A power storage control method and system

Abstract

The application discloses a kind of electric power energy storage control method and system, it is related to energy storage equipment management and control technical field, method includes collecting energy storage system state data and energy storage system work log data;Based on ampere-hour integration method, the energy storage system state data is analyzed in real time, and energy storage battery real-time state data is generated;Every interval setting first length is based on EKF algorithm, the energy storage system state data is analyzed, and energy storage battery state calibration data is generated;The electric power energy storage control method and system, by analyzing the error amount that needs to be calibrated and corresponding battery parameters, state etc. Feature data corresponding prior probability of calibration are constructed error prediction model, then based on the energy storage battery feature input error prediction model of next calibration time that meets historical mode is analyzed, so as to reach the error amount of next calibration time is predicted, realize the effect that real-time estimated battery state parameter is pre-calibrated by predicted error amount.

Description

Technical Field

[0001] This invention relates to the field of energy storage equipment management and control technology, specifically to a power energy storage control method and system. Background Technology

[0002] Energy storage systems are devices or systems that store energy through physical, chemical, or other means. They can release energy during peak electricity demand periods to balance power supply and demand, and improve the flexibility and security of the power system. Energy storage systems play a crucial role in renewable energy consumption, peak shaving, and frequency regulation. The control of power energy storage systems is their core function, directly determining the system's performance, efficiency, and lifespan.

[0003] Existing energy storage systems are prone to errors in estimating the state of energy storage batteries. For example, the ampere-hour integration method is prone to cumulative errors, which makes the estimation error of the energy storage battery state larger and larger over time. Using advanced algorithms to estimate the state of energy storage batteries in real time requires a huge amount of computing resources, which is not conducive to cost control. Summary of the Invention

[0004] The purpose of this invention is to provide a power storage control method and system to overcome the above-mentioned shortcomings in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a power storage control method, comprising the following steps:

[0006] S1. Collect energy storage system status data and energy storage system operation log data;

[0007] S2. Based on the ampere-hour integral method, perform real-time analysis and processing of the energy storage system status data to generate real-time status data of the energy storage battery.

[0008] S3. Every set first time interval, the energy storage system status data is analyzed and processed based on the EKF (Extended Kalman Filter) algorithm to generate energy storage battery status calibration data.

[0009] S4. Based on the deviation between the real-time state data of the energy storage battery and the state calibration data of the energy storage battery at the same moment, the real-time model error data is calculated.

[0010] S5. Based on the energy storage battery status calibration data at each calibration, set the energy storage system status data and energy storage system work log data in the second time-length sliding window, perform feature extraction to obtain the energy storage system battery feature vector, and construct or update the error prediction model based on Bayesian inference based on the energy storage system battery feature vector and the real-time model error data of the most recent calibration number.

[0011] S6. Based on the error prediction model, predict the real-time model error data for the next calibration, and generate real-time model error prediction data and real-time model error prediction data variance.

[0012] S7. Based on the real-time model error prediction data, perform piecewise linear difference compensation on the real-time state data of the energy storage battery at each time between the current energy storage battery state calibration data and the next energy storage battery state calibration data to generate corresponding real-time state pre-calibration data of the energy storage battery.

[0013] S8. Based on the real-time state pre-calibration data of the energy storage battery, perform energy storage system control analysis and processing. The energy storage system control analysis and processing is based on existing analysis and processing foundations, such as model predictive control (MPC) or other optimization-based energy management strategies. The specific process is common knowledge under the existing technology. This invention only optimizes the estimated state of the energy storage battery that needs to be used. Other steps are not changed. Therefore, they are not described in detail in this technical solution, and will not cause any trouble to the art.

[0014] Furthermore, S1 includes the following steps:

[0015] S1.1 Based on the data requirements of the ampere-hour integration method and EKF algorithm, collect the state parameters of the energy storage system and generate the state data of the energy storage system;

[0016] S1.2 Based on the data requirements of Bayesian inference, supplementary data collection is required to collect the cycle count, usage days, health value and cumulative throughput of each battery in the energy storage system, and generate energy storage system working log data.

