Safety scheduling analysis method and system for black start of electrochemical energy storage power station unit

Abstract

The invention provides a safety scheduling analysis method and system for black start of an electrochemical energy storage power station unit, and relates to the technical field of information, and the method comprises the steps: obtaining the real-time charge state data of each energy storage unit in an electrochemical energy storage power station; and carrying out noise removal and smoothing processing on the charge state data by adopting a Kalman filtering algorithm to obtain an accurate charge state estimation value. According to the invention, fault diagnosis and early warning are carried out through a support vector machine algorithm, the state of each unit is monitored in real time, and problem units are isolated timely. In the black start process, residual energy is continuously evaluated, a threshold value and a scheduling strategy are dynamically adjusted, safe and stable operation is ensured, after black start, the influence of the process on an energy storage unit is analyzed, a management strategy and parameter setting are optimized, and preparation is made for subsequent black start. According to the scheme, the black-start reliability and the energy utilization efficiency of the electrochemical energy storage power station are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a safety dispatching analysis method and system for black start of an electrochemical energy storage power station unit. Background Art

[0002] Electrochemical energy storage power stations face a key technical contradiction when performing black starts: in the absence of external power support, the energy storage units inside the power station must have sufficient state of charge (SOC) to provide the energy required to start and maintain the operation of the power station until the external power grid is restored. However, setting the SOC lower limit too high will limit the discharge depth of the power station and reduce the utilization efficiency of the energy storage unit; while setting the SOC lower limit too low may cause the power station to run out of energy during the black start process and be unable to continue operating. Therefore, it is crucial to accurately determine the lower limit threshold of the SOC.

[0003] This requires the power station's energy management system to be able to monitor the SOC status of each energy storage unit in real time, and dynamically adjust the SOC lower limit threshold according to the power station's load demand and the estimated black start duration. At the same time, it is also necessary to consider the impact of factors such as the energy storage unit's self-discharge rate and energy conversion efficiency on the SOC, as well as the impact of external conditions such as ambient temperature on the performance of the energy storage unit. In addition, during the actual black start process, some energy storage units may fail or their performance degrades, which requires the energy management system to detect and isolate the faulty units in a timely manner, and re-optimize the scheduling of other energy storage units to ensure the safe and stable operation of the power station. Summary of the invention

[0004] The purpose of the present invention is to overcome the problems existing in the prior art and to provide a safe scheduling analysis method and system for black start of electrochemical energy storage power station units.

[0005] In order to achieve the above-mentioned object, the first aspect of the present invention provides a safety dispatch analysis method for black start of an electrochemical energy storage power station unit, comprising the following steps:

[0006] Obtain real-time state of charge data of each energy storage unit in the electrochemical energy storage power station;

[0007] The Kalman filter algorithm is used to remove noise and smooth the state of charge data to obtain the estimated value of the state of charge;

[0008] Based on historical operation data, a mathematical model between the state of charge of the energy storage unit and the discharge power, energy conversion efficiency, and ambient temperature is established through least squares fitting;

[0009] The available energy of the energy storage unit under different state of charge lower limit threshold settings is predicted using the state of charge estimation value and mathematical model;

[0010] Real-time load demand data and ambient temperature data are obtained, and the optimal lower limit threshold of the charge state is solved based on the mathematical model using a dynamic programming algorithm.

[0011] Preferably, the Kalman filter algorithm is used to remove noise and smooth the state of charge data, specifically:

[0012] Perform format conversion and outlier filtering on the state of charge data, establish a state space model, and dynamically adjust the process noise covariance matrix Q and the measurement noise covariance matrix R;

[0013] In the filtering process, by comparing the residuals of the filter estimate and the measured value, the Q and R matrices are dynamically adjusted to minimize the mean square error of the residuals;

[0014] The state of charge estimation value after filtering is smoothed, and the exponential smoothing method is used to set the smoothing factor α to perform smoothing calculation on the estimation value sequence.

[0015] Preferably, the mathematical model building process is:

[0016] With the state of charge of the energy storage unit as the independent variable, and the discharge power, energy conversion efficiency and ambient temperature as the dependent variables, mathematical fitting is performed, and the accuracy and generalization ability of the model are evaluated by the cross-validation method, and the model is optimized and adjusted according to the evaluation results.

[0017] Preferably, the method of using the state of charge estimation value and the mathematical model to predict the available energy of the energy storage unit under different state of charge lower limit threshold settings is specifically as follows:

[0018] The estimated state of charge of the current energy storage unit and the ambient temperature are substituted into the mathematical model as input variables to calculate the predicted value of the available energy of the energy storage unit under different threshold settings.

[0019] Preferably, the real-time load demand data and ambient temperature data are obtained, and a dynamic programming algorithm is used to solve the optimal lower limit threshold of the state of charge based on a mathematical model, specifically:

[0020] The average black start duration is obtained using the statistical regression method, and the state equation and objective function are constructed to calculate the black start reliability of the power station and the utilization efficiency of the energy storage unit corresponding to the lower limit thresholds of different charge states.

[0021] Preferably, the step of acquiring real-time load demand data and ambient temperature data and solving the optimal charge state lower limit threshold based on a mathematical model using a dynamic programming algorithm also includes the following steps:

[0022] Acquire real-time monitoring data of energy storage units, clean and normalize the data, extract key characteristic parameters reflecting the performance of energy storage units, and use pre-trained support vector machine models for fault diagnosis;

[0023] When the electrochemical energy storage power station is in black start mode, the state of charge of each energy storage unit is monitored in real time. If the state of charge of a certain energy storage unit drops to the lower limit threshold solved by the dynamic programming algorithm, the energy storage unit is controlled to stop discharging and its load is transferred to other energy storage units.

[0024] During the black start process, the remaining available energy of the power station is continuously evaluated. If the remaining available energy is predicted to be lower than the preset safety margin, the lower limit threshold of the state of charge is recalculated and appropriately increased according to the mathematical model to avoid energy depletion caused by excessive discharge of the energy storage unit. At the same time, the dynamic programming algorithm is re-run to optimize the scheduling of other non-isolated energy storage units to ensure safe and stable operation of the power station.

