Performance judgment method and system for improving static voltage stability index of energy storage power station

Abstract

The invention discloses a performance judgment method for improving a static voltage stability index of an energy storage power station, and the method comprises the steps: obtaining data information of a target power system and the energy storage power station, carrying out the preprocessing, and carrying out the screening through combining a causal decomposition method, and obtaining an energy storage adjustable key variable; constructing a mapping model based on a convolutional network model and training the mapping model to obtain a mapping relation between the energy storage adjusting quantity and the index improvement quantity; dividing the state of the target power system and performing corresponding control; and according to the obtained control result, completing the performance judgment of improving the static voltage stability index of the energy storage power station. The invention also discloses a system for realizing the performance judgment method for improving the static voltage stability index of the energy storage power station. According to the method, the performance judgment of the static voltage stability index of the energy storage power station is improved, the reliability is higher, and the accuracy is better.

Description

Technical Field

[0001] This invention belongs to the field of electrical automation, and specifically relates to a performance evaluation method and system for improving the static voltage stability index of an energy storage power station. Background Technology

[0002] With economic and technological development and the improvement of people's living standards, electricity has become an indispensable secondary energy source in people's production and daily life, bringing endless convenience. Therefore, ensuring a stable and reliable supply of electricity has become one of the most important tasks of the power system.

[0003] Currently, an increasing number of new energy power generation systems are being integrated into the power grid and generating electricity. The fluctuations and randomness of the output from these new energy power generation systems pose significant challenges to the safe and stable operation of the power system. Energy storage systems, due to their ability to rapidly regulate both active and reactive power, can flexibly mitigate fluctuations in new energy output and supplement voltage support, and have become the mainstream technology choice for improving the static voltage stability of the distribution network.

[0004] However, existing performance evaluation schemes for improving static voltage stability in energy storage power stations have significant shortcomings; the existing evaluation schemes have relatively singular indicator selection and cannot make a comprehensive and balanced objective evaluation. Summary of the Invention

[0005] One of the objectives of this invention is to provide a reliable and accurate method for judging the performance of energy storage power stations in improving static voltage stability.

[0006] The second objective of this invention is to provide a system for determining the performance of the energy storage power station in improving static voltage stability.

[0007] The performance evaluation method for improving static voltage stability in an energy storage power station provided by this invention includes the following steps:

[0008] S1. Obtain data information about the target power system and energy storage power station;

[0009] S2. Preprocess the data obtained in step S1 and use the causal decomposition method to screen out the key variables that can be controlled in energy storage;

[0010] S3. Based on the convolutional network model, construct and train a mapping model to obtain the mapping relationship between energy storage regulation amount and index improvement amount;

[0011] S4. Based on the mapping model obtained in step S3, the state of the target power system is divided and corresponding control is performed;

[0012] S5. Based on the control results obtained in step S4, complete the performance assessment of the energy storage power station to improve the static voltage stability index.

[0013] Step S1, which involves acquiring data information about the target power system and energy storage power station, specifically includes the following steps:

[0014] Under the given scenario, acquire the following data from the target power system and energy storage power station:

[0015] Obtain the voltage amplitude at each node. Phase angles of each node Line active power Line reactive power Real-time load demand Wind power output data Photovoltaic power output data And energy storage system data;

[0016] The energy storage system data includes the energy storage system's state of charge. Active charging and discharging power Reactive charging and discharging power and the delay time from receiving the regulation command to the power response ;

[0017] The scenarios described include a wind power output decrease by a set amount within a set time period and a load surge by a set amount.

[0018] Step S2, which involves preprocessing the data obtained in step S1, specifically includes the following steps:

[0019] Based on the data obtained in step S1, construct the basic dataset. for ; For the total number of data, For the i-th set of input data, including voltage amplitude Phase angle , , , , , , and ; The output data for the i-th group includes active power margin. and voltage deviation rate ;

[0020] For the basic dataset An enhanced sample set is generated using a classification-targeted perturbation method. , represented as M represents the total number of data points in the augmented sample set; among them, wind power output data is calculated according to the formula. By perturbing the wind, the wind power output can be enhanced. , This is the wind power output disturbance coefficient. and , The set standard deviation, The maximum fluctuation coefficient of wind power output is set, and it is stipulated that... , For wind power output, The rated power of the wind power is used; the reactive power charging and discharging power data are calculated according to the formula. By perturbing, the reactive power charging and discharging power is enhanced. , The set number of steps for the reactive power disturbance of energy storage. The step value of reactive power disturbance of energy storage and , For the rated reactive power regulation capacity of energy storage, To improve energy storage regulation accuracy; Replace the base dataset In At the same time Replace the base dataset In To obtain enhanced input data Based on enhanced input data The Newton-Raphson power flow calculation scheme was used to calculate the enhanced output data. for , To calculate the enhanced active power margin, The calculated enhanced voltage deviation rate;