[0017] The ampere-hour integration method requires real-time acquisition of the input and output currents of each battery in the energy storage system at various times through high-frequency sampling (1-10Hz). It also acquires the initial state of charge (SOC) and rated capacity of each battery, as well as the coulombic efficiency, current sensor bias, and temperature of each battery at various times. The EKF algorithm further requires acquiring the terminal voltages of each battery corresponding to the input and output currents, obtaining the OCV-SOC curves through experimental calibration, obtaining internal resistance parameters through pulse testing calibration, obtaining polarization resistance and polarization capacitance through electrochemical impedance spectroscopy, and acquiring or setting the initial covariance, process noise covariance, and measurement noise covariance.

[0018] Furthermore, S4 includes the following steps:

[0019] S4.1. Based on the binary search method, search for the real-time state data of the energy storage battery that corresponds to the time of the energy storage battery state calibration data in the real-time state data of the energy storage battery, and obtain the real-time state data of the energy storage battery to be calibrated.

[0020] S4.2. Subtract the real-time state data to be calibrated of the energy storage battery from the energy storage battery state calibration data to obtain real-time model error data. Positive / negative values ​​indicate the direction of error.

[0021] Furthermore, S3 also includes the following steps:

[0022] Record the energy storage battery state calibration data generated each time;

[0023] Determine whether the number of recorded energy storage battery state calibration data is less than the set sample size threshold;

[0024] If so, the energy storage battery status calibration data is used to replace the real-time status data of the energy storage battery at the same time to obtain the real-time status replacement data of the energy storage battery, and the energy storage system control analysis and processing is performed based on the real-time status replacement data of the energy storage battery, and then S1 is returned.

[0025] If not, then execute S4.

[0026] This embodiment allows for the direct replacement of real-time battery status data at the same time with the battery status calibration data when the sample size of the battery status calibration data is small. This achieves the elimination of battery status estimation errors every certain period of time, preventing errors from accumulating over a long period and causing large errors.

[0027] Furthermore, S5 includes the following steps:

[0028] S5.1. Based on the energy storage battery status calibration data at each calibration of the most recently set calibration number, set the energy storage system status data and energy storage system work log data in the second duration sliding window, extract the features of each state parameter of each battery in the energy storage system during charging and discharging, and generate the energy storage system battery feature vector. The extracted features include: such as current-related features, temperature-related features, SOC-related features, cumulative throughput features, battery cycle count features, usage days features, health value features, etc.

[0029] S5.2. Using the real-time model error data from each calibration as output and the corresponding energy storage system battery feature vector as input, Gaussian process regression analysis is performed to generate or update the error prediction model. This allows for the analysis of the prior probabilities of historical energy storage system battery feature vectors and corresponding real-time model error data, resulting in an error prediction model that can predict the corresponding real-time model error based on the energy storage system battery feature vectors. Then, during the next calibration, the error prediction model is updated by adding new energy storage system battery feature vectors and corresponding real-time model error data samples, completing the posterior update of the error prediction model. This ensures that the error prediction model can be updated in real-time to keep pace with the actual situation of the energy storage system, thereby guaranteeing prediction accuracy.

[0030] Furthermore, S6 includes the following steps:

[0031] S6.1 Collect the feature vector of the energy storage system battery after the most recent calibration cycle;

[0032] S6.2 Calculate the vector distance between the energy storage system battery feature vector corresponding to the current calibration number and the energy storage system battery feature vector corresponding to each calibration number, and perform normalization processing to generate calibration number battery feature weight data, where the sum of all calibration number battery feature weight data is equal to 1, and the vector distance can be Euclidean distance or cosine similarity.