[0025] After the black start is completed, the charge state and health status data of each energy storage unit are obtained, and the impact of the black start process on the performance of the energy storage unit is analyzed through the decision tree algorithm, and a comprehensive evaluation is performed in combination with the fault diagnosis results. Based on the analysis results, the energy management strategy and the charge state lower limit threshold setting are optimized, and the mathematical model established by the least squares method and the parameters of the dynamic programming algorithm are updated to prepare for subsequent black starts.

[0026] The second aspect of the present invention provides a safety dispatching and analysis system for black start of electrochemical energy storage power station units, including a data acquisition module, a data preprocessing module, a digital modeling module, a prediction and analysis module, an optimization decision module, and a fault diagnosis and isolation module;

[0027] A data acquisition module is used to obtain real-time state of charge data of each energy storage unit in the electrochemical energy storage power station;

[0028] A data preprocessing module is used to remove noise and smooth the state of charge data using a Kalman filter algorithm to obtain a state of charge estimation value;

[0029] Digital modeling module, used to establish a mathematical model between the state of charge of the energy storage unit and the discharge power, energy conversion efficiency, and ambient temperature through least squares fitting based on historical operating data;

[0030] A prediction and analysis module, used to predict the available energy of the energy storage unit under different state of charge lower limit threshold settings using the state of charge estimation value and the mathematical model;

[0031] The optimization decision module is used to obtain real-time load demand data and ambient temperature data, and uses a dynamic programming algorithm to solve the optimal lower limit threshold of the charge state based on a mathematical model.

[0032] Preferably, the data acquisition module is used to obtain the real-time state of charge data of each energy storage unit in the electrochemical energy storage power station, specifically:

[0033] A state of charge sensor is installed on each energy storage unit to collect state of charge data in real time, and the data is transmitted to the data cache area of ​​the data processing center through the data acquisition module.

[0034] The third aspect of the present invention provides a computer device, comprising: a memory and a processor; the memory stores a computer program, characterized in that: when the processor executes the computer program, the steps of the above-mentioned safety scheduling analysis method for black start of the electrochemical energy storage power station unit are implemented.

[0035] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the above-mentioned safe scheduling analysis method for black start of an electrochemical energy storage power station unit are implemented.

[0036] The technical solution of the present invention may include the following beneficial effects:

[0037] The present invention discloses a black start energy management method for an electrochemical energy storage power station. The method collects the state of charge data of the energy storage unit in real time, removes noise using a Kalman filter algorithm, and obtains an accurate estimate; a mathematical model of parameters such as the state of charge and discharge power is established in combination with historical operation data to predict the available energy under different thresholds. Considering the load demand and ambient temperature, a dynamic programming algorithm is used to solve the optimal state of charge lower limit threshold, maximizing the utilization efficiency while ensuring a reliable black start. The present invention also performs fault diagnosis and early warning through a support vector machine algorithm, monitors the state of each unit in real time, and isolates the problem unit in a timely manner. During the black start process, the remaining energy is continuously evaluated and the threshold and scheduling strategy are dynamically adjusted to ensure safe and stable operation. After the black start, the impact of the process on the energy storage unit is analyzed, the management strategy and parameter settings are optimized, and preparations are made for subsequent black starts. The present invention significantly improves the black start reliability and energy utilization efficiency of the electrochemical energy storage power station. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 The present invention provides a flow chart of a safety dispatch analysis method and system for black start of an electrochemical energy storage power station unit. DETAILED DESCRIPTION

[0039] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0040] A first aspect of an embodiment of the present invention provides a safety dispatch analysis method for black start of an electrochemical energy storage power station unit, such as Figure 1 The safety dispatch analysis method for black start of electrochemical energy storage power station units in this embodiment may specifically include:

[0041] Step S101, obtaining real-time state of charge data of each energy storage unit in the electrochemical energy storage power station, and using a Kalman filter algorithm to remove noise and smooth the obtained state of charge data to obtain an accurate state of charge estimation value.

[0042] According to the distribution of energy storage units in the electrochemical energy storage power station, a state of charge sensor is installed on each energy storage unit to collect state of charge data in real time. The data acquisition frequency is set to once per second, and the real-time state of charge data of the energy storage unit is obtained through the data acquisition module. The data acquisition module transmits the collected data to the data buffer area of ​​the data processing center through industrial Ethernet, and stores it according to the timestamp and energy storage unit number. In the data processing center, the received state of charge data is preprocessed. The data format is converted to convert the original data into a unified data format. The data is filtered for outliers, and the abnormal data points beyond the normal range are eliminated by setting the threshold range to ensure the validity and integrity of the data. For the preprocessed state of charge data, the Kalman filter algorithm is used for noise removal and smoothing. According to the electrochemical characteristics and noise level of the energy storage unit, the state space model is established, and the initial values ​​of the process noise covariance matrix Q and the measurement noise covariance matrix R are reasonably set. In the filtering process, by comparing the residuals of the filter estimate and the measured value, the Q and R matrices are dynamically adjusted to minimize the mean square error of the residual. The estimated state of charge value after Kalman filtering is smoothed, and the exponential smoothing method is used to set the smoothing factor α to smooth the estimated value sequence. By adjusting the α value and controlling the degree of smoothing, the fluctuation and mutation of the estimated value can be effectively reduced, and the trend of the change of the state of charge can be reflected in time, so as to obtain a stable and reliable state of charge estimation curve. The processed state of charge estimation value is stored in the form of a relational database table, which includes the timestamp, energy storage unit number, and state of charge estimation value fields. The database table structure and index are designed to ensure efficient access and query of data. Through the database interface, real-time data update and historical data management are realized, providing reliable data support for subsequent battery health status assessment and energy management strategy formulation. A Web-based visual monitoring interface is developed to display the state of charge estimation results of each energy storage unit in real time in the form of charts and curves. Intuitive user interaction functions are designed. At the same time, data export and report generation functions are provided to support operation and maintenance personnel to conduct offline analysis and decision support for historical data.

