Distributed energy storage power supply scheduling control method, system and equipment based on power grid software supply chain and medium

Abstract

The invention discloses a power grid software supply chain-based distributed energy storage power supply scheduling control method, system and device and a medium, and belongs to the technical field of power system scheduling control, and the method comprises the steps: collecting the operation parameters of a distributed energy storage unit and the supply chain information of power grid software, and carrying out the information fusion to generate credible operation state data; dividing the distributed energy storage unit into a standby pool and an isolation pool, and generating a charging and discharging control curve of the distributed energy storage unit in the standby pool; and inputting the charging and discharging control curve of the distributed energy storage unit in the standby pool into the electric energy scheduling optimization model, executing optimization solution, and outputting a scheduling optimization result. According to the method, power grid software supply chain information and operation parameters are subjected to double verification and fusion quantification, accurate evaluation of the safety state of the energy storage unit and fault probability prediction are achieved, a charging and discharging control curve with safety and stability considered is generated, and the power stability, the standby capacity utilization rate and the operation economical efficiency are improved under multi-target optimization.

Description

Technical Field

[0001] This invention relates to the field of power system dispatch and control technology, specifically to a method, system, equipment, and medium for dispatch and control of distributed energy storage power sources based on the power grid software supply chain. Background Technology

[0002] With the integration of numerous distributed energy storage units into distribution networks and microgrid systems, they can play a crucial role in peak shaving and valley filling, emergency power supply, and improving the integration rate of new energy sources through flexible charging and discharging operation modes. To achieve coordinated operation of multiple energy storage units, it is typically necessary to establish a unified dispatch and control strategy in the power grid dispatch center or regional energy management system to comprehensively manage the operating parameters of each energy storage unit, such as charging and discharging power, state of charge, and reserve capacity.

[0003] Existing distributed energy storage dispatch and control methods are mostly based on grid power flow analysis and operating parameter prediction, combined with conditions such as state of charge constraints and capacity constraints, to construct optimization models to obtain power allocation schemes for each time period. These methods have achieved certain results in improving energy storage utilization efficiency and maintaining grid operation stability. However, the operation and control of energy storage units usually rely on hardware and software devices such as inverter controllers and monitoring terminals, and their firmware versions, control strategies, and software components all come from the grid software supply chain.

[0004] In actual operation, the software supply chain of energy storage control systems may pose security risks due to firmware upgrades, policy adjustments, or vulnerabilities in third-party components. Existing scheduling and control methods often focus on optimizing the physical operating state and lack mechanisms to incorporate the reliability of the software supply chain into scheduling decisions. Once the control software of an energy storage unit has vulnerabilities or abnormal configurations, even if the operating parameters are within the normal range, it may lead to scheduling execution deviations, power fluctuations, or even equipment failure.

[0005] Furthermore, in scenarios where multiple energy storage units participate in grid operation, existing solutions typically conduct operational parameter monitoring and software security testing independently, lacking the ability to integrate these two types of information to construct a unified operational status evaluation. This makes it difficult to accurately reflect the actual availability and safety level of energy storage units during dispatch optimization. This, to some extent, limits the reliable dispatch and safe operation of distributed energy storage in the grid. Summary of the Invention

[0006] In view of the above-mentioned problems, the present invention is proposed.

[0007] Therefore, the technical problem solved by this invention is that in existing distributed energy storage power dispatch and control technologies, operating parameters and grid software supply chain information are often processed independently, lacking a correlation modeling and reliability comprehensive evaluation mechanism based on a unified time base and data structure, which makes it impossible to fully reflect the real safety status of energy storage units in the integrated hardware and software operating environment; at the same time, in existing dispatch optimization models, power flow constraints, state of charge constraints and reserve capacity constraints are usually modeled separately and lack linkage processing, making it difficult to ensure system safety while taking into account dispatch accuracy and resource utilization efficiency.

[0008] To address the aforementioned technical problems, this invention provides the following technical solution: a distributed energy storage power dispatch and control method based on the power grid software supply chain, comprising,

[0009] The system collects the operating parameters of the distributed energy storage unit and the supply chain information of the power grid software, performs matching verification between the operating parameters and the control strategy of the distributed energy storage unit, performs software bill of materials hash verification and controller remote measurement report verification of the supply chain information, and merges the verified operating parameters and verified supply chain information to generate reliable operating status data. The initial safety level of the distributed energy storage unit is calculated using trusted operating status data, and the failure probability of the distributed energy storage unit is predicted based on the historical trend of operating parameters in the trusted operating status data. The initial safety level and failure probability are used as inputs to divide the distributed energy storage unit into a backup pool and an isolation pool, and the charge and discharge control curves of the distributed energy storage unit in the backup pool are generated. The charging and discharging control curves of the distributed energy storage units in the standby pool are input into the power dispatch optimization model, which includes power flow constraints, state of charge constraints, and standby capacity constraints. The optimization solution is then executed, and the dispatch optimization results are output.

[0010] As a preferred embodiment of the distributed energy storage power dispatch and control method based on the power grid software supply chain described in this invention, the matching verification of the execution operating parameters and the distributed energy storage unit control strategy includes: using the name of the operating parameter as an index, searching for the allowable threshold range of the same name item in the preset firmware version parameter table and control strategy parameter table; comparing the measured value of each operating parameter with the searched allowable threshold range; and determining that the operating parameter verification is qualified when all measured values ​​fall within their respective searched allowable threshold ranges. The software bill of materials hash verification and controller remote measurement report verification for executing supply chain information include: comparing the software bill of materials hash value with the controller remote measurement report, calculating the hash value of the software bill of materials and comparing it with a preset trusted hash value; if they match, it is recorded as passed; comparing the measured value of each detection item in the controller remote measurement report with the baseline value of the same detection item in the preset trusted baseline data; if the absolute value of the difference between the two does not exceed a preset deviation limit, it is recorded as passed; when both the software bill of materials hash value comparison and the measurement report comparison are recorded as passed, the supply chain information verification is deemed qualified.

[0011] As a preferred embodiment of the distributed energy storage power dispatch and control method based on power grid software supply chain described in this invention, the step of fusing verified operating parameters and verified supply chain information to generate reliable operating status data includes: according to the equipment identification code of the distributed energy storage unit, searching for operating parameter records and supply chain information records with the same equipment identification code in the operating parameters and supply chain information respectively, establishing an index association relationship, and forming an operating parameter index table and a supply chain information index table. Based on a unified time base, time alignment processing is performed on the records in the operation parameter index table and the supply chain information index table to generate a time-aligned set of operation parameters and a time-aligned set of supply chain information. Map the time-aligned set of operating parameters to operating parameter trust vectors according to parameter type, and map the time-aligned set of supply chain information to supply chain trust vectors according to information fields; An element-wise weighted operation is performed on the trusted vector of operating parameters and the trusted vector of the supply chain to obtain a fused trusted vector. The fused trusted vector is then bound to the device identification code according to the device identification code to generate trusted operating status data containing the device identification code and the fused trusted vector.

[0012] As a preferred embodiment of the distributed energy storage power dispatch and control method based on power grid software supply chain described in this invention, the step of dividing the distributed energy storage unit into a backup pool and an isolation pool includes continuously sampling the fused reliable vector in the reliable operating status data at equal time intervals according to the equipment identification code, and generating a fused reliable vector time series matrix with time series as rows and vector components as columns.