[0033] S6.3. Based on the battery feature weight data of the calibration count, the feature vectors of the energy storage system batteries for the corresponding calibration count are weighted, and then all weighted feature vectors of the energy storage system batteries are summed to obtain the predicted feature vector of the energy storage system battery corresponding to the next calibration. For example, if the most recent calibration count is set to the most recent N times, then the feature vector of the energy storage system battery corresponding to the current calibration count is the 0th time (it is necessary to ensure that 0 is greater than N, that is, the current calibration count must be greater than the most recent calibration count). Then the predicted feature vector of the energy storage system battery... The formula is as follows:

[0034]

[0035] in, To calculate the Euclidean distance, i represents the first i calibrations. , Item is with Corresponding calibration count and battery feature weight data;

[0036] S6.4. Based on the error prediction model, predict the real-time model error prediction data and the variance of the real-time model error prediction data corresponding to the predicted feature vector of the energy storage system battery.

[0037] Furthermore, S7 includes the following steps:

[0038] Obtain the current time corresponding to the real-time status data of the energy storage battery;

[0039] Obtain the time corresponding to the current energy storage battery state calibration data corresponding to the current real-time state data of the current energy storage battery, and get the current calibration time;

[0040] Based on the current time, the current calibration time, and the set second duration, the proportion of the time from the current calibration time to the current time to the set second duration is calculated to obtain the compensation coefficient.

[0041] The compensation amount at the current moment is obtained by multiplying the compensation coefficient by the real-time model error prediction data;

[0042] The current compensation amount is used to compensate the current real-time state data of the energy storage battery to obtain the current real-time state pre-calibration data of the energy storage battery. The specific formula is as follows:

[0043] Real-time status pre-calibration data of energy storage battery = real-time status data of energy storage battery + compensation amount at the current moment.

[0044] Furthermore, step S7 also includes setting pre-calibration boundary data based on the variance of the real-time model error prediction data, including the following steps:

[0045] The lower boundary data of the real-time model error prediction is obtained by subtracting the product of the set safety factor and the standard deviation of the real-time model error prediction data from the real-time model error prediction data. The square of the standard deviation of the real-time model error prediction data is equal to the variance of the real-time model error prediction data.

[0046] The upper boundary data of the real-time model error prediction is obtained by adding the product of the set safety factor and the standard deviation of the real-time model error prediction data to the real-time model error prediction data.

[0047] The pre-calibrated lower boundary data is obtained by multiplying the real-time state data of the energy storage battery with the product of the corresponding compensation coefficient and the real-time model error prediction lower boundary data.

[0048] The pre-calibrated upper boundary data is obtained by multiplying the real-time status data of the energy storage battery with the product of the corresponding compensation coefficient and the real-time model error prediction upper boundary data.

[0049] The precalibrated lower boundary data and precalibrated upper boundary data are collected and combined to obtain precalibrated boundary data.

[0050] The safety factor is set based on the confidence interval of the normal distribution. For example, it can be set to 1.96 for a 95% confidence level, 2.33 for a 98% confidence level, and 2.58 for a 99% confidence level.

[0051] An energy storage control system includes an energy storage battery pack, a BMS (Battery Management System), an EMS (Energy Management System), and a PCS (Energy Storage Converter).

[0052] The energy storage battery pack includes multiple energy storage batteries, and the energy storage battery pack can also consist of multiple groups, each group consisting of multiple energy storage batteries.

[0053] The BMS is used to collect and analyze the state parameters of the energy storage battery according to the power storage control method according to any one of claims-8, and to send the collection and analysis results to the EMS and PCS;

[0054] The EMS is used to perform energy storage system control analysis and processing based on the data collected and analyzed by the BMS, generate decision control data, and send the decision control data to the BMS and PCS.

[0055] The PCS is used to charge and discharge the energy storage battery pack based on the results of the data collection and analysis of the BMS and the decision control data of the EMS.