[0021] For enhanced output data Remove or After obtaining the samples, they are then compared with the base dataset. By merging, we obtain the total sample set. for , The input data is for the total sample set. This is the output data for the total sample set;

[0022] For the total sample set Remove Or samples where the line power exceeds the set value. This is the set minimum node voltage. The maximum node voltage is set; then, 3 is used. The criteria and DBSCAN clustering algorithm are used to identify and remove outlier samples; calculations are performed. rate of change, excluding The sample whose rate of change is greater than the maximum charge and discharge rate of the energy storage system;

[0023] For proportional features, an interval scaling method is used to map them to... The proportional features mentioned in the interval specifically include line active power, line reactive power, wind power output data, photovoltaic power output data, energy storage active charging power, energy storage active discharging power, energy storage reactive charging power, and energy storage reactive discharging power; for numerical features, a Z-score standardization scheme is used for processing; for angular features, a... The process is handled using a periodic standardization method;

[0024] Finally, before supplementing The statistical characteristics and difference characteristics of the sliding window at each time step are used to obtain the preprocessed sample set. for ,in To preprocess the input data of the sample set, The output data for the preprocessed sample set, where T is the total number of data in the preprocessed sample set; the supplementary data... The sliding window statistical characteristics and difference characteristics at each time point include the following steps:

[0025] Construction of sliding window statistical features: For each input variable, calculate the mean, variance, maximum and minimum values ​​over the first k time steps to construct sliding window statistical features; sliding window statistical features are used to reflect the overall distribution characteristics of the variable within the time window;

[0026] Construction of difference features: Calculate the difference between the current time and the previous time for each input variable, as well as the difference between the current time and the mean of the previous k time steps, to construct difference features; difference features are used to reflect the dynamic changing trend of variables;

[0027] Feature Supplementation: The constructed sliding window statistical features and difference features are concatenated with the original input variables to form preprocessed input data with expanded dimensions. .

[0028] Step S2, which involves using causal decomposition to screen key variables for energy storage control, specifically includes the following steps:

[0029] Based on the obtained preprocessed sample set ,calculate Input variables and output vectors The Pearson correlation coefficient was used to select input variables with a Pearson correlation coefficient greater than a set value to form a candidate variable set. ;right Multicollinearity was tested, and the variance inflation factor (VIF) was calculated. Variables with VIF values ​​greater than a set threshold were removed to obtain a set of intermediate variables without significant multicollinearity. ;from Through screening, variables that can be actively controlled by energy storage and key controllable variables on the system side are obtained, resulting in a set of key variables. The process involves screening out variables that can be actively controlled in energy storage and key controllable variables on the system side, specifically including the following steps:

[0030] Screening of active regulation variables for energy storage: Screening variables that the energy storage system can directly adjust, including active charging power, active discharging power, reactive charging power, reactive discharging power, and state of charge.

[0031] Screening of key controllable variables on the system side: Screening variables that have a greater impact on voltage stability indicators than the set value and can be indirectly controlled through scheduling means, including node voltage reference values ​​and line power flow distribution coefficients.

[0032] A causal graph model of variables is constructed using the causal decomposition method, and conditional independence is tested using the PC algorithm. Variables related to energy storage systems The causal strength is determined by selecting variables whose causal strength is greater than a set value.

[0033] Define the potential coefficient for improvement for , for The changes in variables related to energy storage systems. For the corresponding Changes; Calculation of preprocessed sample sets Each input variable in Value, select The top few groups of input variables with the highest values;

[0034] Finally, variables with causal strength greater than a set value (preferably 0.5) obtained through conditional independence testing using the PC algorithm are compared with the improvement potential coefficient. The intersection of the top 60% of the highest-valued input variables is used to obtain the key variables for energy storage control.