[0013] The running parameter sub-vector and the supply chain information sub-vector are extracted from the time series matrix according to the preset parameter field index. The sliding window covariance sequence, the sliding window correlation coefficient sequence and the power spectrum distribution of the discrete Fourier transform are calculated in the time series dimension and combined to form a set of running and supply chain coupling features. The set of coupling features is input into a security level determination model obtained through supervised learning using historical operational failure records and software vulnerability event records as training samples, and the initial security level of the distributed energy storage unit is output. Based on the aforementioned set of coupling features, within each time window, the first-order difference of the operational stability index sequence along the time axis is calculated to obtain the operational stability gradient vector, and the first-order difference of the vulnerability risk index sequence along the time axis is calculated to obtain the vulnerability risk gradient vector; the arithmetic mean of the absolute values ​​of each component of the operational stability gradient vector is calculated to obtain the operational stability change rate; the arithmetic mean of the absolute values ​​of each component of the vulnerability risk gradient vector is calculated to obtain the vulnerability risk accumulation rate. The failure probability is obtained by retrieving a preset failure probability mapping table based on the combined interval of the rate of change in operational stability and the rate of accumulation of vulnerability risk. The initial safety level and failure probability are mapped to a combination of two-dimensional coordinates. The boundary between the reserve area and the isolation area in the threshold matrix is ​​used for area matching. If the fault falls into the reserve area, it is assigned to the reserve pool; if it falls into the isolation area, it is assigned to the isolation pool.

[0014] As a preferred embodiment of the distributed energy storage power dispatch and control method based on the power grid software supply chain described in this invention, the generation of the charge and discharge control curves of the distributed energy storage units in the backup pool includes, according to the equipment identification code, collecting the rate of change of the state of charge, the maximum allowable charge and discharge power, the rate of temperature rise and the rate of change of internal resistance of each distributed energy storage unit in the backup pool, forming a parameter vector sequence arranged in time slices. Using the initial safety level as an index, select a derating factor that matches the initial safety level from the preset derating factor table, and multiply the maximum allowable charge and discharge power component in the parameter vector sequence with the derating factor to generate a safe power upper limit sequence. Under the constraint of the safe power upper limit sequence, the charge state change rate component is accumulated with the power output value of the previous time slice to determine the target charge and discharge power value of each time slice. The temperature rise rate component is compared with the preset temperature threshold to calculate the maximum continuous running time of each time slice, and a time-segmented power table containing power value and running time is obtained. Using the internal resistance change rate component as a smoothing weight, weighted interpolation is performed on adjacent power values ​​of the time-segmented power meter, and amplitude limiting is performed when the interpolation result exceeds the safe power limit of the time slice, to obtain a smoothed and corrected continuous power sequence. The continuous power sequence is mapped to the scheduling and control time axis according to a unified time base to form the charging and discharging control curve of the distributed energy storage unit.

[0015] As a preferred embodiment of the distributed energy storage power dispatch and control method based on the power grid software supply chain described in this invention, the output dispatch optimization result includes: indexing and integrating the charge and discharge control curves, state of charge sequence, available capacity sequence and power change rate sequence of the distributed energy storage units in the standby pool according to the equipment identification code to form an input parameter vector sequence arranged in time slices; Based on the node voltage, branch power and phase angle difference components in the input parameter vector sequence, interval pruning, boundary limiting and rate of change limiting are performed according to the power flow constraint parameter set to obtain the first modified sequence that satisfies the power flow constraint. Based on the state of charge and the rate of change of the state of charge in the input parameter vector sequence, interval pruning, boundary limiting and rate of change limiting are performed according to the state of charge constraint parameter set to obtain a second modified sequence that satisfies the state of charge constraint. Based on the available capacity components in the input parameter vector sequence, interval pruning, boundary limiting, and rate of change limiting are performed according to the spare capacity constraint parameter set to obtain a third modified sequence that satisfies the spare capacity constraint.

[0016] The first, second, and third correction sequences are merged element-wise under a unified time base to generate a constraint-corrected scheduling input sequence. The constrained and modified scheduling input sequence is input into the power dispatch optimization model. The objective function is to minimize system power fluctuation, maximize reserve capacity utilization, and minimize charging and discharging losses. The optimization solution is executed, and the dispatch optimization results are output.

[0017] As a preferred embodiment of the distributed energy storage power dispatch and control method based on the power grid software supply chain described in this invention, the output dispatch optimization result further includes parsing the optimal active power sequence, optimal reactive power sequence, available capacity utilization rate sequence, and time slice duration sequence from the output result of the power dispatch optimization model according to the equipment identification code. Based on the available capacity utilization rate sequence, the rated capacity is proportionally allocated to generate a reserve capacity ratio sequence; Calculate the power rise slope sequence and power fall slope sequence based on the optimal active power sequence and the duration sequence of the time slice; convert the optimal active power sequence of the time slice into a DC side current reference sequence based on the rated DC voltage and converter efficiency in the preset inverter nameplate parameter table; use the optimal reactive power sequence of the time slice as the reactive power reference sequence, and combine them to form an inverter control parameter set that includes the power rise slope sequence, the power fall slope sequence, the DC side current reference sequence, and the reactive power reference sequence; The standby capacity ratio sequence and the inverter control parameter set are encapsulated into a time-slice scheduling instruction data frame, which is then sent to the distributed energy storage unit controller in chronological order and according to the device identification code. The distributed energy storage unit controller performs active power tracking control and slope limiting control in each time slice, performs reactive power tracking control, sets the upper limit of available power and the state of charge window according to the reserve capacity ratio sequence, and completes the scheduling operation.

[0018] This invention provides a distributed energy storage power dispatch and control system based on the power grid software supply chain.

[0019] To solve the above technical problems, the present invention provides the following technical solution: a distributed energy storage power dispatch and control system based on power grid software supply chain, comprising: a data acquisition module, used to acquire the operating parameters of the distributed energy storage unit and the supply chain information of the power grid software, perform matching verification between the operating parameters and the control strategy of the distributed energy storage unit, perform software bill of materials hash verification and controller remote measurement report verification of the supply chain information, and fuse the verified operating parameters and the verified supply chain information to generate reliable operating status data; The analysis module is used to calculate the initial safety level of the distributed energy storage unit using trusted operating status data, and predict the failure probability of the distributed energy storage unit based on the historical trend of operating parameters in the trusted operating status data. Taking the initial safety level and failure probability as input, the distributed energy storage unit is divided into a backup pool and an isolation pool, and the charging and discharging control curves of the distributed energy storage unit in the backup pool are generated. The scheduling module is used to input the charging and discharging control curves of the distributed energy storage units in the standby pool into the power scheduling optimization model, which includes power flow constraints, state of charge constraints, and standby capacity constraints, and output the scheduling optimization results.

[0020] The present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, characterized in that the processor executes the computer program to implement the steps of the distributed energy storage power dispatch and control method based on the power grid software supply chain.

[0021] 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, it implements the steps of the distributed energy storage power dispatch and control method based on the power grid software supply chain.

[0022] The beneficial effects of this invention are as follows: By introducing a dual verification mechanism of grid software supply chain information and operating parameters into the distributed energy storage scheduling process, it is possible to effectively identify and exclude energy storage units with firmware vulnerabilities, abnormal control strategies, or abnormal operating states, thereby reducing the grid operation risks caused by software security vulnerabilities or hardware failures.

[0023] The fused and generated reliable operating status data enables unified quantification of operating parameters and supply chain security status, providing a complete and consistent data foundation for security level assessment and failure probability prediction, and improving the accuracy of energy storage unit security status determination.

[0024] The charge and discharge control curves generated based on safety level and operating status characteristics can achieve a smooth transition of power output while ensuring the safety of equipment operation, and reduce power surges and power quality fluctuations caused by them during the charge and discharge process.

[0025] During the scheduling optimization process, power flow constraints, state of charge constraints, and reserve capacity constraints are uniformly incorporated into the multi-objective optimization model. This ensures that the optimization results meet the requirements for safe operation while minimizing system power fluctuations, maximizing reserve capacity utilization, and minimizing charging and discharging losses, thereby improving the overall stability of power grid scheduling.