[0056] Beneficial effects:

[0057] 1. Compared with the prior art, the power energy storage control method and system provided by the present invention calibrates the state parameters of the energy storage battery at regular intervals using an advanced estimation algorithm. The intermittent use of the advanced estimation algorithm to calibrate the state parameters of the energy storage battery ensures that the error of the estimated state parameters of the energy storage battery does not accumulate too much, while reducing the resource requirements and the cost of data analysis.

[0058] 2. Compared with the prior art, the power energy storage control method and system provided by the present invention constructs an error prediction model by analyzing the error amount that needs to be calibrated during historical calibration and the prior probabilities corresponding to the characteristic data such as battery parameters and states. Then, based on the similarity between the characteristic data of the energy storage battery at the current calibration and the characteristic data of the energy storage battery at the previous calibrations, the characteristic data of the energy storage battery at the previous calibrations are weighted and summed to analyze and obtain the characteristic input error prediction model of the energy storage battery at the next calibration that conforms to the historical pattern. In this way, the error amount at the next calibration is predicted, and the effect of pre-calibrating the real-time estimated battery state parameters is achieved by predicting the error amount.

[0059] 3. Compared with the prior art, the power storage control method and system provided by the present invention can perform real-time pre-calibration of the estimated battery state parameters based on the predicted error before the next calibration using the advanced estimation algorithm, ensuring the estimation accuracy of the battery state parameters between two calibrations. At the same time, by obtaining the corresponding energy storage battery characteristics and error when the next calibration time point is reached, calculating their corresponding posterior probabilities and updating the error prediction model, the error prediction model is updated in real time and can match the battery state of the energy storage system. Attached Figure Description

[0060] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0061] Figure 1 This is a flowchart illustrating the method steps provided in an embodiment of the present invention;

[0062] Figure 2 This is a system structure block diagram provided for an embodiment of the present invention. Detailed Implementation

[0063] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0064] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0065] Exemplary embodiments will be described more fully below with reference to the accompanying drawings; however, these exemplary embodiments may be embodied in different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will enable those skilled in the art to fully understand the scope of this disclosure.

[0066] Where there is no conflict, the various embodiments of this disclosure and the features thereof in the embodiments may be combined with each other.

[0067] As used herein, the term “and / or” includes any and all combinations of one or more related enumerated entries.

[0068] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. As used herein, the singular forms “a” and “the” are also intended to include the plural forms unless the context clearly indicates otherwise. It will also be understood that when the terms “comprising” and / or “made of” are used in this specification, the presence of the stated feature, integral, step, operation, element, and / or component is specified, but the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof is not excluded.

[0069] The embodiments described herein can be described with reference to plan views and / or cross-sectional views using the ideal schematic diagrams of this disclosure. Therefore, the example illustrations can be modified according to manufacturing techniques and / or tolerances. Therefore, the embodiments are not limited to those shown in the drawings, but include modifications to configurations formed based on manufacturing processes. Therefore, the areas illustrated in the drawings are schematic in nature, and the shapes of the areas shown in the figures illustrate specific shapes of areas of an element, but are not intended to be limiting.

[0070] Unless otherwise specified, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having the same meaning as they have in the context of the relevant art and this disclosure.

[0071] Please see Figure 1 A power storage control method includes the following steps:

[0072] S1. Collect energy storage system status data and energy storage system operation log data, including the following steps:

[0073] S1.1 Based on the ampere-hour integral method and EKF algorithm, the data requirements are collected to obtain the state parameters of the energy storage system and generate the state data of the energy storage system;

[0074] S1.2 Based on the data requirements of Bayesian inference, supplementary data collection is required to collect the cycle count, usage days, health value and cumulative throughput of each battery in the energy storage system, and generate energy storage system working log data.

[0075] The ampere-hour integration method requires real-time acquisition of the input and output currents of each battery in the energy storage system at various times through high-frequency sampling (1-10Hz). It also acquires the initial state of charge (SOC) and rated capacity of each battery, as well as the coulombic efficiency, current sensor bias, and temperature of each battery at various times. The EKF algorithm further requires acquiring the terminal voltages of each battery corresponding to the input and output currents, obtaining the OCV-SOC curves through experimental calibration, obtaining internal resistance parameters through pulse testing calibration, obtaining polarization resistance and polarization capacitance through electrochemical impedance spectroscopy, and acquiring or setting the initial covariance, process noise covariance, and measurement noise covariance.