[0035] Step S3, which involves constructing and training a mapping model based on a convolutional network model to obtain the mapping relationship between energy storage regulation and index improvement, specifically includes the following steps:

[0036] The key variables for energy storage controllability obtained in step S2 are used as input variables. ;

[0037] A mapping model is constructed using an inverse feature enhancement temporal convolutional network model: the model includes three causal convolutional modules, a normalization module, and an activation function module;

[0038] In the three causal convolutional modules, the output of the l-th causal convolutional module is represented as

[0039] In the formula This is the output of the l-th causal convolutional module; This describes the processing procedure of the causal convolution module. This is the output of the module above, and it is set... ; The kernel size; The expansion rate is satisfied, and the receptive field is satisfied. , This represents the receptive field of the l-th module;

[0040] The output of the last causal convolutional module is standardized using the following formula:

[0041] In the formula This is the standardized output; This is the output of the last causal convolutional module; This is the first parameter to be learned; This is the second parameter to be learned; This is the batch average. This represents the batch variance. This is a set minimum value to prevent the denominator from being 0;

[0042] Constructing the node-adjustment sensitivity matrix The element in the i-th row and j-th column The calculation formula is , Let be the reactive power regulation of the j-th energy storage unit;

[0043] The input variables are calculated using the following formula. Sensitivity coefficients of each variable:

[0044] In the formula Let be the sensitivity coefficient of the j-th input variable; This represents the change in the static voltage stability index. Let j be the j-th input variable;

[0045] Training the constructed mapping model: using a preprocessed sample set The mapping model is trained; the output of the mapping model is the energy storage regulation amount. , represented as During training, the weighted MSE loss function is used, and the Adam optimizer is employed for training.

[0046] Step S4, based on the mapping model obtained in step S3, divides the state of the target power system and performs corresponding control, specifically including the following steps:

[0047] Based on the input of the mapping model obtained in step S3, obtain the data information of the target power system at the current moment;

[0048] The acquired data is input into the mapping model obtained in step S3 to obtain the benchmark value of the voltage stability index of the target power system at the current moment, expressed as:

[0049] In the formula This is the baseline value for active power margin; This represents the initial active power margin without energy storage regulation. The active power margin improvement is the amount corresponding to the baseline adjustment output by the mapping model. This is the reference value for voltage deviation rate; This represents the initial voltage deviation rate without energy storage regulation. The voltage deviation rate improvement corresponding to the baseline adjustment output of the mapping model;

[0050] Based on the obtained voltage stability index benchmark value, the state of the target power system is classified as follows:

[0051] like and If so, the target power system is determined to be operating in a safe zone;

[0052] like or If so, the target power system is determined to be operating in the warning zone;

[0053] like or If so, the target power system is determined to be operating in the critical region;

[0054] Based on the state division results of the target power system, corresponding control measures are implemented:

[0055] Controls located in the safe zone:

[0056] Maintain the current control method unchanged; if The rate of change is lower than the set value or If the rate of change is higher than the set value, the control will switch to the control in the warning zone;

[0057] Control measures located in the warning zone:

[0058] The following formula is used for calculation:

[0059] In the formula Let be the energy storage reactive power control adjustment amount at time t; This is the rated reactive power capacity for energy storage; It is a symbolic function; The target reactive power regulation; Let be the energy storage active power control adjustment at time t; This is the rated active capacity of the energy storage. The target active power adjustment;

[0060] Real-time calculation and And perform corresponding control on the energy storage system;

[0061] Control in the critical region:

[0062] The energy storage system's regulation output is maximized, and the reactive power regulation is determined based on the direction of the deviation between the node voltage and the rated voltage; until... and The conditions for the warning zone are met, and the area within the warning zone is placed under control.

[0063] Step S5, which involves determining the performance of the energy storage power station in improving static voltage stability based on the control results obtained in step S4, specifically includes the following steps:

[0064] Based on the control results obtained in step S4, data information of the target power system before and after control is acquired.

[0065] The absolute improvement in static voltage stability is calculated using the following formula:

[0066] In the formula This represents the increase in active power margin. This refers to the active power margin after control. This refers to the active power margin before control is implemented. This represents the reduction in voltage deviation rate. The voltage deviation rate before control; The controlled voltage deviation rate;

[0067] The relative improvement rate of static voltage stability index is calculated using the following formula:

[0068] In the formula The relative improvement rate of active power margin; The relative improvement rate of voltage deviation rate;

[0069] The energy storage input-output ratio is calculated using the following formula:

[0070] In the formula The overall benefits brought about by improved voltage stability; The energy consumption cost of the energy storage regulation process;

[0071] Performance evaluation of energy storage power stations to improve static voltage stability:

[0072] Power grid dispatch support: or In areas where energy storage capacity is increased;

[0073] Energy storage configuration optimization: During the planning phase, optimize the energy storage capacity and power configuration to ensure... Greater than the set value; and according to Determine the lower limit of the reactive power regulation capacity of the energy storage system;

[0074] Operational performance evaluation: Monitoring the indicators of each energy storage system; Energy storage systems that have been below the set value for several consecutive months, or For energy storage systems that are below the set value, maintenance or optimization should be carried out.