[0043] Exemplarily, the data is formatted and the raw data is converted into a unified data format, such as JSON or CSV format. Intuitive user interaction functions are designed, such as data filtering, time range selection, and curve comparison, to facilitate operation and maintenance personnel to perform data analysis and abnormal diagnosis. In an electrochemical energy storage power station, the state of charge sensor installed on each energy storage unit collects data in real time, and these data are updated once a second to ensure the timeliness and accuracy of the data. In the filtering process, by comparing the residuals of the filter estimate and the measured value, the Q and R matrices are dynamically adjusted to minimize the mean square error of the residual, thereby optimizing the performance of the filter and improving the accuracy of the state of charge estimation. For example, if the battery voltage of an energy storage unit suddenly drops, this change can be captured immediately and quickly transmitted to the data processing center via industrial Ethernet. In the data processing center, the received data is first formatted, such as converting the original binary format data into a JSON format that is easier to process and store. This step is necessary because data in JSON format is easier to parse and process by various programming languages ​​and database technologies. Next, filtering the data for outliers is another key step. Set a reasonable threshold range. For example, if the voltage of a certain energy storage unit exceeds the normal operating range (assuming the normal range is 0 to 7 volts), the system will automatically identify this and mark it as abnormal data. Doing so can avoid misleading information in subsequent analysis and ensure the accuracy of data analysis. After data preprocessing is completed, using the Kalman filter algorithm to remove noise and smooth the data is an important part of improving data quality. Kalman filtering is an effective recursive filter that can estimate the state of a dynamic system in a noisy environment. In practical applications, according to the electrochemical characteristics and noise level of the energy storage unit, a state space model can be established, and the process noise covariance matrix Q and the measurement noise covariance matrix R can be reasonably set. For example, if the energy storage unit is greatly affected by temperature fluctuations, the process noise covariance can cleverly adjust the Q and R matrices, so that the filter can track the actual state of charge more accurately, thereby minimizing the mean square error of the residual. The processed state of charge estimate also needs to be smoothed to reduce the volatility of the estimate. Using the exponential smoothing method, by setting the smoothing factor α (for example, α = 2), the stability of the estimate and the response speed to new changes can be effectively balanced. The processed data is not only smooth, but also able to timely reflect the real trend of the state of charge. Finally, the processed data is stored in a relational database, and a reasonable database table structure and index are designed to ensure efficient access and query of data. For example, a database table containing timestamps, energy storage unit numbers, and estimated state of charge values ​​can be created. By creating indexes, query speed can be accelerated to support efficient data analysis and decision making.In addition, a Web-based visual monitoring interface is developed to display the state of charge of each energy storage unit in real time. Through charts and curves, operation and maintenance personnel can intuitively understand the system status, perform data screening and time range selection, and facilitate data analysis and abnormal diagnosis. This system design not only improves the operating efficiency of the electrochemical energy storage power station, but also enhances the safety and reliability of the system.

[0044] Step S102, based on the historical operation data of the electrochemical energy storage power station, a mathematical model between the state of charge of the energy storage unit and its discharge power, energy conversion efficiency, and ambient temperature is established by least squares fitting. This model is used to predict the available energy of the energy storage unit under different state of charge lower limit threshold settings in combination with the accurate state of charge estimate obtained in the previous step and the current ambient temperature.

[0045] The historical operation data of the electrochemical energy storage power station is obtained, including the state of charge, discharge power, energy conversion efficiency and ambient temperature parameters of the energy storage unit. The historical operation data obtained is preprocessed, abnormal data is eliminated, and the data is normalized. The least squares method is used to fit and establish a mathematical model with the state of charge of the energy storage unit as the independent variable and the discharge power, energy conversion efficiency and ambient temperature as the dependent variables. The accuracy and generalization ability of the established mathematical model are evaluated by the cross-validation method, and the model is optimized and adjusted according to the evaluation results. The accurate state of charge estimation value and ambient temperature of the current energy storage unit are obtained and substituted into the established mathematical model as input variables. Different lower limit thresholds of the state of charge are set in the model, and the predicted value of the available energy of the energy storage unit under different threshold settings is obtained through model calculation. The predicted available energy is compared with the actual available energy to evaluate the accuracy of the prediction model, and the model is further optimized according to the evaluation results to improve the prediction accuracy.

[0046] For example, in the operation and management of electrochemical energy storage power stations, the acquisition and processing of historical operation data is a crucial link. For example, through sensors installed on energy storage units, key parameters such as state of charge, discharge power, energy conversion efficiency and ambient temperature can be monitored and recorded in real time. After these data are collected, they need to be preprocessed first to ensure the quality and availability of the data. In the preprocessing process, the removal of abnormal data is a typical operation. For example, if the discharge power of a data point suddenly jumps to an unreasonable high value, this is likely caused by sensor failure or data transmission error, and such data should be excluded from the analysis data set. Data normalization is also an important step in preprocessing, which can eliminate the effects of different dimensions and make model training more stable. For example, converting all temperature data to between 0 and 1 can avoid the adverse effects of scale differences between dependent variables on the results when performing multivariate analysis. After the data preprocessing is completed, it is a common method to use the least squares method to establish a mathematical model. With the state of charge as the independent variable and the discharge power, energy conversion efficiency and ambient temperature as the dependent variables, a multivariate linear regression model can be established. This model can help understand how the discharge efficiency of energy storage units and environmental factors affect energy conversion efficiency under different states of charge. In order to verify the accuracy and generalization ability of the model, cross-validation is an effective method. By dividing the data set into multiple groups, using one group as the test set in turn and the rest as the training set, the performance of the model on unseen data can be evaluated. This method can effectively avoid model overfitting and ensure that the model has good generalization ability. After the model is established and verified, the real-time data of the current energy storage unit is input into the model to predict the discharge power and energy conversion efficiency under different states of charge. For example, if the current state of charge is 80% and the ambient temperature is 25℃, the model can help predict the discharge power and energy conversion efficiency under such conditions. In addition, by setting different lower limit thresholds of the state of charge, the model can be used to calculate the predicted value of the available energy of the energy storage unit under these thresholds. This is of great significance for the operation and management of power plants. The operation strategy of the power plant can be adjusted according to the prediction results, energy use can be optimized, and economic benefits can be improved. Finally, comparing the predicted available energy with the actual available energy is a direct way to evaluate the accuracy of the prediction model. This comparison can help further adjust and optimize the model and improve the accuracy of predictions, thereby better serving the operation and management of power plants.