[0076] S2. Real-time analysis and processing of energy storage system status data is performed based on the ampere-hour integration method to generate real-time status data of energy storage batteries.

[0077] S3. At set intervals, the energy storage system status data is analyzed and processed using the EKF (Extended Kalman Filter) algorithm to generate energy storage battery status calibration data. The ampere-hour integration method and the EKF algorithm are well-known in existing technology and are applied directly without modification; therefore, they will not be elaborated upon in this technical solution, and will not cause any problems for those in the field. Furthermore, the analysis based on the EKF algorithm can be performed using the computing power of a cloud server, further saving costs.

[0078] Furthermore, S3 also includes the following steps:

[0079] Record the energy storage battery state calibration data generated each time;

[0080] Determine whether the number of recorded energy storage battery state calibration data is less than the set sample size threshold;

[0081] If so, the energy storage battery status calibration data is used to replace the real-time status data of the energy storage battery at the same time to obtain the real-time status replacement data of the energy storage battery. Based on the real-time status replacement data of the energy storage battery, the energy storage system control analysis and processing are performed, and the process is returned to S1.

[0082] If not, then execute S4.

[0083] This embodiment allows for the direct replacement of real-time battery status data at the same time with the battery status calibration data when the sample size of the battery status calibration data is small. This achieves the elimination of battery status estimation errors every certain period of time, preventing errors from accumulating over a long period and causing large errors.

[0084] S4. Based on the deviation between the real-time state data of the energy storage battery and the state calibration data of the energy storage battery at the same moment, the real-time model error data is calculated, including the following steps:

[0085] S4.1. Based on the binary search method, search for the real-time state data of the energy storage battery that corresponds to the time of the energy storage battery state calibration data in the real-time state data of the energy storage battery, and obtain the real-time state data of the energy storage battery to be calibrated.

[0086] S4.2. Subtract the real-time state data to be calibrated from the energy storage battery state calibration data to obtain the real-time model error data. Positive / negative values ​​indicate the direction of error.

[0087] S5. Based on the energy storage battery status calibration data at each calibration, set the energy storage system status data and energy storage system operation log data within the second time-duration sliding window, perform feature extraction to obtain the energy storage system battery feature vector, and construct or update the Bayesian inference-based error prediction model based on the energy storage system battery feature vector and the real-time model error data of the most recent calibration, including the following steps:

[0088] S5.1. Based on the energy storage battery status calibration data at each calibration of the most recently set calibration number, set the energy storage system status data and energy storage system work log data in the second duration sliding window, extract the features of each state parameter of each battery in the energy storage system during charging and discharging, and generate the energy storage system battery feature vector. The extracted features include: such as current-related features, temperature-related features, SOC-related features, cumulative throughput features, battery cycle count features, usage days features, health value features, etc.

[0089] S5.2. Using the real-time model error data from each calibration as output and the corresponding energy storage system battery feature vector as input, Gaussian process regression analysis is performed to generate or update the error prediction model. This allows for the analysis of the prior probabilities of historical energy storage system battery feature vectors and corresponding real-time model error data, resulting in an error prediction model that can predict the corresponding real-time model error based on the energy storage system battery feature vectors. Then, during the next calibration, the error prediction model is updated by adding new energy storage system battery feature vectors and corresponding real-time model error data samples, completing the posterior update of the error prediction model. This ensures that the error prediction model can be updated in real-time to keep pace with the actual situation of the energy storage system, thereby guaranteeing prediction accuracy.