[0075] This invention also provides a system for determining the performance of the method for improving the static voltage stability index of the energy storage power station, comprising a data acquisition module, a data processing module, an index mapping module, a system control module, and a performance determination module; the data acquisition module, data processing module, index mapping module, system control module, and performance determination module are connected in series; the data acquisition module acquires data information of the target power system and the energy storage power station and uploads the data information to the data processing module; the data processing module preprocesses the acquired data information based on the received data information, and uses a causal decomposition method to screen out the key controllable variables of energy storage, and uploads the data information to the index mapping module; the index mapping module constructs and trains a mapping model based on a convolutional network model based on the received data information to obtain the mapping relationship between energy storage regulation quantity and index improvement quantity, and uploads the data information to the system control module; the system control module classifies the state of the target power system based on the obtained mapping model based on the received data information, performs corresponding control, and uploads the data information to the performance determination module; the performance determination module determines the performance of the energy storage power station in improving the static voltage stability index based on the received data information and the obtained control results.

[0076] The method and system for judging the performance of improving static voltage stability index of energy storage power station provided by the present invention acquires, filters and maps data of target power system and energy storage power station, and monitors the control effect of energy storage system. It not only realizes the performance judgment of improving static voltage stability index of energy storage power station, but also has higher reliability and better accuracy. Attached Figure Description

[0077] Figure 1 This is a schematic diagram of the method flow of the present invention.

[0078] Figure 2 This is a schematic diagram of the functional modules of the system of the present invention. Detailed Implementation

[0079] like Figure 1 The diagram shown is a flowchart of the method of the present invention: The performance evaluation method for improving the static voltage stability index of an energy storage power station disclosed in this invention includes the following steps:

[0080] S1. Obtain data information from the target power system and energy storage power station; specifically including the following steps:

[0081] Under the given scenario, acquire the following data from the target power system and energy storage power station:

[0082] Obtain the voltage amplitude at each node. Phase angles of each node Line active power Line reactive power Real-time load demand Wind power output data Photovoltaic power output data And energy storage system data;

[0083] The energy storage system data includes the energy storage system's state of charge. Active charging and discharging power Reactive charging and discharging power and the delay time from receiving the regulation command to the power response ;

[0084] The scenarios set include scenarios where wind power output decreases by a set amount within a set time period (e.g., wind power output decreases by 20% within a set very short time period) and scenarios where load increases by a set amount (e.g., load increases by 15%).

[0085] S2. Preprocess the data obtained in step S1 and use the causal decomposition method to screen out the key variables that can be controlled in energy storage;

[0086] Preprocessing specifically includes the following steps:

[0087] Based on the data obtained in step S1, construct the basic dataset. for ; For the total number of data, For the i-th set of input data, including voltage amplitude Phase angle , , , , , , and ; The output data for the i-th group includes active power margin. and voltage deviation rate ;

[0088] For the basic dataset An enhanced sample set is generated using a classification-targeted perturbation method. , represented as M represents the total number of data points in the augmented sample set; among them, wind power output data is calculated according to the formula. By perturbing the wind, the wind power output can be enhanced. , This is the wind power output disturbance coefficient. and , The set standard deviation, The maximum fluctuation coefficient of wind power output is set, and it is stipulated that... , For wind power output, The rated power of the wind power is used; the reactive power charging and discharging power data are calculated according to the formula. By perturbing, the reactive power charging and discharging power is enhanced. , The set number of steps for the reactive power disturbance of energy storage. The step value of reactive power disturbance of energy storage and , For the rated reactive power regulation capacity of energy storage, To improve energy storage regulation accuracy; Replace the base dataset In At the same time Replace the base dataset In To obtain enhanced input data Based on enhanced input data The Newton-Raphson power flow calculation scheme was used to calculate the enhanced output data. for , To calculate the enhanced active power margin, The calculated enhanced voltage deviation rate;

[0089] For enhanced output data Remove or After obtaining the samples, they are then compared with the base dataset. By merging, we obtain the total sample set. for , The input data is for the total sample set. This is the output data for the total sample set;

[0090] For the total sample set Remove Or samples where the line power exceeds the set value. This is the set minimum node voltage. The maximum node voltage is set; then, 3 is used. The criteria and DBSCAN clustering algorithm are used to identify and remove outlier samples; calculations are performed. rate of change, excluding The sample whose rate of change is greater than the maximum charge and discharge rate of the energy storage system;