[0026] The full-process scheduling and control mechanism constructed by this method realizes closed-loop control of energy storage units from status acquisition, reliability verification, operation characteristic analysis to optimized scheduling execution, which significantly enhances the availability, controllability and predictability of distributed energy storage resources in complex power grid operation environments. Attached Figure Description

[0027] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 The above is a flowchart of a distributed energy storage power dispatch and control method based on a power grid software supply chain, provided as an embodiment of the present invention. Detailed Implementation

[0029] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0030] Example 1, referring to Figure 1 This is one embodiment of the present invention, which provides a distributed energy storage power dispatch and control method based on the power grid software supply chain, including: Step 1: Collect the operating parameters of the distributed energy storage unit and the supply chain information of the grid software, perform matching verification between the operating parameters and the control strategy of the distributed energy storage unit, perform software bill of materials hash verification and controller remote measurement report verification of the supply chain information, and merge the verified operating parameters and verified supply chain information to generate reliable operating status data. In step 1, the operating parameters include voltage, frequency, state of charge, power change rate, temperature, cumulative charge-discharge cycle count, and internal resistance change trend. Supply chain information includes the version identifier of the power grid software, firmware version number, control strategy identifier, software bill of materials hash value, vulnerability severity level, vulnerability remediation status, and baseline data of remote measurement reports for controllers; Using the name of the running parameter as an index, the allowable threshold range of the same item is searched in the preset firmware version parameter table and control strategy parameter table. The measured value of each running parameter is compared with the searched allowable threshold range. When all measured values ​​fall within their respective searched allowable threshold ranges, the running parameter is deemed to be qualified. The software bill of materials hash value is compared with the remote measurement report of the controller to verify the supply chain information. The hash value of the software bill of materials is calculated and compared with the preset trusted hash value. If they match, it is considered as passing. The measured value of each detection item in the remote measurement report of the controller is compared with the baseline value of the same detection item in the preset trusted baseline data. If the absolute value of the difference between the two does not exceed the preset deviation limit, it is considered as passing. When both the software bill of materials hash value comparison and the measurement report comparison are considered as passing, the supply chain information verification is deemed qualified.

[0031] Based on the equipment identification code of the distributed energy storage unit, the operating parameter records and supply chain information records with the same equipment identification code are retrieved from the operating parameters and supply chain information, respectively, and an index relationship is established to form an operating parameter index table and a supply chain information index table. Based on a unified time base, time alignment processing is performed on the records in the operation parameter index table and the supply chain information index table to generate a time-aligned set of operation parameters and a time-aligned set of supply chain information. Map the time-aligned set of operating parameters to operating parameter trust vectors according to parameter type, and map the time-aligned set of supply chain information to supply chain trust vectors according to information fields; An element-wise weighted operation is performed on the trusted vector of operating parameters and the trusted vector of the supply chain to obtain a fused trusted vector. The fused trusted vector is then bound to the device identification code according to the device identification code to generate trusted operating status data containing the device identification code and the fused trusted vector.

[0032] In this embodiment, the method for determining the threshold range of operating parameters adopts a dynamic adjustment approach based on historical equipment data. First, the initial standard operating ranges of various operating parameters, including voltage, frequency, state of charge, and power change rate, are obtained from the equipment's technical manual as reference thresholds. Then, during equipment operation, the measured values ​​of each operating parameter are collected and recorded in real time. This data is transmitted and stored through a monitoring system. The collected data undergoes preprocessing to remove outliers and noise, ensuring data accuracy. After data preprocessing, the dynamic threshold adjustment stage begins. The specific steps are as follows: First, a time window is set, for example, data from the past 30 days, and the equipment operating data within this time window is statistically analyzed. The maximum, minimum, mean, and standard deviation of each operating parameter are calculated, and a new threshold range is set. If parameters such as voltage or frequency fluctuate and exceed the standard range within a certain time period, the upper or lower threshold is adjusted by increasing or decreasing it to ensure stable operation of the equipment under different loads and environmental conditions. The dynamically adjusted thresholds allow the equipment to adapt to the current operating state, thereby improving overall safety and stability.

[0033] During the verification of supply chain information, the software bill of materials (BOM) and controller remote measurement reports are first hashed to ensure their content has not been tampered with. Specifically, each item in the software BOM is hashed using the SHA-256 algorithm, and this hash is compared with a pre-stored trusted hash value in the database. If the calculated hash value matches the stored hash value, the BOM is considered valid and tamper-proof. Similarly, for the controller remote measurement reports, the hash value of each data item in the report is calculated and compared with a preset trusted hash value. If the comparison results match, the report data is deemed valid. In practice, for each test data item in the report, the measured value is compared with a preset baseline value. If the difference is within the allowable deviation range, the data is considered to meet the expected requirements. Only when the hash values ​​of all BOMs and controller reports match the test data are the supply chain information verification considered successful.

[0034] In generating trusted operational status data, the operating parameters are first associated with supply chain information based on the device identification code. The device identification code maps the operating parameter records to the supply chain information records, ensuring that the operating data of each device is synchronized with its associated supply chain information. Next, based on a unified time base, time alignment is performed, arranging the device's operating parameter records and supply chain information records on the same timeline to ensure data at the same point in time matches. After time alignment, operating parameters are mapped by type to trusted operating parameter vectors, and supply chain information is mapped by field to trusted supply chain vectors. Then, element-wise weighted calculations are performed, assigning different weights to important operating parameters (such as voltage and frequency) and auxiliary parameters (such as temperature and internal resistance) to obtain a fused trusted vector. The purpose of the weighted calculation is to reasonably weight each parameter according to its impact on device operation, so that the final fused vector accurately reflects the overall operating status of the device. Finally, the fused trusted vector is bound to the device identification code to generate trusted operational status data containing both the device identification code and the fused trusted vector.

[0035] Step 2: Calculate the initial safety level of the distributed energy storage unit using trusted operating status data, and predict the failure probability of the distributed energy storage unit based on the historical trend of operating parameters in the trusted operating status data. Using the initial safety level and failure probability as input, divide the distributed energy storage unit into a backup pool and an isolation pool, and generate the charging and discharging control curves of the distributed energy storage unit in the backup pool. In step 2, dividing the distributed energy storage unit into a backup pool and an isolation pool includes continuously sampling the fused trusted vector in the trusted operating status data at equal time intervals according to the device identification code, and generating a fused trusted vector time series matrix with time series as rows and vector components as columns; The running parameter sub-vector and the supply chain information sub-vector are extracted from the time series matrix according to the preset parameter field index. The sliding window covariance sequence, the sliding window correlation coefficient sequence and the power spectrum distribution of the discrete Fourier transform are calculated in the time series dimension and combined to form a set of running and supply chain coupling features. The set of coupling features is input into a security level determination model obtained through supervised learning using historical operational failure records and software vulnerability event records as training samples, and the initial security level of the distributed energy storage unit is output. Based on the aforementioned set of coupling features, within each time window, the first-order difference of the operational stability index sequence along the time axis is calculated to obtain the operational stability gradient vector, and the first-order difference of the vulnerability risk index sequence along the time axis is calculated to obtain the vulnerability risk gradient vector; the arithmetic mean of the absolute values ​​of each component of the operational stability gradient vector is calculated to obtain the operational stability change rate; the arithmetic mean of the absolute values ​​of each component of the vulnerability risk gradient vector is calculated to obtain the vulnerability risk accumulation rate. The failure probability is obtained by retrieving a preset failure probability mapping table based on the combined interval of the rate of change in operational stability and the rate of accumulation of vulnerability risk. The initial safety level and failure probability are mapped to a combination of two-dimensional coordinates. The boundary between the reserve area and the isolation area in the threshold matrix is ​​used for area matching. If the fault falls into the reserve area, it is assigned to the reserve pool; if it falls into the isolation area, it is assigned to the isolation pool.

[0036] Specifically, based on the device identification code (ID), the fused trusted vector in the trusted operating status data is continuously sampled at equal time intervals to generate a fused trusted vector time series matrix with time series as rows and vector components as columns.

[0037] Let the device identification code be ID, and the fused trust vector for each time slice be... Then the time series matrix Represented as: , in, To fuse the trusted vector time series matrix, we have a matrix representing all fused trusted vector data from the time series. Rows in the matrix represent time slices, and columns represent the feature data for each time slice. The number of rows in the matrix is... The number of time slices represents the number of columns. The feature dimension representing each time slice, For the first Okay, number The element of the column represents the first element. The first time film One eigenvalue; This is the index for the time slice, indicating the data time point corresponding to that row; This is an index for the feature dimension, representing the data feature corresponding to that column. It is the number of feature dimensions; The number of time slices indicates the total number of time slices of data. The feature dimension for each time slice represents the number of different types of features contained in each time slice.