[0047] Step S103, obtain the real-time load demand data and ambient temperature data of the electrochemical energy storage power station, combine the average black start duration obtained based on the statistical analysis of historical data, and use the dynamic programming algorithm to solve the optimal lower limit threshold of the charge state. The mathematical model established in the second step is considered in the threshold calculation process to maximize the utilization efficiency of the energy storage unit while ensuring the reliable black start of the power station.

[0048] The real-time load demand data and ambient temperature data of the electrochemical energy storage power station are obtained and used as the input of the optimization algorithm. According to the historical operation data of the electrochemical energy storage power station, the statistical regression method is used to obtain the average black start duration as the time span of the optimization algorithm. The mathematical model of the electrochemical energy storage power station is established. Combining the dynamic characteristics, state of charge and ambient temperature factors of the energy storage battery, the state equation and objective function are constructed to calculate the black start reliability of the power station and the utilization efficiency of the energy storage unit corresponding to different lower limit thresholds of the state of charge. The dynamic programming algorithm is used to solve the mathematical model, and the lower limit threshold of the optimal state of charge is calculated with the black start reliability of the power station and the utilization efficiency of the energy storage unit as the optimization objectives. The basic idea of ​​the dynamic programming algorithm is to decompose the problem into interrelated sub-problems, and solve them from bottom to top in a recursive way, and finally obtain the global optimal solution. If the optimal lower limit threshold of the state of charge obtained meets the reliable black start requirements of the power station, then the threshold is used as the basis for the operation control of the power station; otherwise, the threshold is appropriately increased, and the optimization calculation in step 4 is repeated until the requirements are met. According to the determined lower limit threshold of the optimal state of charge, the charging and discharging operation of the electrochemical energy storage power station is controlled. When the state of charge is lower than the threshold, charging is given priority to ensure that the power station has reliable black start capability. When the state of charge is higher than the threshold, it can be flexibly charged and discharged according to the grid dispatching needs to improve the utilization efficiency of the energy storage unit.

[0049] For example, obtaining real-time load demand data and ambient temperature data of the electrochemical energy storage power station is crucial for the optimization algorithm. Real-time load demand data, such as the power consumption of a factory at different time points, can help understand when and how much power the energy storage power station needs to discharge. Ambient temperature data, such as the current temperature is 25 degrees Celsius, will affect the performance and efficiency of the battery, so it needs to be incorporated. These data are like ingredients for cooking. With them, delicious "dishes" (optimal operation strategies) can be made according to the "recipe" (optimization algorithm). The average black start duration can be obtained through statistical regression analysis of historical operation data. For example, records of black starts of energy storage power stations in the past year were collected, and it was found that the average duration was 3 hours each time. This time will be used as the time span of the optimization algorithm, which means ensuring that the energy storage power station can provide stable power supply within 3 hours after the power outage. This is like planning a long journey. You need to know how far the destination is to arrange the itinerary. The mathematical model of the electrochemical energy storage power station is established to simulate the actual operation of the power station. It is necessary to consider the dynamic characteristics of the energy storage battery, such as the charging and discharging rate; the state of charge, such as the current battery remaining power of 80%; and the ambient temperature, such as 25 degrees Celsius. These factors will affect the performance of the energy storage power station. With this information, the state equation and objective function can be constructed, just like building a virtual power station for simulation and optimization in the computer. For example, the objective function can be set to maximize the black start reliability and the utilization efficiency of the energy storage unit. The dynamic programming algorithm is like a shrewd planner. It breaks down the problem into multiple small problems and solves them one by one. For example, it will first calculate the impact of different discharge strategies on the black start reliability and the utilization efficiency of the energy storage unit in the next hour under the current state of charge. Then, based on the results of the next hour, the situation in the next two hours and three hours is calculated until the entire average black start duration is covered. In this way, the dynamic programming algorithm can find the optimal state of charge lower limit threshold, thereby achieving the best strategy for power station operation. If the optimal state of charge lower limit threshold calculated by the dynamic programming algorithm, such as 60%, can meet the requirements of reliable black start of the power station, then this threshold can be used as the basis for power station operation control. If the requirements cannot be met, for example, the calculation results show that the 60% lower limit of the state of charge threshold cannot guarantee a reliable black start, then the threshold needs to be appropriately increased, for example, to 70%, and then the optimization calculation is re-performed until an optimal threshold is found that can both guarantee a reliable black start and take into account the utilization efficiency of the energy storage unit. This is like adjusting the grinding degree of a coffee machine until you find the coffee that best suits your taste. Based on the ultimately determined optimal lower limit of the state of charge threshold, the charging and discharging operation of the electrochemical energy storage power station can be controlled.For example, if the lower threshold of the optimal state of charge is set to 70%, when the state of charge of the energy storage power station is lower than 70%, the system will give priority to charging the battery to ensure that the power station has reliable black start capability, just like when the fuel level of a car is lower than the warning line, it will give priority to refueling. When the state of charge is higher than 70%, it can be flexibly charged and discharged according to the grid dispatching needs, such as charging during low power consumption and discharging during peak power consumption, thereby improving the utilization efficiency of the energy storage unit and obtaining economic benefits. This is like being able to arrange expenses more flexibly when the budget is sufficient.

[0050] Step S104: Based on the real-time monitoring data of the energy storage unit, a support vector machine algorithm is used to perform fault diagnosis to achieve early warning of performance degradation or faulty energy storage units. The diagnosed problematic energy storage units are isolated in a timely manner.

[0051] The real-time monitoring data of the energy storage unit is obtained, and the data is cleaned and normalized to extract the key characteristic parameters reflecting the performance of the energy storage unit. According to the extracted characteristic parameters, the pre-trained support vector machine model is used to diagnose the fault of the energy storage unit to obtain the fault type and severity diagnosis results. According to the fault diagnosis results, it is judged whether the energy storage unit has performance degradation or failure. If the diagnosis results show that there is a serious fault or the performance is significantly degraded, the warning signal is triggered and the energy storage unit is marked as a problem energy storage unit. The location information of the problem energy storage unit is obtained, and the corresponding isolation switch is controlled to implement electrical isolation and disconnect it from the system to prevent the fault from affecting other energy storage units. The isolated energy storage unit is continuously monitored, its operating data is regularly collected, and its performance status is evaluated. If the continuous monitoring results show that its performance has returned to normal, the isolation switch is controlled to release the isolation and reconnect it to the system. The diagnosis results, isolation operation records and performance recovery information of the faulty energy storage unit are reported to the monitoring center in a standardized format, and a diagnosis report is automatically generated to provide data support and decision-making basis for subsequent operation and maintenance management.