[0091] For proportional features, an interval scaling method is used to map them to... The proportional features mentioned in the interval specifically include line active power, line reactive power, wind power output data, photovoltaic power output data, energy storage active charging power, energy storage active discharging power, energy storage reactive charging power, and energy storage reactive discharging power; for numerical features, a Z-score standardization scheme is used for processing; for angular features, a... The process is handled using a periodic standardization method;

[0092] Finally, before supplementing The statistical characteristics and difference characteristics of the sliding window at each time step are used to obtain the preprocessed sample set. for ,in To preprocess the input data of the sample set, The output data for the preprocessed sample set, where T is the total number of data in the preprocessed sample set; the supplementary data... The sliding window statistical characteristics and difference characteristics at each time point include the following steps:

[0093] Construction of sliding window statistical features: For each input variable, calculate the mean, variance, maximum and minimum values ​​over the first k time steps to construct sliding window statistical features; sliding window statistical features are used to reflect the overall distribution characteristics of the variable within the time window;

[0094] Construction of difference features: Calculate the difference between the current time and the previous time for each input variable, as well as the difference between the current time and the mean of the previous k time steps, to construct difference features; difference features are used to reflect the dynamic changing trend of variables;

[0095] Feature Supplementation: The constructed sliding window statistical features and difference features are concatenated with the original input variables to form preprocessed input data with expanded dimensions. ;

[0096] Step S2, which involves using causal decomposition to screen key variables for energy storage control, specifically includes the following steps:

[0097] Based on the obtained preprocessed sample set ,calculate Input variables and output vectors The Pearson correlation coefficient was used to select input variables with a Pearson correlation coefficient greater than a set value to form a candidate variable set. ;right Multicollinearity was tested, and the variance inflation factor (VIF) was calculated. Variables with VIF values ​​greater than a set threshold were removed to obtain a set of intermediate variables without significant multicollinearity. ;from Through screening, variables that can be actively controlled by energy storage and key controllable variables on the system side are obtained, resulting in a set of key variables. The process involves screening out variables that can be actively controlled in energy storage and key controllable variables on the system side, specifically including the following steps:

[0098] Screening of active regulation variables for energy storage: Screening variables that the energy storage system can directly adjust, including active charging power, active discharging power, reactive charging power, reactive discharging power, and state of charge.

[0099] Screening of key controllable variables on the system side: Screening variables that have a greater impact on voltage stability indicators than the set value and can be indirectly controlled through scheduling means, including node voltage reference values ​​and line power flow distribution coefficients;

[0100] A causal graph model of variables is constructed using the causal decomposition method, and conditional independence is tested using the PC algorithm. Variables related to energy storage systems The causal strength is determined by selecting variables whose causal strength is greater than a set value.

[0101] Define the potential coefficient for improvement for , for The changes in variables related to energy storage systems. For the corresponding Changes; Calculation of preprocessed sample sets Each input variable in Value, select The top few groups of input variables with the highest values;

[0102] Finally, variables with causal strength greater than a set value (preferably 0.5) obtained through conditional independence testing using the PC algorithm are compared with the improvement potential coefficient. The intersection of the top 60% of the highest-valued input variables is used to obtain the key variables for energy storage control.

[0103] Step S3, which involves constructing and training a mapping model based on a convolutional network model to obtain the mapping relationship between energy storage regulation and index improvement, specifically includes the following steps:

[0104] The key variables for energy storage controllability obtained in step S2 are used as input variables. ;

[0105] A mapping model is constructed using an inverse feature enhancement temporal convolutional network model: the model includes three causal convolutional modules, a normalization module, and an activation function module;

[0106] In the three causal convolutional modules, the output of the l-th causal convolutional module is represented as

[0107] In the formula This is the output of the l-th causal convolutional module; This describes the processing procedure of the causal convolution module. This is the output of the module above, and it is set... ; The kernel size; The expansion rate is satisfied, and the receptive field is satisfied. , This represents the receptive field of the l-th module;

[0108] The output of the last causal convolutional module is standardized using the following formula:

[0109] In the formula This is the standardized output; This is the output of the last causal convolutional module; This is the first parameter to be learned; This is the second parameter to be learned; This is the batch average. This represents the batch variance. This is a set minimum value to prevent the denominator from being 0;

[0110] Constructing the node-adjustment sensitivity matrix The element in the i-th row and j-th column The calculation formula is , Let be the reactive power regulation of the j-th energy storage unit;

[0111] The input variables are calculated using the following formula. Sensitivity coefficients of each variable:

[0112] In the formula Let be the sensitivity coefficient of the j-th input variable; This represents the change in the static voltage stability index. Let j be the j-th input variable;

[0113] Training the constructed mapping model: using a preprocessed sample set The mapping model is trained; the output of the mapping model is the energy storage regulation amount. , represented as During training, the weighted MSE loss function is used, and the Adam optimizer is employed for training.