[0090] S6. Based on the error prediction model, predict the real-time model error data for the next calibration, and generate real-time model error prediction data and real-time model error prediction data variance, including the following steps:

[0091] S6.1 Collect the feature vector of the energy storage system battery after the most recent calibration cycle;

[0092] S6.2 Calculate the vector distance between the energy storage system battery feature vector corresponding to the current calibration number and the energy storage system battery feature vector corresponding to each calibration number, and perform normalization processing to generate calibration number battery feature weight data, where the sum of all calibration number battery feature weight data is equal to 1, and the vector distance can be Euclidean distance or cosine similarity.

[0093] S6.3. Based on the battery feature weight data of the calibration count, the feature vectors of the energy storage system batteries for the corresponding calibration count are weighted, and then all weighted feature vectors of the energy storage system batteries are summed to obtain the predicted feature vector of the energy storage system battery corresponding to the next calibration. For example, if the most recent calibration count is set to the most recent N times, then the feature vector of the energy storage system battery corresponding to the current calibration count is the 0th time (it is necessary to ensure that 0 is greater than N, that is, the current calibration count must be greater than the most recent calibration count). Then the predicted feature vector of the energy storage system battery... The prediction formula is as follows:

[0094]

[0095] in, To calculate the Euclidean distance, i represents the first i calibrations. , Item is with Corresponding calibration count and battery feature weight data;

[0096] S6.4. Based on the error prediction model, predict the real-time model error prediction data and the variance of the real-time model error prediction data corresponding to the predicted feature vector of the energy storage system battery.

[0097] S7. Based on the real-time model error prediction data, perform piecewise linear difference compensation on the real-time state data of the energy storage battery at each time point between the current energy storage battery state calibration data and the next energy storage battery state calibration data to generate corresponding real-time state pre-calibration data of the energy storage battery. S7 includes the following steps:

[0098] S7.01. Obtain the time corresponding to the current real-time status data of the energy storage battery to get the current time;

[0099] S7.02. Obtain the time corresponding to the current energy storage battery state calibration data corresponding to the current real-time state data of the energy storage battery, and obtain the current calibration time;

[0100] S7.03. Based on the current time, the current calibration time, and the set second duration, calculate the proportion of the time from the current calibration time to the current time to the set second duration, and obtain the compensation coefficient;

[0101] S7.04. Based on the product of the compensation coefficient and the real-time model error prediction data, the compensation amount at the current moment is obtained;

[0102] S7.05. Compensate the current real-time state data of the energy storage battery using the compensation amount at the current moment to obtain the pre-calibrated real-time state data of the energy storage battery at the current moment. The specific formula is as follows:

[0103] Real-time status pre-calibration data of energy storage battery = real-time status data of energy storage battery + compensation amount at the current moment.

[0104] S7 also includes setting pre-calibration boundary data based on the variance of real-time model error prediction data, including the following steps:

[0105] S7.06. Subtract the product of the set safety factor and the standard deviation of the real-time model error prediction data from the real-time model error prediction data to obtain the lower boundary data of the real-time model error prediction. The square of the standard deviation of the real-time model error prediction data is equal to the variance of the real-time model error prediction data.

[0106] S7.07. Add the product of the set safety factor and the standard deviation of the real-time model error prediction data to the real-time model error prediction data to calculate the upper boundary data of the real-time model error prediction.

[0107] S7.08. Calculate the pre-calibrated lower boundary data by adding the corresponding compensation coefficient to the real-time state data of the energy storage battery and the product of the real-time model error prediction lower boundary data.

[0108] S7.09. Calculate the pre-calibrated upper boundary data by adding the corresponding compensation coefficient to the real-time status data of the energy storage battery and the product of the real-time model error prediction upper boundary data.

[0109] S7.10. Collect and combine the pre-calibration lower boundary data and the pre-calibration upper boundary data to obtain the pre-calibration boundary data.

[0110] The safety factor is set based on the confidence interval of the normal distribution. For example, it can be set to 1.96 for a 95% confidence level, 2.33 for a 98% confidence level, and 2.58 for a 99% confidence level.