[0038] From the fusion of the trusted vector time series matrix In the process, the runtime parameter sub-vector and supply chain information sub-vector are extracted according to the preset parameter field index. For the first... For each time slice, the subvector of runtime parameters and the subvector of supply chain information are as follows: , in, For the first The time slice's runtime parameter subvector represents the set of all runtime parameters within that time slice. It is the first The first time in the film The values ​​of the running parameters, This represents the first operating parameter of that time slice, such as voltage; This represents the second operating parameter for that time slice, such as frequency, and so on, until... ; The index of the time slice indicates the time point corresponding to the vector; The total number of runtime parameters indicates the number of parameters in each time slice. Different operating parameters For the first The supply chain information subvector of a time slice represents the set of all supply chain information within that time slice. It is the first The first time in the film The value of a supply chain information, Indicates the first The first time in the film Dimensional supply chain information value, The first supply chain information representing that time slice, such as the software version; This represents the second type of supply chain information, such as control strategies, and so on, until... , The index of the time slice indicates the time point corresponding to the vector. The total number of supply chain information, representing the amount of information in each time slice. Different supply chain information.

[0039] A sliding window method will be used to calculate the covariance and correlation coefficient sequences to measure the correlation between operating parameters and supply chain information. The window size will be set to... Each window contains The formulas for calculating covariance and correlation coefficient within a sliding window are as follows: (The text abruptly ends here, so the translation stops as well.) The sliding window covariance calculation is expressed as: , in, Indicates the first The sliding window covariance of time slices measures the relationship between operating parameters and supply chain information within a time window; For the first The runtime parameter values ​​for the time slice indicate the time slice's runtime parameters. The Middle The values ​​of each running parameter; For the first The supply chain information value of a time slice represents the time slice. The Middle Numerical values ​​for supply chain information; The mean of the running parameters represents the average value of all running parameter values ​​within the current sliding window; The mean of the supply chain information represents the average value of all supply chain information values ​​within the current sliding window. The size of the sliding window represents the number of time slices used to calculate the covariance; This is the index of the current time slice, representing the current time point when calculating the covariance.

[0040] The covariance formula is used to measure the mutual changing trends of operating parameters and supply chain information within a given time window. By calculating the covariance using a sliding window approach, the linear relationship between operating parameters and supply chain information within each time slice can be obtained.

[0041] The sliding window correlation coefficient is calculated as follows: , in, For the first The sliding window correlation coefficient of a time slice represents the correlation between operating parameters and supply chain information within the time window; For the first Sliding window covariance of time slices; For the first The variance of the operating parameters of a time slice represents the degree of fluctuation of all operating parameters within that time slice; For the first The variance of supply chain information within a time slice represents the degree of fluctuation of all supply chain information within that time slice.

[0042] The power spectral distribution of the discrete Fourier transform is expressed as: , in, This represents the power spectrum distribution after the Discrete Fourier Transform (DFT). The power spectrum represents the frequency composition of a signal, especially its variation trend in the time domain. Let be a time-domain signal sequence, representing the first... The signal value at a given time point; For frequency variables, it represents the frequency components of the signal in the frequency domain; The imaginary unit represents the imaginary part in complex number operations; For time index, it represents the index of the discrete time series. A point in time; The total number of time points represents the length of the time series on which the Discrete Fourier Transform is based.

[0043] Sliding window covariance sequence Sliding window correlation coefficient sequence Power spectrum distribution of Fourier transform Combined into a set of operational and supply chain coupling characteristics Represented as: , in, For the first The set of coupling features of time slices.

[0044] Coupled feature set Input into a supervised learning model In this process, the initial security level of the distributed energy storage unit is obtained. The security level determination model is expressed as follows: , in, For the first The security level of a time slice indicates the security status of a distributed energy storage unit at that specific point in time. For security level determination model For the set of coupling features The output of .

[0045] Within each time window, calculate the runtime stability gradient vector. and vulnerability risk gradient vector The formula is: , in, For the first The time slice stability gradient represents the change of operating parameters between two consecutive time points; For the first The vulnerability risk gradient of a time slice represents the change in supply chain information between two consecutive points in time. For the first The operating parameter values ​​of the time slice; For the first Supply chain information value of time slices.

[0046] Calculate the average of the absolute values ​​of the gradient vectors: , in The rate of change of operational stability represents the degree of change in system stability over a period of time; The vulnerability risk accumulation rate indicates the degree to which system vulnerability risks accumulate over a period of time; The dimension of the gradient vector represents the number of time slices.

[0047] Based on the rate of change in operational stability and the rate of accumulation of vulnerability risk, a pre-defined fault probability mapping table is retrieved. To obtain the failure probability : , in, For the first The failure probability of a time slice represents the failure probability of a distributed energy storage unit at that point in time. This is a fault probability mapping table that outputs the corresponding fault probabilities based on the rate of change in operational stability and the rate of accumulation of vulnerability risks. The fault probability mapping table is used to calculate the fault probability based on changes in system stability and the accumulation of vulnerability risks.

[0048] Initial security level and failure probability Combined into two-dimensional coordinates And according to the preset threshold matrix Perform region matching: , Among them, Zone For the first The time slice region type indicates that the region of the time slice is determined by a combination of security level and failure probability; According to security level and failure probability Output area type; BA is the standby area, indicating the area where distributed energy storage units are allocated to the standby pool; IA is the isolation area, indicating the area where distributed energy storage units are allocated to the isolation pool.

[0049] In a preferred embodiment, the safety level determination model is based on a multilayer perceptron architecture, comprising an input layer, multiple hidden layers, and an output layer. Each layer of the model optimizes the safety status of the energy storage unit, with input data including operating parameters obtained after feature extraction, supply chain information, and failure probabilities. The design of each layer is as follows: Input layer: The input layer receives the fused feature vectors. The number of neurons in this layer is equal to the dimension of the input features, and each feature is connected to one neuron.

[0050] Hidden layers: Three hidden layers were designed in the model, each using the ReLU activation function. The number of neurons in each hidden layer was optimized experimentally, initially set to twice the number of input features.

[0051] The role of hidden layers is to extract higher-order relationships between features through nonlinear transformations. For example, there may be complex nonlinear relationships between input features and output, so each hidden layer applies the ReLU activation function to enhance the model's expressive power.

[0052] Output Layer: The output layer uses a single neuron to represent the safety level of the distributed energy storage unit. The safety level ranges from 1 to 5, representing different safety state levels. A Softmax activation function is then used after the output layer to ensure that the output safety level conforms to a probability distribution and meets multi-class classification requirements.

[0053] The output layer uses the Softmax function to calculate the final safety level prediction result, which can be mapped to the safety status of the energy storage unit: 1 indicates extremely safe and 5 indicates extremely dangerous.

[0054] The training process includes: First, the data is standardized so that the mean of all input data is 0 and the variance is 1. This helps to speed up the training process and avoids affecting the convergence of the model due to some feature values ​​being too large or too small.

[0055] The feature set is received through the input layer. That is, the first The time slice covariance sequence, correlation coefficient sequence, and power spectrum distribution after Fourier transform are presented. Each feature is passed from the input layer to the hidden layer, where it is processed by weight matrix multiplication and activation function (ReLU).

[0056] At the output layer, based on the Softmax activation function, the model outputs a predicted probability distribution of the safety level of a distributed energy storage unit. Then, the loss is calculated based on the weighted cross-entropy loss function, and the model parameters are adjusted through backpropagation.

[0057] The gradient of each parameter is calculated using the backpropagation algorithm, and the model weights are updated using the Adam optimizer. This process is iterated until the value of the loss function converges to a small value.

[0058] To improve the model's generalization ability and accuracy, hyperparameters, including the learning rate, the number of hidden layer neurons, and the sample size per batch, were adjusted using cross-validation. Optimal hyperparameters were further sought using methods such as grid search or Bayesian optimization.