[0052] Exemplarily, real-time monitoring data of energy storage units are obtained, the data are cleaned and normalized, and key characteristic parameters such as voltage, current and temperature that reflect the performance of energy storage units are extracted. In electrochemical energy storage systems, real-time monitoring of energy storage unit data is the key to ensuring stable operation of the system. For example, the voltage, current and temperature data collected by sensors can reflect the immediate working status of the energy storage unit. These data first need to be cleaned to remove noise and outliers, and then normalized to ensure that the data are on the same scale for subsequent analysis and calculation. Using a support vector machine model for fault diagnosis is an effective method. Assume that during the training phase, the model has learned the key features that distinguish between normal and abnormal energy storage units. In actual applications, when new monitoring data is input into the model, the model will judge the state of the energy storage unit based on these features. For example, if the temperature of an energy storage unit rises abnormally, and the current and voltage show unstable fluctuations, the model may diagnose an overheating fault. Once the model diagnoses a serious fault, the system automatically triggers an early warning signal and marks the energy storage unit as a problem unit. This step is crucial because it prevents the fault from spreading to the entire energy storage system and causing greater losses. For example, if an energy storage unit has an internal short circuit, it may cause a fire if it is not isolated in time. Isolating the problem energy storage unit is achieved by controlling the electrical isolation switch. Specifically, the system controls the corresponding isolation switch based on the location information of the problem energy storage unit to cut off its electrical connection with the system. This process needs to be accurate and fast to ensure that other parts of the system are not affected. After isolation, the system will continue to monitor the status of the energy storage unit and regularly collect and analyze its operating data. If the data shows that the performance of the energy storage unit has returned to normal, for example, the voltage, current and temperature are stable within the normal range for several consecutive monitoring results, the system will control the isolation switch to release the isolation and reconnect the energy storage unit to the system. All fault diagnosis results, isolation operations and performance recovery will be recorded in detail and reported to the monitoring center in a standardized format. These records not only provide data support for operation and maintenance management, but also provide valuable information for future fault prevention and system optimization. For example, by analyzing the fault records of multiple energy storage units, the common cause of the fault can be found, and then corresponding improvement measures can be taken in system design and maintenance.

[0053] Step S105, when the electrochemical energy storage power station is in black start mode, the energy management system monitors the state of charge of each energy storage unit in real time. If the state of charge of a certain energy storage unit drops to the lower limit threshold obtained by the dynamic programming algorithm, the energy storage unit is controlled to stop discharging and its load is transferred to other energy storage units.

[0054] The real-time operation data of the electrochemical energy storage power station is obtained, including the state of charge SOC, voltage U and current I parameters of each energy storage unit, and the data is transmitted to the energy management system in real time through industrial Ethernet. The energy management system uses the optimization algorithm in dynamic programming to calculate the optimal state of charge lower limit threshold SOCmin of each energy storage unit based on historical operation data and current operation status. The optimization goal is to maximize the discharge of the energy storage unit under the premise of ensuring the safe and stable operation of the power station. The energy management system compares the actual state of charge SOC of each energy storage unit with the calculated lower limit threshold SOCmin in real time. If the SOC of a certain energy storage unit is lower than SOCmin, it is determined that the energy storage unit needs to stop discharging. For the energy storage unit that needs to stop discharging, the energy management system sends a stop discharge control instruction to it through the Modbus communication protocol, and sends the current active power, reactive power and other load data of the energy storage unit to the load management module through the OPC protocol. After receiving the load data to be transferred, the load management module uses the least squares method to predict the load, and combines the real-time SOC and capacity of other operating energy storage units to calculate the optimal power distribution plan for transferring the load to other energy storage units, with the goal of minimizing the loss of the energy storage power station. According to the calculation results of the load transfer, the load management module sends active and reactive power adjustment instructions to the target energy storage unit through Modbus, and transmits the updated SOC data back to the energy management system through OPC. The energy management system dynamically adjusts the operation strategy of the energy storage power station according to the real-time SOC and load distribution of each energy storage unit. When the power grid returns to normal, the energy management system uses a fuzzy control algorithm, combined with factors such as electricity price, SOC of the energy storage power station, and power grid frequency, to control the charging and discharging power of each energy storage unit, smooth the fluctuations of new energy power generation, and participate in the frequency regulation of the power grid to improve the safety and economy of the power grid.

[0055] For example, in an electrochemical energy storage power station, real-time monitoring of the state of charge SOC, voltage U, current I and other parameters of each energy storage unit is the key to ensure the efficient operation of the system. For example, real-time data transmission to the energy management system through industrial Ethernet can quickly respond to changes in various operating states. Assume that at a certain moment, the SOC of an energy storage unit suddenly drops to 30%, and its optimal state of charge lower limit threshold SOCmin is 40%. At this time, the energy management system will immediately recognize that the SOC of the energy storage unit is lower than the safety threshold. At this time, the energy management system sends a control instruction to stop discharging to the energy storage unit through the Modbus communication protocol to prevent further loss. At the same time, the system sends the load data such as active power and reactive power of the energy storage unit to the load management module through the OPC protocol. After receiving these data, the load management module uses the least squares method to predict the future load, which is a statistical method that finds the best function match for the data by minimizing the sum of squares of the error. Assuming that the prediction results show that the load demand of the energy storage unit will increase by 10% in the next hour, this part of the load needs to be transferred to other energy storage units. Next, the load management module calculates the optimal power allocation scheme for transferring this part of the load to other energy storage units. For example, if the current SOCs of the other two energy storage units are 50% and 60% respectively, 6% of the load can be allocated to the unit with an SOC of 50% and 4% of the load to the unit with an SOC of 60% according to their capacity and current load, so as to achieve load balance and stable operation of the system. After the power grid returns to normal, the energy management system uses a fuzzy control algorithm to adjust the charging and discharging power of each energy storage unit. Fuzzy control is a control method based on fuzzy logic, which processes uncertain or fuzzy information through fuzzy sets, fuzzy rules and fuzzy reasoning. For example, considering factors such as electricity prices, the SOC of energy storage power stations and the frequency of the power grid, the system may decide to increase the charging power of energy storage units when the electricity price is low, and increase the discharging power when the electricity price is high, so as to smooth the fluctuations of new energy generation and effectively participate in the frequency regulation of the power grid. Such a strategy not only improves the security of the power grid, but also improves the economy. Through this series of operations, the energy storage power station can achieve precise control of the state of each energy storage unit, optimize load distribution, and improve the overall efficiency and reliability of the system.