[0111] S8. Based on the real-time state pre-calibration data of the energy storage battery, the energy storage system control analysis and processing is performed. The energy storage system control analysis and processing is based on existing analysis and processing foundations, such as model predictive control (MPC) or other optimization-based energy management strategies. The specific process is common knowledge under the existing technology. This invention only optimizes the estimated state of the energy storage battery that needs to be used. Other steps are not changed. Therefore, they are not described in detail in this technical solution, and will not cause any trouble to the art.

[0112] Please see Figure 2The present invention also provides a power energy storage control system, including an energy storage battery pack, a BMS (Battery Management System), an EMS (Energy Management System), and a PCS (Energy Storage Converter).

[0113] An energy storage battery pack includes multiple energy storage batteries, and an energy storage battery pack can also consist of multiple groups, each group consisting of multiple energy storage batteries;

[0114] The BMS is used to collect and analyze the state parameters of the energy storage battery according to the power storage control method provided by the present invention, and to send the collection and analysis results to the EMS and PCS.

[0115] EMS is used to perform control analysis and processing of energy storage system based on the data collected and analyzed by BMS, generate decision control data, and send the decision control data to BMS and PCS.

[0116] PCS is used to collect and analyze data from BMS and make decisions and controls data from EMS for charging and discharging energy storage battery packs.

[0117] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A power storage control method, characterized in that, Includes the following steps: S1. Collect energy storage system status data and energy storage system operation log data; S2. Based on the ampere-hour integral method, perform real-time analysis and processing of the energy storage system status data to generate real-time status data of the energy storage battery. S3. Every set first time interval, the energy storage system status data is analyzed and processed based on the EKF algorithm to generate energy storage battery status calibration data. S4. Based on the deviation between the real-time state data of the energy storage battery and the state calibration data of the energy storage battery at the same moment, the real-time model error data is calculated. S5. Based on the energy storage battery status calibration data at each calibration, set the energy storage system status data and energy storage system work log data in the second time-length sliding window, perform feature extraction to obtain the energy storage system battery feature vector, and construct or update the error prediction model based on Bayesian inference based on the energy storage system battery feature vector and the real-time model error data of the most recent calibration number. S6. Based on the error prediction model, predict the real-time model error data for the next calibration, and generate real-time model error prediction data and real-time model error prediction data variance. S7. Based on the real-time model error prediction data, perform piecewise linear difference compensation on the real-time state data of the energy storage battery at each time between the current energy storage battery state calibration data and the next energy storage battery state calibration data to generate corresponding real-time state pre-calibration data of the energy storage battery. S8. Perform energy storage system control analysis and processing based on the real-time state pre-calibration data of the energy storage battery.

2. The power storage control method according to claim 1, characterized in that, S1 includes the following steps: S1.1 Based on the data requirements of the ampere-hour integration method and EKF algorithm, collect the state parameters of the energy storage system and generate the state data of the energy storage system; S1.2 Based on the data requirements of Bayesian inference, supplementary data collection is required to collect the cycle count, usage days, health value and cumulative throughput of each battery in the energy storage system, and generate energy storage system working log data.

3. The power storage control method according to claim 1, characterized in that... S4 includes the following steps: S4.

1. Based on the binary search method, search for the real-time state data of the energy storage battery that corresponds to the time of the energy storage battery state calibration data in the real-time state data of the energy storage battery, and obtain the real-time state data of the energy storage battery to be calibrated. S4.

2. Subtract the real-time state data to be calibrated of the energy storage battery from the energy storage battery state calibration data to obtain the real-time model error data.

4. The power storage control method according to claim 1, characterized in that, S3 further includes the following steps: Record the energy storage battery state calibration data generated each time; Determine whether the number of recorded energy storage battery state calibration data is less than the set sample size threshold; If so, the energy storage battery status calibration data is used to replace the real-time status data of the energy storage battery at the same time to obtain the real-time status replacement data of the energy storage battery, and the energy storage system control analysis and processing is performed based on the real-time status replacement data of the energy storage battery, and then S1 is returned. If not, then execute S4.