[0114] Step S4, based on the mapping model obtained in step S3, divides the state of the target power system and performs corresponding control, specifically including the following steps:

[0115] Based on the input of the mapping model obtained in step S3, obtain the data information of the target power system at the current moment;

[0116] The acquired data is input into the mapping model obtained in step S3 to obtain the benchmark value of the voltage stability index of the target power system at the current moment, expressed as:

[0117] In the formula This is the baseline value for active power margin; This represents the initial active power margin without energy storage regulation. The active power margin improvement is the amount corresponding to the baseline adjustment output by the mapping model. This is the reference value for voltage deviation rate; This represents the initial voltage deviation rate without energy storage regulation. The voltage deviation rate improvement corresponding to the baseline adjustment output of the mapping model;

[0118] Based on the obtained voltage stability index benchmark value, the state of the target power system is classified as follows:

[0119] like and If so, the target power system is determined to be operating in a safe zone;

[0120] like or If so, the target power system is determined to be operating in the warning zone;

[0121] like or If so, the target power system is determined to be operating in the critical region;

[0122] Based on the state division results of the target power system, corresponding control measures are implemented:

[0123] Controls located in the safe zone:

[0124] Maintain the current control method unchanged; if The rate of change is lower than the set value or If the rate of change is higher than the set value, the control will switch to the control in the warning zone;

[0125] Control measures located in the warning zone:

[0126] The following formula is used for calculation:

[0127] In the formula Let be the energy storage reactive power control adjustment amount at time t; This is the rated reactive power capacity for energy storage; It is a symbolic function; The target reactive power regulation; Let be the energy storage active power control adjustment at time t; This is the rated active capacity of the energy storage. The target active power adjustment;

[0128] Real-time calculation and And perform corresponding control on the energy storage system;

[0129] Control in the critical region:

[0130] The energy storage system's regulation output is maximized, and the reactive power regulation is determined based on the direction of the deviation between the node voltage and the rated voltage; until... and The conditions for the warning zone are met, and the area within the warning zone is placed under control.

[0131] Step S5, which involves determining the performance of the energy storage power station in improving static voltage stability based on the control results obtained in step S4, specifically includes the following steps:

[0132] Based on the control results obtained in step S4, data information of the target power system before and after control is acquired.

[0133] The absolute improvement in static voltage stability is calculated using the following formula:

[0134] In the formula This represents the increase in active power margin. This refers to the active power margin after control. This refers to the active power margin before control is implemented. This represents the reduction in voltage deviation rate. The voltage deviation rate before control; The controlled voltage deviation rate;

[0135] The relative improvement rate of static voltage stability index is calculated using the following formula:

[0136] In the formula The relative improvement rate of active power margin; The relative improvement rate of voltage deviation rate; when or When the value approaches 0, the industry benchmark value of the corresponding indicator is used as the denominator to ensure the validity of the calculation results;

[0137] The energy storage input-output ratio is calculated using the following formula:

[0138] In the formula The overall benefits of improved voltage stability include avoiding power outages caused by voltage instability and improving industrial production efficiency due to voltage quality compliance (calculated based on a 0.5% increase in production efficiency for every 1% reduction in voltage deviation rate). The energy consumption cost of the energy storage regulation process is calculated as the product of the energy loss during energy storage charging and discharging and the electricity price; This indicates that the benefits of energy storage regulation outweigh the costs, making it economically feasible.

[0139] Performance evaluation of energy storage power stations to improve static voltage stability:

[0140] Power grid dispatch support: or In areas where energy storage capacity is increased;

[0141] Energy storage configuration optimization: During the planning phase, optimize the energy storage capacity and power configuration to ensure... Greater than the set value (preferably 1.2); and simultaneously according to Determine the lower limit of the reactive power regulation capacity of the energy storage system;

[0142] Operational performance evaluation: Monitoring the indicators of each energy storage system; An energy storage system that has been below a set value (preferably 15%) for several consecutive months (preferably 3 months), or For energy storage systems that are below the set value (preferably 1), maintenance or optimization should be carried out.

