; in response to the wind speed being lower than or higher than the preset power generation interval, the physical characteristic score gradually decreases based on the magnitude of the wind speed deviation from the preset power generation interval;

In response to the renewable energy corresponding to the subgrid being hydropower, the physical characteristic score is based on current water flow and water storage; wherein, the closer the water flow and water storage are to preset target values, the higher the physical characteristic score is.

5. An intelligent dispatching system based on multi-source microgrid fluctuation prediction, characterized in that the intelligent dispatching method based on multi-source microgrid fluctuation prediction as described in any one of claims 1 to 4 is executed; the multi-source microgrid includes at least three subgrids; each of the subgrids corresponds to a type of renewable energy, and the system includes: a collection unit, a construction unit, a prediction unit, and a dispatching unit; wherein,

The acquisition unit is used to acquire the physical data of each type of renewable energy in real time; based on the physical data of each type of renewable energy, determine the fluctuation mode of each type of renewable energy;

The construction unit is used to construct a neural network model for predicting the power generation fluctuation data of each type of renewable energy based on the fluctuation mode of each type of renewable energy.

The prediction unit is used to perform fluctuation prediction on the physical data of each type of renewable energy by using the neural network model corresponding to each type of renewable energy to generate fluctuation data;

The dispatching unit is used to dynamically adjust the energy storage system and power output of each subgrid based on the fluctuation data and the real-time load of each subgrid; the dynamic adjustment of the energy storage system and power output of each subgrid also includes: determining the volatility score of each subgrid based on the fluctuation data of each subgrid; determining the physical characteristic score of each subgrid based on the change information of the physical data of the renewable energy corresponding to each subgrid; wherein the change information includes: light change information, wind speed change information, and water flow change information; based on the volatility score and physical characteristic score of each subgrid, weighted calculation of the power supply priority of each subgrid; wherein the volatility score is calculated based on a dynamic time period; the determination of the volatility score of each subgrid includes: adjusting the calculation time period of the volatility score based on the real-time load of each subgrid; wherein, the closer the real-time load of the subgrid is to the preset load threshold of the subgrid, the shorter the calculation time period of the volatility score.