5. The power storage control method according to claim 1, characterized in that, S5 includes the following steps: S5.

1. Based on the energy storage battery status calibration data at each calibration of the most recently set calibration number, set the energy storage system status data and energy storage system work log data in the second time-length sliding window, extract the features of each state parameter of each battery in the energy storage system during charging and discharging, and generate the energy storage system battery feature vector. S5.

2. Using the real-time model error data from each calibration as the output and the corresponding energy storage system battery feature vector as the input, perform Gaussian process regression analysis to generate or update the error prediction model.

6. The power storage control method according to claim 1, characterized in that, S6 includes the following steps: S6.1 Collect the feature vector of the energy storage system battery after the most recent calibration cycle; S6.2 Calculate the vector distance between the energy storage system battery feature vector corresponding to the current calibration number and the energy storage system battery feature vector corresponding to each calibration number, and perform normalization processing to generate calibration number battery feature weight data, wherein the sum of all calibration number battery feature weight data is equal to 1; S6.

3. Based on the battery feature weight data of the calibration number, the feature vector of the energy storage system battery for the corresponding calibration number is weighted, and then all the weighted feature vectors of the energy storage system battery are summed to obtain the predicted feature vector of the energy storage system battery corresponding to the next calibration. S6.

4. Based on the error prediction model, predict the real-time model error prediction data and the variance of the real-time model error prediction data corresponding to the predicted feature vector of the energy storage system battery.

7. The power storage control method according to claim 1, characterized in that, S7 includes the following steps: Obtain the current time corresponding to the real-time status data of the energy storage battery; Obtain the time corresponding to the current energy storage battery state calibration data corresponding to the current real-time state data of the current energy storage battery, and get the current calibration time; Based on the current time, the current calibration time, and the set second duration, the proportion of the time from the current calibration time to the current time to the set second duration is calculated to obtain the compensation coefficient. The compensation amount at the current moment is obtained by multiplying the compensation coefficient by the real-time model error prediction data; The current moment compensation amount is used to compensate the current real-time state data of the energy storage battery to obtain the current moment's real-time state pre-calibration data of the energy storage battery.

8. The power storage control method according to claim 7, characterized in that, S7 further includes setting pre-calibration boundary data based on the variance of the real-time model error prediction data, including the following steps: The lower boundary data of the real-time model error prediction is obtained by subtracting the product of the set safety factor and the standard deviation of the real-time model error prediction data from the real-time model error prediction data. The square of the standard deviation of the real-time model error prediction data is equal to the variance of the real-time model error prediction data. The upper boundary data of the real-time model error prediction is obtained by adding the product of the set safety factor and the standard deviation of the real-time model error prediction data to the real-time model error prediction data. The pre-calibrated lower boundary data is obtained by multiplying the real-time state data of the energy storage battery with the product of the corresponding compensation coefficient and the real-time model error prediction lower boundary data. The pre-calibrated upper boundary data is obtained by multiplying the real-time status data of the energy storage battery with the product of the corresponding compensation coefficient and the real-time model error prediction upper boundary data. The precalibrated lower boundary data and precalibrated upper boundary data are collected and combined to obtain precalibrated boundary data.

9. A power energy storage control system, used to execute the power energy storage control method according to any one of claims 1-8, characterized in that: This includes energy storage battery packs, BMS, EMS, and PCS; The energy storage battery pack includes multiple energy storage batteries; BMS is used to collect and analyze the state parameters of the energy storage battery according to the power storage control method according to any one of claims 1-8, and send the collection and analysis results to EMS and PCS; The EMS is used to perform energy storage system control analysis and processing based on the data collected and analyzed by the BMS, generate decision control data, and send the decision control data to the BMS and PCS. The PCS is used to charge and discharge the energy storage battery pack based on the results of the data collection and analysis of the BMS and the decision control data of the EMS.

Images