[0143] The key variable selection of this invention is more precise. It quantifies and identifies core factors such as SOC and voltage support sensitivity through causal decomposition, eliminating reliance on experience and significantly improving the efficiency of indicator improvement, effectively avoiding the waste of adjustment resources. The mapping accuracy of the model in this invention is higher. It integrates energy storage voltage support sensitivity to construct an inverse feature-enhanced temporal convolutional network, which can accurately capture nonlinear temporal correlations. Under extreme conditions, the prediction error is much lower than that of traditional methods, providing a reliable basis for control strategies. The control strategy of this invention is more adaptable. It performs hierarchical control according to system safety, early warning, and critical state, avoiding the problems of over-adjustment or under-adjustment caused by "one-size-fits-all" approach, and achieving precise control and stability. The effectiveness quantification of this invention is more comprehensive. In addition to technical indicators, it adds the calculation of energy storage input-output ratio, taking into account both technical effects and economic feasibility, providing full-dimensional data support for grid dispatch and energy storage configuration optimization, and is especially suitable for high-proportion new energy distribution networks.

[0144] like Figure 2 The diagram shows the functional modules of the system of the present invention: The system disclosed in this invention for determining the performance of the method for improving the static voltage stability index of the energy storage power station includes a data acquisition module, a data processing module, an index mapping module, a system control module, and a performance determination module; the data acquisition module, data processing module, index mapping module, system control module, and performance determination module are connected in series; the data acquisition module is used to acquire data information of the target power system and the energy storage power station, and upload the data information to the data processing module; the data processing module is used to preprocess the acquired data information according to the received data information, and filter out the controllable key variables of energy storage using the causal decomposition method, and upload the data information to the index mapping module; the index mapping module is used to construct and train a mapping model based on a convolutional network model according to the received data information, obtain the mapping relationship between energy storage adjustment quantity and index improvement quantity, and upload the data information to the system control module; the system control module is used to classify the state of the target power system according to the received data information and the obtained mapping model, and perform corresponding control, and upload the data information to the performance determination module; the performance determination module is used to complete the performance determination of the energy storage power station's improvement of the static voltage stability index according to the received data information and the obtained control results.

Claims

1. A method for determining the performance of an energy storage power station in improving static voltage stability, comprising the following steps: S1. Obtain data information about the target power system and energy storage power station; S2. Preprocess the data obtained in step S1 and use the causal decomposition method to screen out the key variables that can be controlled in energy storage; S3. Based on the convolutional network model, construct and train a mapping model to obtain the mapping relationship between energy storage regulation amount and index improvement amount; S4. Based on the mapping model obtained in step S3, the state of the target power system is divided and corresponding control is performed; S5. Based on the control results obtained in step S4, complete the performance assessment of the energy storage power station to improve the static voltage stability index.

2. The performance evaluation method for improving static voltage stability in an energy storage power station according to claim 1, characterized in that... Step S1, which involves acquiring data information about the target power system and energy storage power station, specifically includes the following steps: Under the given scenario, acquire the following data from the target power system and energy storage power station: Obtain the voltage amplitude at each node. Phase angles of each node Line active power Line reactive power Real-time load demand Wind power output data Photovoltaic power output data And energy storage system data; The energy storage system data includes the energy storage system's state of charge. Active charging and discharging power Reactive charging and discharging power and the delay time from receiving the regulation command to the power response ; The scenarios described include a wind power output decrease by a set amount within a set time period and a load surge by a set amount.