[0056] Step S106, during the black start process, the energy management system continuously evaluates the remaining available energy of the power station. If the remaining energy is predicted to be lower than the preset safety margin (e.g., the remaining energy is insufficient to support 120% of the estimated black start time), the lower limit threshold of the charge state is recalculated and appropriately increased according to the mathematical model established by the least squares method to avoid energy depletion caused by excessive discharge of the energy storage unit. At the same time, the dynamic programming algorithm is re-run to optimize the scheduling of other non-isolated energy storage units to ensure safe and stable operation of the power station.

[0057] The current remaining available energy data of the power station is obtained, and according to the preset safety margin threshold, it is determined whether the remaining energy is lower than the threshold. If the remaining energy is lower than the safety margin threshold, the least squares method is used to establish a linear regression model of the state of charge and the remaining energy based on historical data. Through the linear regression model, the new lower limit threshold of the state of charge is calculated, and the threshold is set as the lower limit of the discharge of the energy storage unit to avoid energy exhaustion caused by excessive discharge of the energy storage unit. The status information of all non-isolated energy storage units in the power station is obtained, including parameters such as state of charge, capacity, and discharge power. These parameters are used as input to construct a linear dynamic programming model, and the objective function is to maximize the scheduling benefit of the energy storage unit under the premise of ensuring the safe and stable operation of the power station. The simplex method is used to solve the linear dynamic programming model to obtain the optimal scheduling strategy of each energy storage unit, including charging and discharging power and duration. The optimal scheduling strategy is sent to the control system of each energy storage unit, and the operating status data of each unit is collected in real time through the SCADA system. According to the actual state of charge and remaining energy, the scheduling strategy is dynamically adjusted using the predictive control algorithm to ensure the safe and stable operation of the power station. When the remaining energy is lower than the lower limit of the new charge state, an alarm is triggered and the emergency dispatch plan is activated to reduce the discharge of the energy storage unit and start the backup power supply when necessary to ensure the power supply reliability of the power station.

[0058] For example, in an electrochemical energy storage power station, real-time monitoring of the remaining available energy is crucial. For example, assume that the safety margin threshold of the power station is set to 20%. When the remaining energy drops below this threshold through the SCADA system, the least squares method is immediately used to respond. The least squares method can establish a linear relationship between the state of charge and the remaining energy through historical data. Assume that historical data shows that for every 1% decrease in the state of charge, the remaining energy decreases by an average of 5%. In this way, the energy management system can predict the remaining energy based on the current state of charge and adjust the lower limit of the state of charge accordingly to prevent excessive discharge. In addition, obtaining the status information of all non-isolated energy storage units, such as state of charge, capacity, and discharge power, is the basis for formulating a scheduling strategy. Based on this information, a linear dynamic programming model is constructed with the goal of maximizing the scheduling benefit while ensuring the safety of power station operation. For example, if an energy storage unit has a large capacity but a low state of charge, it may be decided to reduce its discharge power and extend the discharge time to protect the health of the battery and extend its service life. The simplex method is used to solve this dynamic programming model to obtain the optimal charging and discharging strategy for each energy storage unit. The simplex method is an effective optimization algorithm that is suitable for dealing with linear problems and iteratively finds the optimal solution. For example, for a specific energy storage unit, if the model calculation results show that increasing charging can obtain higher economic benefits under the current electricity price and grid demand, the system will adjust the charging power of the unit. These optimal strategies are sent to each energy storage unit through the control system, and the actual operating status continues to be monitored through the SCADA system. This real-time data feedback allows the use of predictive control algorithms to dynamically adjust the strategy to respond to emergencies or prediction errors. For example, if the actual charging speed of a certain energy storage unit is slower than expected, the system may immediately reduce its discharge power to avoid energy exhaustion. Finally, when the remaining energy is detected to be lower than the newly set lower limit of the charge state, the system will trigger an alarm and initiate an emergency plan. This may include reducing the discharge power of the entire power station, or starting the backup power supply when necessary to ensure the power supply reliability of the power station. This emergency response strategy is an important part of power station management, ensuring that the power station can still operate safely and stably under adverse conditions.

[0059] Step S107: After the black start is completed, the state of charge and health status data of each energy storage unit (including parameters such as internal resistance and capacity decay rate) are obtained, and the impact of the black start process on the performance of the energy storage unit is analyzed through the decision tree algorithm, and a comprehensive evaluation is performed in combination with the fault diagnosis results. According to the analysis results, the energy management strategy and the lower limit threshold setting of the state of charge are optimized, and the mathematical model established by the least squares method and the parameters of the dynamic programming algorithm are updated to prepare for the subsequent black start.

[0060] First, according to the actual operation data of the energy storage unit, the least squares method is used to identify the parameters of the mathematical model to obtain the initial model parameters. Then, during the operation of the energy storage unit, the input and output data are obtained in real time, and the model parameters are iteratively updated using the Kalman filter algorithm to make the model output as close to the actual output as possible. The continuously updated mathematical model provides support for the subsequent optimization of energy management strategies. Based on the updated mathematical model, the dynamic programming algorithm is used to optimize the operating conditions of the energy storage unit. The optimization goal is to minimize the capacity attenuation and charge and discharge times of the energy storage unit while meeting the black start power requirements. The following constraints need to be considered during the optimization process: the state of charge of the energy storage unit must be maintained within a reasonable range; the charge and discharge power cannot exceed the rated power of the energy storage unit; the single charge and discharge time cannot be too long to avoid accelerating the aging of the energy storage unit. By solving the above optimization problems with the dynamic programming algorithm, the optimal operation strategy of the energy storage unit under different working conditions can be obtained. The optimized energy storage operation strategy is sent to the control system of the energy storage unit to schedule and manage the charging and discharging process of the energy storage unit in real time, ensuring that the service life of the energy storage unit is maximized while meeting the black start requirements, providing reliable energy guarantee for the safe and stable operation of the power system.