[0059] After training, the model is evaluated using a validation set, primarily through metrics such as accuracy, precision, recall, and F1 score, to ensure that the model can accurately predict the safety status of energy storage units in practical applications.

[0060] The trained model is integrated into the power grid dispatch and control system to predict the safety status of energy storage units in a real-time environment. And take corresponding scheduling strategies based on the prediction results.

[0061] The weighted cross-entropy loss function is expressed as: , in, This represents the output of the loss function; The number of categories (in this implementation, the number of security levels, set to 5); It is the first in the real label Each security level label (security level determined by experts); It is the first output of the model The predicted probabilities of each category; It is a category The weights are used to adjust the importance of each category in the loss calculation; It is the logarithm of the predicted probability, used to calculate the loss for classification tasks.

[0062] To ensure efficient model training, the Adam optimizer was chosen. It adaptively adjusts the learning rate and corrects gradients, avoiding problems caused by excessively high or low learning rates. By adjusting hyperparameters, the Adam optimizer effectively prevents vanishing and exploding gradients in practice, improving model convergence speed.

[0063] The update formula for the Adam optimizer is expressed as: , in, This represents the updated model parameters. Here are the parameter values ​​at the current time t, representing the model parameters that need to be optimized. and These are the decay rate hyperparameters that control the estimation of the first and second moments, typically set to 0.9 and 0.999, respectively. and Let represent the exponential decay values ​​of the decay rates of the first and second moments at each time step t, respectively. It is the gradient at the current time t. This represents the square of the gradient. It is a first-moment estimate of the gradient. It is a second-moment estimate of the gradient. It is the gradient at the current time step. and These are the first-order moment estimate and the second-order moment estimate after bias correction, respectively. It's the learning rate. It is a small constant used to prevent division by zero errors, and is set to 0. .

[0064] According to the equipment identification code, the rate of change of state of charge, maximum allowable charge and discharge power, temperature rise rate and internal resistance change rate of each distributed energy storage unit in the backup pool are collected to form a parameter vector sequence arranged in time slices. Using the initial safety level as an index, select a derating factor that matches the initial safety level from the preset derating factor table, and multiply the maximum allowable charge and discharge power component in the parameter vector sequence with the derating factor to generate a safe power upper limit sequence. Under the constraint of the safe power upper limit sequence, the charge state change rate component is accumulated with the power output value of the previous time slice to determine the target charge and discharge power value of each time slice. The temperature rise rate component is compared with the preset temperature threshold to calculate the maximum continuous running time of each time slice, and a time-segmented power table containing power value and running time is obtained. Using the internal resistance change rate component as a smoothing weight, weighted interpolation is performed on adjacent power values ​​of the time-segmented power meter, and amplitude limiting is performed when the interpolation result exceeds the safe power limit of the time slice, to obtain a smoothed and corrected continuous power sequence. The continuous power sequence is mapped to the scheduling and control time axis according to a unified time base to form the charging and discharging control curve of the distributed energy storage unit.

[0065] Specifically, based on the device identification code, the rate of change of state of charge, maximum permissible charge / discharge power, temperature rise rate, and internal resistance change rate of each distributed energy storage unit are collected from the backup pool, forming a parameter vector sequence arranged by time slices. In practice, firstly, these parameters are read from each distributed energy storage unit, and the data for each time slice is indexed according to the device identification code. The rate of change of state of charge, maximum permissible charge / discharge power, temperature rise rate, and internal resistance change rate for each time slice form a set of parameter values, which serve as the parameter vector for that time slice. Assume the time slice starts from... arrive The parameter vector for each device is represented as: , in, The rate of change of the state of charge. For the maximum permissible charge and discharge power, For the rate of temperature rise, This represents the rate of change of internal resistance.

[0066] Select a derating factor from the preset derating factor table that matches the initial safety level to adjust the maximum charge and discharge power. Assume the initial safety level is... The corresponding reduction coefficient is The maximum charge / discharge power adjusted using the derating factor is expressed by the following formula: , in, To match the initial security level The corresponding reduction coefficient, For the first The maximum permissible charge and discharge power of the time slice.

[0067] Next, under the constraint of the safe power upper limit sequence, the target charge / discharge power value for each time slot is calculated. Specifically, using the rate of change of state of charge and the power output value of the previous time slot, an accumulation calculation is performed to obtain the target charge / discharge power. The formula is as follows: , in, This represents the power output value from the previous time slice. For the first The rate of change of the state of charge of the time slice, For time step.

[0068] Furthermore, the maximum continuous runtime for each time slice is calculated. The maximum continuous runtime is limited by temperature variation; therefore, the maximum runtime that each time slice can run is calculated based on the difference between the temperature rise rate and a preset temperature threshold. The formula is as follows: , in, It is the temperature threshold. This is the current temperature. This is the rate of temperature rise. Using this formula, the maximum continuous operating time for each time slot can be controlled based on temperature, thereby preventing equipment damage due to overheating.

[0069] Next, the power value is smoothed using the rate of change of internal resistance. A weighted interpolation method is used to smooth the power values ​​between adjacent time slots to ensure a smooth transition in power output. The interpolation formula is as follows: , in, The power value is after smoothing. For the target power, For the first Smoothing weights of time slices This is the smoothed power value from the previous time slice. This method uses the rate of change of internal resistance to smooth the power value, reducing power fluctuations and ensuring stable equipment operation.

[0070] If the smoothed power value exceeds the safe power limit, a limiting process is performed to ensure that the power does not exceed the safe range. The limiting process formula is: , in, This is the final power output value. This is the upper limit for safe power.

[0071] Finally, the smoothed power sequence is mapped to the dispatch control time axis according to a unified time base. This step is to ensure that the power output of the energy storage units is consistent with the grid dispatch requirements. The mapping formula is as follows: , in, This refers to the power output value on the scheduling and control time axis.

[0072] In this embodiment, the derating factor is determined based on the initial safety level of the energy storage unit. The initial safety level is calculated by comprehensively considering various factors such as the energy storage unit's operating parameters, supply chain information, and failure probability, using a specific safety level determination model. Based on the calculated initial safety level, the corresponding derating factor is retrieved from a pre-defined derating factor table. This factor is selected based on the current safety status of the energy storage unit, aiming to reasonably limit its maximum charge and discharge power. Each safety level corresponds to a different derating factor. Specifically, a lower safety level corresponds to a larger derating factor, meaning the energy storage unit's power is more restricted to prevent safety issues caused by overcharging and discharging. After obtaining the derating factor, it is combined with the energy storage unit's maximum permissible charge and discharge power to drate the power, ensuring the energy storage unit's power output remains within a safe range. The calculation formula multiplies the maximum permissible charge and discharge power by the derating factor to obtain the safe power upper limit. This upper limit limits the energy storage unit's power output in each time slice.

[0073] The smoothing weight is calculated based on the change in internal resistance rate. The internal resistance change rate reflects the battery health and stability of the energy storage unit during charging and discharging. Typically, a large change in internal resistance during charging and discharging indicates poor stability of the energy storage unit, necessitating smoothing. Specifically, within each time slice, the internal resistance change rate data of the energy storage unit is collected through a monitoring system. A large internal resistance change rate indicates significant power fluctuations in that time slice, requiring smoothing. The internal resistance change rate is used as a smoothing weight in the weighted interpolation of the power value to reduce power fluctuations caused by internal resistance changes. The weight calculation method standardizes the internal resistance change rate for each time slice, giving higher smoothing weights to time slices with larger internal resistance change rates. The smoothed power value is then obtained by weighted averaging of the power from the current time slice and the previous time slice, yielding the smoothed target power. This ensures the stability of the energy storage unit during charging and discharging, preventing power fluctuations from affecting the equipment.

[0074] Step 3: Input the charging and discharging control curves of the distributed energy storage units in the standby pool into the power dispatch optimization model that includes power flow constraints, state of charge constraints and standby capacity constraints, perform optimization solution, and output the dispatch optimization results.

[0075] In step 3, the charge and discharge control curves, state of charge sequence, available capacity sequence and power change rate sequence of the distributed energy storage units in the backup pool are indexed and integrated according to the equipment identification code to form an input parameter vector sequence arranged by time slice; Based on the node voltage, branch power and phase angle difference components in the input parameter vector sequence, interval pruning, boundary limiting and rate of change limiting are performed according to the power flow constraint parameter set to obtain the first modified sequence that satisfies the power flow constraint. Based on the state of charge and the rate of change of the state of charge in the input parameter vector sequence, interval pruning, boundary limiting and rate of change limiting are performed according to the state of charge constraint parameter set to obtain a second modified sequence that satisfies the state of charge constraint. Based on the available capacity components in the input parameter vector sequence, interval pruning, boundary limiting, and rate of change limiting are performed according to the spare capacity constraint parameter set to obtain a third modified sequence that satisfies the spare capacity constraint. The first, second, and third correction sequences are merged element-wise under a unified time base to generate a constraint-corrected scheduling input sequence. The constrained and modified scheduling input sequence is input into the power dispatch optimization model. The objective function is to minimize system power fluctuation, maximize reserve capacity utilization, and minimize charging and discharging losses. The optimization solution is executed, and the dispatch optimization results are output.

[0076] Specifically, firstly, an input parameter vector is established for each energy storage unit according to the device identification code, represented as follows: , , in, Indicates time slice Time storage unit The input parameter column vector; Indicates time slice Time storage unit The reference active power; Indicates time slice Time storage unit The state of charge; Indicates time slice Time storage unit The upper limit of available capacity (in power equivalent). Indicates time slice Time storage unit The rate of change of power; This indicates the time slices obtained by splicing them together in the order of device identification codes. The sequence of input parameter vectors; This represents the set of energy storage units within the standby pool that participate in scheduling; express Cell index in; By injecting the active power output of the energy storage into the node, the mapping between the node-injected active power and the linearized power flow is obtained: , in, Represents a node In time slice The effective injection; Indicates energy storage unit Whether to connect to the node The elements of the mapping matrix, with a value of 1, indicate access; This refers to injections other than energy storage (conventional units, distributed power sources, net load). Indicates time slice The node phase angle vector; This represents the generalized inverse of the electric Narlatanus matrix after removing the reference node; Indicates time slice The node active injection vector, its th Each component is ; Indicates time slice branch road Phase angle difference between the two ends; Indicates branch index, parentheses Indicate its two endpoints; Represents the set of power grid branches; Representing time slices node ,node The phase angle; Indicates time slice branch road The meritorious trend; Indicates a branch The equivalent susceptance coefficient; Indicates time slice node The voltage amplitude; Represents a node The amplitude of the ground-state voltage; Elements representing the sensitivity matrix of node voltage to active power injection; Represents the set of power grid nodes; express The node index in the data.

[0077] By performing minimum deviation projection on the reference active power under power flow constraints, the first correction sequence is obtained: , , in, Indicates time slice The power flow constraint corrects the active power vector; Represents the column vector of optimization variables, whose first... Components ; Indicates time slice The reference active power column vector, its first... Components ; Representing nodes respectively The lower and upper limits of voltage; Indicates will Substitute the nodes obtained by the power flow mapping Voltage; Indicates a branch The upper limit of the trend; Indicates will Substituting the branches obtained by the power flow mapping trend; Indicates a branch The upper limit of the phase angle difference; Indicates will Substituting the branches obtained by the power flow mapping Phase angle difference; Indicates time slice Energy storage unit The candidate active power; This indicates the reference active power of the previous time slice; Indicates energy storage unit The upper limit of the rate of change of power, It is the 2-norm (Euclidean norm) of a vector, representing the square root of the sum of the squares of the vector's components.

[0078] Projecting the first corrected sequence under the state-of-charge constraint yields the second corrected sequence: , , in, Indicates time slice The active power vector is corrected by the state of charge constraint; Indicates time slice The charged state vector; The column vector representing the remaining power (State of Charge, SOC) optimization variables, its i.e., Components ; Indicates time slice Energy storage unit SOC; Indicates energy storage unit The charging efficiency and discharging efficiency; express The non-negative part; express The absolute value of the non-positive part; Indicates the length of the time slice; These represent the lower and upper limits of SOC, respectively. This indicates the upper limit of SOC change between adjacent time slices; This represents the candidate active power column vector.

[0079] Projecting the second corrected sequence again under the spare capacity constraint yields the third corrected sequence: , in, Indicates time slice The active power vector is corrected based on the reserve capacity constraint; Indicates time slice Energy storage unit The upper limit of the available capacity converted to power; This indicates the candidate active power in the previous time slice; This indicates the upper limit of the rate of change of power.

[0080] The three types of corrected sequences are element-wise weighted and merged, and a uniform projection is performed on the intersection of the three feasible regions to obtain feasible inputs for objective optimization: , in, Indicates time slice Element-wise weighted combined active power vector; These represent the combined weights for the three types of corrections: power flow, state of charge (SOC), and reserve capacity. Indicates time slice The feasible active power vector after uniform projection; Represents a set Euclidean projection operator; This represents the feasible region that satisfies the power flow constraints. This represents the feasible region that satisfies the SOC constraint; This represents the feasible region that satisfies the spare capacity constraint.

[0081] by To determine the decision-making complexity, joint optimization is performed across all time slices to obtain the optimal scheduling result: , in, Represents the set of feasible active power vectors for all time slices; Represents the set of scheduling time slices; These represent the weight coefficients of each term in the objective function; Indicates time slice A measure of system power fluctuation; express The Quantity; Indicates time slice Energy storage unit The upper bound of the capacity; This represents a measure of capacity utilization. Indicates time slice The system power fluctuation. As a secondary loss coefficient, it characterizes the effect of internal resistance power loss increasing with the square of power. As a primary loss coefficient, it characterizes the loss that is linearly related to power in the energy conversion process.

[0082] Specifically, time slices The system power fluctuation is expressed as: , in, Indicates time slice The total amount of active power injection changes in the system; Indicates time slice node The effective injection; Represents the set of power grid nodes; express Node index in; This indicates the index of the previous time slice.

[0083] It should be noted that when constructing the power flow constraint correction model, the upper and lower limits of the node voltage are... This is obtained from power system design specifications. In this embodiment, for a 10kV distribution network node, the allowable voltage fluctuation range is set to the rated value. Branch power limit The upper limit of the branch phase angle difference is determined by the cross-section of the branch conductor, its current carrying capacity, and its thermal stability. In engineering practice, it is directly determined through design drawings or equipment nameplate parameters. Based on system static stability analysis, this embodiment uses a limit value of no more than 0.35 radians to avoid loss of voltage stability. Upper limit of power change rate. The maximum ramping capability of the energy storage unit inverter is determined by fitting the measured value. For example, a continuous ramping test is conducted under rated operating conditions, the maximum amplitude of power variation per unit time is recorded, and this value is taken as a constraint parameter.

[0084] When constructing the state-of-charge constraint correction model, the upper and lower limits of SOC are... The value is determined based on the recommended values ​​given by the battery manufacturer, for example, for lithium iron phosphate batteries. to The optimal loop within the range is chosen, therefore... As the lower limit, The upper limit. The upper limit of SOC variation between adjacent time slices. Based on battery life experiments, the maximum change in SOC was gradually increased during accelerated cycling tests, and the rate of lifespan degradation was recorded. The maximum change that did not cause a sharp decrease in lifespan was selected as the parameter value. Charging efficiency. With discharge efficiency The ratio of charging input energy to discharging output energy was calculated by measuring the sample battery in a constant current and constant voltage charge-discharge experiment, and then fitted to a constant value in long-term experiments.