[0061] For example, in the power system, the effective management of the energy storage unit is crucial to ensure the stable operation of the system. First, the initial model parameters of the energy storage unit are identified by the least squares method. This step can be achieved by collecting the input and output data of the energy storage unit at different states of charge. For example, if an energy storage unit is at a state of charge of 50%, the input is 100 kWh of electricity, and the output is 95 kWh. This data can help establish a preliminary energy loss model. Subsequently, in order to improve the accuracy and adaptability of the model, the Kalman filter algorithm is used to update the model parameters in real time. Kalman filtering is an excellent prediction correction method. It can continuously receive new input and output data during the operation of the energy storage unit, and correct the model parameters through algorithm iteration to make the output predicted by the model as close as possible to the actual output. For example, if it is found in actual operation that the actual output of a certain energy storage unit is always lower than the predicted output, the Kalman filter can adjust the model parameters to reduce this deviation. Using the updated mathematical model, the operation strategy of the energy storage unit can be optimized through a dynamic programming algorithm. The dynamic programming algorithm can find an operation strategy that minimizes the capacity decay and charge and discharge times of the energy storage unit, taking into account factors such as the state of charge, charge and discharge power, and time constraints of the energy storage unit over multiple operation cycles. For example, for a 1-MWh energy storage unit, if the grid load is predicted to be low in the next few hours, the dynamic programming algorithm may recommend reducing the discharge rate and extending the discharge time to reduce the number of battery cycles and extend its service life. Finally, the optimized operation strategy is sent to the energy storage unit through the control system. During this process, the control system monitors the charge and discharge status of the energy storage unit in real time and adjusts the charge and discharge power according to the optimization strategy. This method not only ensures that the power system can meet key requirements such as black start, but also maximizes the service life of the energy storage unit, providing strong support for the safe and stable operation of the power system.

[0062] Obtain real-time load demand data and ambient temperature data, and use a dynamic programming algorithm to solve the optimal lower limit threshold of the charge state based on a mathematical model, which also includes the following steps:

[0063] Acquire real-time monitoring data of energy storage units, clean and normalize the data, extract key characteristic parameters reflecting the performance of energy storage units, and use pre-trained support vector machine models for fault diagnosis;

[0064] When the electrochemical energy storage power station is in black start mode, the state of charge of each energy storage unit is monitored in real time. If the state of charge of a certain energy storage unit drops to the lower limit threshold solved by the dynamic programming algorithm, the energy storage unit is controlled to stop discharging and its load is transferred to other energy storage units.

[0065] During the black start process, the remaining available energy of the power station is continuously evaluated. If the remaining available energy is predicted to be lower than the preset safety margin, the lower limit threshold of the state of charge is recalculated and appropriately increased according to the mathematical model to avoid energy depletion caused by excessive discharge of the energy storage unit. At the same time, the dynamic programming algorithm is re-run to optimize the scheduling of other non-isolated energy storage units to ensure safe and stable operation of the power station.

[0066] After the black start is completed, the charge state and health status data of each energy storage unit are obtained, and the impact of the black start process on the performance of the energy storage unit is analyzed through the decision tree algorithm, and a comprehensive evaluation is performed in combination with the fault diagnosis results. Based on the analysis results, the energy management strategy and the charge state lower limit threshold setting are optimized, and the mathematical model established by the least squares method and the parameters of the dynamic programming algorithm are updated to prepare for subsequent black starts.

[0067] Based on the same inventive concept, the second aspect of the embodiment of the present invention provides a safety dispatching and analysis system for black start of electrochemical energy storage power station units, including a data acquisition module, a data preprocessing module, a digital modeling module, a prediction and analysis module, an optimization decision module, and a fault diagnosis and isolation module;

[0068] A data acquisition module is used to obtain real-time state of charge data of each energy storage unit in the electrochemical energy storage power station;

[0069] A data preprocessing module is used to remove noise and smooth the state of charge data using a Kalman filter algorithm to obtain a state of charge estimation value;

[0070] Digital modeling module, used to establish a mathematical model between the state of charge of the energy storage unit and the discharge power, energy conversion efficiency, and ambient temperature through least squares fitting based on historical operating data;

[0071] A prediction and analysis module, used to predict the available energy of the energy storage unit under different state of charge lower limit threshold settings using the state of charge estimation value and the mathematical model;

[0072] The optimization decision module is used to obtain real-time load demand data and ambient temperature data, and uses a dynamic programming algorithm to solve the optimal lower limit threshold of the charge state based on a mathematical model.

[0073] Furthermore, the data acquisition module is used to obtain the real-time state of charge data of each energy storage unit in the electrochemical energy storage power station, specifically:

[0074] A state of charge sensor is installed on each energy storage unit to collect state of charge data in real time, and the data is transmitted to the data cache area of ​​the data processing center through the data acquisition module.

[0075] Based on the same inventive concept, the fourth aspect of an embodiment of the present invention provides a computer device, including: a memory and a processor; the memory stores a computer program, characterized in that: when the processor executes the computer program, the steps of the above-mentioned safe scheduling analysis method for black start of the electrochemical energy storage power station unit are implemented.

[0076] Based on the same inventive concept, the fifth aspect of an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-mentioned safe scheduling analysis method for black start of an electrochemical energy storage power station unit.

[0077] In summary, the present invention discloses a black start energy management method for an electrochemical energy storage power station. The method collects the state of charge data of the energy storage unit in real time, and uses the Kalman filter algorithm to remove noise to obtain an accurate estimate; combines historical operation data to establish a mathematical model of parameters such as the state of charge and discharge power, and predicts the available energy under different thresholds. Considering the load demand and ambient temperature, a dynamic programming algorithm is used to solve the optimal lower limit threshold of the state of charge, maximizing the utilization efficiency while ensuring a reliable black start. The present invention also uses a support vector machine algorithm to perform fault diagnosis and early warning, monitor the status of each unit in real time, and isolate the problem unit in a timely manner. During the black start process, the remaining energy is continuously evaluated and the threshold and scheduling strategy are dynamically adjusted to ensure safe and stable operation. After the black start, the impact of the process on the energy storage unit is analyzed, and the management strategy and parameter settings are optimized to prepare for subsequent black starts. The present invention significantly improves the black start reliability and energy utilization efficiency of the electrochemical energy storage power station.