[0085] When constructing the backup capacity constraint correction model, the upper limit of available capacity is... The capacity is obtained through battery capacity calibration tests. Specifically, the battery is fully charged and discharged at its rated rate, and the output capacity is measured. A curve is established showing the relationship between the measured capacity and the current state of charge, and then the capacity is measured at any given time interval. The remaining discharge capacity of the battery is converted into the upper limit of usable power. This value needs to be adjusted according to ambient temperature and aging conditions. Therefore, in practice, a correction factor for temperature and cycle number is established based on experiments to form a time-varying parameter curve of usable capacity.

[0086] In the objective function part, the weighting coefficients This is achieved through scheduling simulation and fitting with historical operational data. Specifically, the importance of three indicators—system power fluctuation, reserve capacity utilization, and charging / discharging losses—is statistically analyzed from historical power grid scheduling data. Scheduling experts assign priorities, and multiple weighted combinations are used in simulation optimization to compare the overall system efficiency. The optimal combination that balances stability, economy, and reliability is ultimately selected. For example, if the simulation reveals that system power fluctuation has the most significant impact on operation, then the weighting is increased. The value is determined until the fluctuation satisfies the constraint without significantly increasing the loss.

[0087] equivalent loss coefficient The relationship between measured energy loss and output power was obtained by fitting a curve. First, the energy storage unit was operated at different power levels on an experimental platform, and the difference between input and output electrical energy was recorded to obtain the loss value. Then, least squares regression was performed on the relationship between the loss value and output power to fit a quadratic polynomial coefficient, where... This represents the loss effect of the power square term. This represents the loss effect of the primary power term. In practical applications, this coefficient is periodically calibrated and updated to ensure it matches the aging state of the energy storage unit.

[0088] In the power flow calculation mapping section, the node voltage sensitivity matrix... Equivalent susceptance coefficient of branch Direct calculation using grid parameters. Specifically, this involves establishing the admittance matrix using the system's line parameters (resistance, reactance), and then solving the linearized power flow equations to obtain the sensitivity matrix. Ground state voltage. The rated operating node voltage is then directly set using the system steady-state calculation. Non-energy storage active power injection. Derived from load forecasting and conventional power generation scheduling plans, this embodiment uses day-ahead load forecasting results and thermal power unit baseline output as inputs.

[0089] Finally, the weights of the three types of modified merging. The criteria are determined through multi-objective normalization and expert trade-offs. During the training phase, the violation rate, feasibility, and improvement effect on system performance indicators of the sequence after correction of the three types of constraints are evaluated. The importance of the three types of constraints is quantified as a percentage and mapped to weight coefficients. For example, when the SOC constraint has the highest violation probability during system operation, it is increased in the weight calculation. The value of makes the satisfaction priority of SOC constraints higher.

[0090] From the output of the power dispatch optimization model, the optimal active power sequence, optimal reactive power sequence, available capacity utilization rate sequence, and time slice duration sequence are parsed according to the equipment identification code. Based on the available capacity utilization rate sequence, the rated capacity is proportionally allocated to generate a reserve capacity ratio sequence; Calculate the power rise slope sequence and power fall slope sequence based on the optimal active power sequence and the duration sequence of the time slice; convert the optimal active power sequence of the time slice into a DC side current reference sequence based on the rated DC voltage and converter efficiency in the preset inverter nameplate parameter table; use the optimal reactive power sequence of the time slice as the reactive power reference sequence, and combine them to form an inverter control parameter set that includes the power rise slope sequence, the power fall slope sequence, the DC side current reference sequence, and the reactive power reference sequence; The standby capacity ratio sequence and the inverter control parameter set are encapsulated into a time-slice scheduling instruction data frame, which is then sent to the distributed energy storage unit controller in chronological order and according to the device identification code. The distributed energy storage unit controller performs active power tracking control and slope limiting control in each time slice, performs reactive power tracking control, sets the upper limit of available power and the state of charge window according to the reserve capacity ratio sequence, and completes the scheduling operation.

[0091] Furthermore, during the scheduling process, when it is detected that the initial safety level of any distributed energy storage unit in the backup pool is lower than a preset threshold, the charge and discharge control curve generated by the distributed energy storage unit is invoked to switch the load of the distributed energy storage unit to a distributed energy storage unit in the backup pool other than the distributed energy storage unit. A control barrier function is used to limit the voltage and frequency changes during the switching process, and the distributed energy storage unit with an initial safety level lower than the threshold is moved into the isolation pool.

[0092] Example 2 is an embodiment of the present invention, which provides a distributed energy storage power dispatch and control system based on the power grid software supply chain, including: The data acquisition module is used to collect the operating parameters of the distributed energy storage unit and the supply chain information of the power grid software, perform matching verification between the operating parameters and the control strategy of the distributed energy storage unit, perform software bill of materials hash verification and controller remote measurement report verification of the supply chain information, and integrate the verified operating parameters and verified supply chain information to generate reliable operating status data. The analysis module is used to calculate the initial safety level of the distributed energy storage unit using trusted operating status data, and predict the failure probability of the distributed energy storage unit based on the historical trend of operating parameters in the trusted operating status data. Taking the initial safety level and failure probability as input, the distributed energy storage unit is divided into a backup pool and an isolation pool, and the charging and discharging control curves of the distributed energy storage unit in the backup pool are generated. The scheduling module is used to input the charging and discharging control curves of the distributed energy storage units in the standby pool into the power scheduling optimization model, which includes power flow constraints, state of charge constraints, and standby capacity constraints, and output the scheduling optimization results.

[0093] This embodiment also provides an electronic device applicable to the distributed energy storage power dispatch and control method based on the power grid software supply chain, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the distributed energy storage power dispatch and control method based on the power grid software supply chain proposed in the above embodiment.

[0094] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements the distributed energy storage power dispatch and control method based on the power grid software supply chain proposed in the above embodiments.

[0095] The storage medium proposed in this embodiment and the method for implementing distributed energy storage power dispatch and control based on the power grid software supply chain proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0096] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0097] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A distributed energy storage power source dispatch and control method based on power grid software supply chain, characterized in that: include, The system collects the operating parameters of the distributed energy storage unit and the supply chain information of the power grid software, performs matching verification between the operating parameters and the control strategy of the distributed energy storage unit, performs software bill of materials hash verification and controller remote measurement report verification of the supply chain information, and merges the verified operating parameters and verified supply chain information to generate reliable operating status data. The initial safety level of the distributed energy storage unit is calculated using trusted operating status data, and the failure probability of the distributed energy storage unit is predicted based on the historical trend of operating parameters in the trusted operating status data. The initial safety level and failure probability are used as inputs to divide the distributed energy storage unit into a backup pool and an isolation pool, and the charge and discharge control curves of the distributed energy storage unit in the backup pool are generated. The charging and discharging control curves of the distributed energy storage units in the standby pool are input into the power dispatch optimization model, which includes power flow constraints, state of charge constraints, and standby capacity constraints. The optimization solution is then executed, and the dispatch optimization results are output.

2. The distributed energy storage power dispatch and control method based on the power grid software supply chain as described in claim 1, characterized in that: The matching verification of the execution operating parameters and the control strategy of the distributed energy storage unit includes: using the name of the operating parameter as an index, searching for the allowable threshold range of the same name item in the preset firmware version parameter table and control strategy parameter table; comparing the measured value of each operating parameter with the searched allowable threshold range; and determining that the operating parameter verification is qualified when all measured values ​​fall within their respective searched allowable threshold ranges. The software bill of materials hash verification and controller remote measurement report verification for executing supply chain information include: comparing the software bill of materials hash value with the controller remote measurement report for the supply chain information; calculating the hash value of the software bill of materials and comparing it with a preset trusted hash value; if they match, it is recorded as passed. The measured value of each detection item in the remote measurement report of the controller is compared with the baseline value of the same detection item in the preset trusted baseline data. If the absolute value of the difference between the two does not exceed the preset deviation limit, it is considered to pass. When both the comparison of the software bill of materials hash value and the comparison of the measurement report are considered to pass, the supply chain information verification is deemed to be qualified.

3. The distributed energy storage power dispatch and control method based on the power grid software supply chain as described in claim 2, characterized in that: The process of fusing verified operating parameters with verified supply chain information to generate reliable operating status data includes: according to the equipment identification code of the distributed energy storage unit, searching for operating parameter records and supply chain information records with the same equipment identification code in the operating parameters and supply chain information respectively, establishing an index association relationship, and forming an operating parameter index table and a supply chain information index table; Based on a unified time base, time alignment processing is performed on the records in the operation parameter index table and the supply chain information index table to generate a time-aligned set of operation parameters and a time-aligned set of supply chain information. Map the time-aligned set of operating parameters to operating parameter trust vectors according to parameter type, and map the time-aligned set of supply chain information to supply chain trust vectors according to information fields; An element-wise weighted operation is performed on the trusted vector of operating parameters and the trusted vector of the supply chain to obtain a fused trusted vector. The fused trusted vector is then bound to the device identification code according to the device identification code to generate trusted operating status data containing the device identification code and the fused trusted vector.

4. The distributed energy storage power dispatch and control method based on the power grid software supply chain as described in claim 3, characterized in that: The step of dividing the distributed energy storage unit into a backup pool and an isolation pool includes continuously sampling the fused trusted vector in the trusted operating status data at equal time intervals according to the device identification code, and generating a fused trusted vector time series matrix with time series as rows and vector components as columns. The running parameter sub-vector and the supply chain information sub-vector are extracted from the time series matrix according to the preset parameter field index. The sliding window covariance sequence, the sliding window correlation coefficient sequence and the power spectrum distribution of the discrete Fourier transform are calculated in the time series dimension and combined to form a set of running and supply chain coupling features. The set of coupling features is input into a security level determination model obtained through supervised learning using historical operational failure records and software vulnerability event records as training samples, and the initial security level of the distributed energy storage unit is output. Based on the set of coupling features, within each time window, the first-order difference of the operation stability index sequence along the time axis is calculated to obtain the operation stability gradient vector, and the first-order difference of the vulnerability risk index sequence along the time axis is calculated to obtain the vulnerability risk gradient vector. The rate of change of operational stability is obtained by taking the arithmetic mean of the absolute values ​​of each component of the operational stability gradient vector. The accumulation rate of vulnerability risk is obtained by taking the arithmetic mean of the absolute values ​​of each component of the vulnerability risk gradient vector. The failure probability is obtained by retrieving a preset failure probability mapping table based on the combined interval of the rate of change in operational stability and the rate of accumulation of vulnerability risk. The initial safety level and failure probability are mapped to a combination of two-dimensional coordinates. The boundary between the reserve area and the isolation area in the threshold matrix is ​​used for area matching. If the fault falls into the reserve area, it is assigned to the reserve pool; if it falls into the isolation area, it is assigned to the isolation pool.

5. The distributed energy storage power dispatch and control method based on the power grid software supply chain as described in claim 4, characterized in that: The generation of charge and discharge control curves for distributed energy storage units in the backup pool includes collecting the rate of change of state of charge, maximum allowable charge and discharge power, temperature rise rate, and internal resistance change rate of each distributed energy storage unit in the backup pool according to the equipment identification code, forming a parameter vector sequence arranged in time slices. Using the initial safety level as an index, select a derating factor that matches the initial safety level from the preset derating factor table, and multiply the maximum allowable charge and discharge power component in the parameter vector sequence with the derating factor to generate a safe power upper limit sequence. Under the constraint of the safe power upper limit sequence, the charge state change rate component is accumulated with the power output value of the previous time slice to determine the target charge and discharge power value of each time slice. The temperature rise rate component is compared with the preset temperature threshold to calculate the maximum continuous running time of each time slice, and a time-segmented power table containing power value and running time is obtained. Using the internal resistance change rate component as a smoothing weight, weighted interpolation is performed on adjacent power values ​​of the time-segmented power meter, and amplitude limiting is performed when the interpolation result exceeds the safe power limit of the time slice, to obtain a smoothed and corrected continuous power sequence. The continuous power sequence is mapped to the scheduling and control time axis according to a unified time base to form the charging and discharging control curve of the distributed energy storage unit.

6. The distributed energy storage power dispatch and control method based on the power grid software supply chain as described in claim 5, characterized in that: The output scheduling optimization results include indexing and integrating the charge and discharge control curves, state of charge sequence, available capacity sequence and power change rate sequence of the distributed energy storage units in the backup pool according to the equipment identification code to form an input parameter vector sequence arranged in time slices; Based on the node voltage, branch power and phase angle difference components in the input parameter vector sequence, interval pruning, boundary limiting and rate of change limiting are performed according to the power flow constraint parameter set to obtain the first modified sequence that satisfies the power flow constraint. Based on the state of charge and the rate of change of the state of charge in the input parameter vector sequence, interval pruning, boundary limiting and rate of change limiting are performed according to the state of charge constraint parameter set to obtain a second modified sequence that satisfies the state of charge constraint. Based on the available capacity components in the input parameter vector sequence, interval pruning, boundary limiting, and rate of change limiting are performed according to the spare capacity constraint parameter set to obtain a third modified sequence that satisfies the spare capacity constraint. The first, second, and third correction sequences are merged element-wise under a unified time base to generate a constraint-corrected scheduling input sequence. The constrained and modified scheduling input sequence is input into the power dispatch optimization model. The objective function is to minimize system power fluctuation, maximize reserve capacity utilization, and minimize charging and discharging losses. The optimization solution is executed, and the dispatch optimization results are output.

7. The distributed energy storage power dispatch and control method based on the power grid software supply chain as described in claim 6, characterized in that: The output scheduling optimization results also include parsing the optimal active power sequence, optimal reactive power sequence, available capacity utilization rate sequence, and time slice duration sequence from the output results of the power scheduling optimization model according to the equipment identification code. Based on the available capacity utilization rate sequence, the rated capacity is proportionally allocated to generate a reserve capacity ratio sequence; Calculate the power rise slope sequence and power fall slope sequence based on the optimal active power sequence of the time slice and the duration sequence of the time slice; Based on the rated DC voltage and commutation efficiency in the preset inverter nameplate parameter table, the optimal active power sequence of the time slot is converted into the DC side current reference sequence; the optimal reactive power sequence of the time slot is used as the reactive power reference sequence, and the two sequences are combined to form a set of inverter control parameters including the power rise slope sequence, the power fall slope sequence, the DC side current reference sequence, and the reactive power reference sequence. The standby capacity ratio sequence and the inverter control parameter set are encapsulated into a time-slice scheduling instruction data frame, which is then sent to the distributed energy storage unit controller in chronological order and according to the device identification code. The distributed energy storage unit controller performs active power tracking control and slope limiting control in each time slice, performs reactive power tracking control, sets the upper limit of available power and the state of charge window according to the reserve capacity ratio sequence, and completes the scheduling operation.

8. A distributed energy storage power dispatch and control system based on the power grid software supply chain, employing the distributed energy storage power dispatch and control method based on the power grid software supply chain as described in any one of claims 1 to 7, characterized in that, include: The data acquisition module is used to collect the operating parameters of the distributed energy storage unit and the supply chain information of the power grid software, perform matching verification between the operating parameters and the control strategy of the distributed energy storage unit, perform software bill of materials hash verification and controller remote measurement report verification of the supply chain information, and integrate the verified operating parameters and verified supply chain information to generate reliable operating status data. The analysis module is used to calculate the initial safety level of the distributed energy storage unit using trusted operating status data, and predict the failure probability of the distributed energy storage unit based on the historical trend of operating parameters in the trusted operating status data. Taking the initial safety level and failure probability as input, the distributed energy storage unit is divided into a backup pool and an isolation pool, and the charging and discharging control curves of the distributed energy storage unit in the backup pool are generated. The scheduling module is used to input the charging and discharging control curves of the distributed energy storage units in the standby pool into the power scheduling optimization model, which includes power flow constraints, state of charge constraints, and standby capacity constraints, and output the scheduling optimization results.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the distributed energy storage power dispatch and control method based on the power grid software supply chain as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the distributed energy storage power dispatch and control method based on the power grid software supply chain as described in any one of claims 1 to 7.

Images