[0078] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A safety dispatch analysis method for black start of electrochemical energy storage power station units, characterized in that: The following steps are involved: Obtain real-time state of charge data of each energy storage unit in the electrochemical energy storage power station; The Kalman filter algorithm is used to remove noise and smooth the state of charge data to obtain the estimated value of the state of charge; Based on historical operation data, a mathematical model between the state of charge of the energy storage unit and the discharge power, energy conversion efficiency, and ambient temperature is established through least squares fitting; The available energy of the energy storage unit under different state of charge lower limit threshold settings is predicted using the state of charge estimation value and mathematical model; Real-time load demand data and ambient temperature data are obtained, and the optimal lower limit threshold of the charge state is solved based on the mathematical model using a dynamic programming algorithm.

2. The method according to claim 1, characterized in that The Kalman filter algorithm is used to remove noise and smooth the state of charge data, specifically: Perform format conversion and outlier filtering on the state of charge data, establish a state space model, and dynamically adjust the process noise covariance matrix Q and the measurement noise covariance matrix R; In the filtering process, by comparing the residuals of the filter estimate and the measured value, the Q and R matrices are dynamically adjusted to minimize the mean square error of the residuals; The state of charge estimation value after filtering is smoothed, and the exponential smoothing method is used to set the smoothing factor α to perform smoothing calculation on the estimation value sequence.

3. The method according to claim 1, characterized in that The mathematical model establishment process is as follows: With the state of charge of the energy storage unit as the independent variable, and the discharge power, energy conversion efficiency and ambient temperature as the dependent variables, mathematical fitting is performed, and the accuracy and generalization ability of the model are evaluated by the cross-validation method, and the model is optimized and adjusted according to the evaluation results.

4. The method according to claim 3, characterized in that The method of using the estimated state of charge value and the mathematical model to predict the available energy of the energy storage unit under different state of charge lower limit threshold settings is specifically as follows: The estimated state of charge of the current energy storage unit and the ambient temperature are substituted into the mathematical model as input variables to calculate the predicted value of the available energy of the energy storage unit under different threshold settings.

5. The method according to any one of claims 1 to 4, characterized in that: The real-time load demand data and ambient temperature data are obtained, and a dynamic programming algorithm is used to solve the optimal lower limit threshold of the charge state based on a mathematical model, specifically: The average black start duration is obtained using the statistical regression method, and the state equation and objective function are constructed to calculate the black start reliability of the power station and the utilization efficiency of the energy storage unit corresponding to the lower limit thresholds of different charge states.

6. The method according to any one of claims 1 to 4, characterized in that: The method of obtaining real-time load demand data and ambient temperature data and solving the optimal charge state lower limit threshold based on a mathematical model using a dynamic programming algorithm also includes the following steps: Acquire real-time monitoring data of energy storage units, clean and normalize the data, extract key characteristic parameters reflecting the performance of energy storage units, and use pre-trained support vector machine models for fault diagnosis; When the electrochemical energy storage power station is in black start mode, the state of charge of each energy storage unit is monitored in real time. If the state of charge of a certain energy storage unit drops to the lower limit threshold solved by the dynamic programming algorithm, the energy storage unit is controlled to stop discharging and its load is transferred to other energy storage units. During the black start process, the remaining available energy of the power station is continuously evaluated. If the remaining available energy is predicted to be lower than the preset safety margin, the lower limit threshold of the state of charge is recalculated and appropriately increased according to the mathematical model to avoid energy depletion caused by excessive discharge of the energy storage unit; At the same time, the dynamic programming algorithm is re-run to optimize the scheduling of other non-isolated energy storage units to ensure safe and stable operation of the power station; After the black start is completed, the state of charge and health status data of each energy storage unit are obtained, and the impact of the black start process on the performance of the energy storage unit is analyzed through the decision tree algorithm, and a comprehensive evaluation is performed in combination with the fault diagnosis results; According to the analysis results, the energy management strategy and the lower limit threshold setting of the charge state are optimized, and the parameters of the mathematical model established by the least squares method and the dynamic programming algorithm are updated to prepare for the subsequent black start.

7. A safety dispatching and analysis system for black start of electrochemical energy storage power station units, characterized in that: It includes data acquisition module, data preprocessing module, digital modeling module, prediction analysis module, optimization decision module and fault diagnosis and isolation module; A data acquisition module is used to obtain real-time state of charge data of each energy storage unit in the electrochemical energy storage power station; A data preprocessing module is used to remove noise and smooth the state of charge data using a Kalman filter algorithm to obtain a state of charge estimation value; Digital modeling module, used to establish a mathematical model between the state of charge of the energy storage unit and the discharge power, energy conversion efficiency, and ambient temperature through least squares fitting based on historical operating data; A prediction and analysis module, used to predict the available energy of the energy storage unit under different state of charge lower limit threshold settings using the state of charge estimation value and the mathematical model; The optimization decision module is used to obtain real-time load demand data and ambient temperature data, and uses a dynamic programming algorithm to solve the optimal lower limit threshold of the charge state based on a mathematical model.

8. The system according to claim 7, characterized in that The data acquisition module is used to obtain the real-time state of charge data of each energy storage unit in the electrochemical energy storage power station, specifically: A state of charge sensor is installed on each energy storage unit to collect state of charge data in real time, and the data is transmitted to the data cache area of ​​the data processing center through the data acquisition module.

9. A computer device comprising: Memory and processor; The memory stores a computer program, characterized in that: when the processor executes the computer program, the steps of the safety scheduling analysis method for black start of an electrochemical energy storage power station unit according to any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the safe scheduling analysis method for black start of an electrochemical energy storage power station unit according to any one of claims 1 to 6 are implemented.

Images