3. The performance evaluation method for improving static voltage stability in an energy storage power station according to claim 2, characterized in that... Step S2, which involves preprocessing the data obtained in step S1, specifically includes the following steps: Based on the data obtained in step S1, construct the basic dataset. for ; For the total number of data, For the i-th set of input data, including voltage amplitude Phase angle , , , , , , and ; The output data for the i-th group includes active power margin. and voltage deviation rate ; For the basic dataset An enhanced sample set is generated using a classification-targeted perturbation method. , represented as M represents the total number of data points in the augmented sample set; among them, wind power output data is calculated according to the formula. By perturbing the wind, the wind power output can be enhanced. , This is the wind power output disturbance coefficient. and , The set standard deviation, The maximum fluctuation coefficient of wind power output is set, and it is stipulated that... , For wind power output, The rated power of the wind power is used; the reactive power charging and discharging power data are calculated according to the formula. By perturbing, the reactive power charging and discharging power is enhanced. , The set number of steps for the reactive power disturbance of energy storage. The step value of reactive power disturbance of energy storage and , For the rated reactive power regulation capacity of energy storage, To improve energy storage regulation accuracy; Replace the base dataset In At the same time Replace the base dataset In To obtain enhanced input data Based on enhanced input data The Newton-Raphson power flow calculation scheme was used to calculate the enhanced output data. for , To calculate the enhanced active power margin, The calculated enhanced voltage deviation rate; For enhanced output data Remove or After obtaining the samples, they are then compared with the base dataset. By merging, we obtain the total sample set. for , The input data is for the total sample set. This is the output data for the total sample set; For the total sample set Remove Or samples where the line power exceeds the set value. This is the set minimum node voltage. The maximum node voltage is set; then, 3 is used. The criteria and DBSCAN clustering algorithm are used to identify and remove outlier samples; calculations are performed. rate of change, excluding The sample whose rate of change is greater than the maximum charge and discharge rate of the energy storage system; For proportional features, an interval scaling method is used to map them to... The proportional features mentioned in the interval specifically include line active power, line reactive power, wind power output data, photovoltaic power output data, energy storage active charging power, energy storage active discharging power, energy storage reactive charging power, and energy storage reactive discharging power; for numerical features, a Z-score standardization scheme is used for processing; for angular features, a... The process is handled using a periodic standardization method; Finally, before supplementing The statistical characteristics and difference characteristics of the sliding window at each time step are used to obtain the preprocessed sample set. for ,in To preprocess the input data of the sample set, The output data for the preprocessed sample set, where T is the total number of data in the preprocessed sample set; the supplementary data... The sliding window statistical characteristics and difference characteristics at each time point include the following steps: Construction of sliding window statistical features: For each input variable, calculate the mean, variance, maximum and minimum values ​​over the first k time steps to construct sliding window statistical features; sliding window statistical features are used to reflect the overall distribution characteristics of the variable within the time window; Construction of difference features: Calculate the difference between the current time and the previous time for each input variable, as well as the difference between the current time and the mean of the previous k time steps, to construct difference features; difference features are used to reflect the dynamic changing trend of variables; Feature Supplementation: The constructed sliding window statistical features and difference features are concatenated with the original input variables to form preprocessed input data with expanded dimensions. .

4. The performance evaluation method for improving static voltage stability index of energy storage power stations according to claim 3, characterized in that... Step S2, which involves using causal decomposition to screen key variables for energy storage control, specifically includes the following steps: Based on the obtained preprocessed sample set ,calculate Input variables and output vectors The Pearson correlation coefficient was used to select input variables with a Pearson correlation coefficient greater than a set value to form a candidate variable set. ;right Multicollinearity was tested, and the variance inflation factor (VIF) was calculated. Variables with VIF values ​​greater than a set threshold were removed to obtain a set of intermediate variables without significant multicollinearity. ;from Through screening, variables that can be actively controlled by energy storage and key controllable variables on the system side are obtained, resulting in a set of key variables. The process involves screening out variables that can be actively controlled in energy storage and key controllable variables on the system side, specifically including the following steps: Screening of active regulation variables for energy storage: Screening variables that the energy storage system can directly adjust, including active charging power, active discharging power, reactive charging power, reactive discharging power, and state of charge. Screening of key controllable variables on the system side: Screening variables that have a greater impact on voltage stability indicators than the set value and can be controlled by scheduling means, including node voltage reference values ​​and line power flow distribution coefficients; A causal graph model of variables is constructed using the causal decomposition method, and conditional independence is tested using the PC algorithm. Variables related to energy storage systems The causal strength is determined by selecting variables whose causal strength is greater than a set value. Define the potential coefficient for improvement for , for The changes in variables related to energy storage systems. For the corresponding The amount of change; Computational preprocessing of sample sets Each input variable in Value, select The top few groups of input variables with the highest values; Finally, variables whose causal strength is greater than the set value as determined by the PC algorithm for conditional independence testing will be compared with the improvement potential coefficient. The intersection of the top few input variables with the highest values ​​yields the key variables for energy storage control.

5. The performance evaluation method for improving static voltage stability index of energy storage power stations according to claim 4, characterized in that... Step S3, which involves constructing and training a mapping model based on a convolutional network model to obtain the mapping relationship between energy storage regulation and index improvement, specifically includes the following steps: The key variables for energy storage controllability obtained in step S2 are used as input variables. ; A mapping model is constructed using an inverse feature enhancement temporal convolutional network model: the model includes three causal convolutional modules, a normalization module, and an activation function module; In the three causal convolutional modules, the output of the l-th causal convolutional module is represented as In the formula This is the output of the l-th causal convolutional module; This describes the processing procedure of the causal convolution module. This is the output of the module above, and it is set... ; The kernel size; The expansion rate is satisfied, and the receptive field is satisfied. , This represents the receptive field of the l-th module; The output of the last causal convolutional module is standardized using the following formula: In the formula This is the standardized output; This is the output of the last causal convolutional module; This is the first parameter to be learned; This is the second parameter to be learned; This is the batch average. This represents the batch variance. This is a set minimum value to prevent the denominator from being 0; Constructing the node-adjustment sensitivity matrix The element in the i-th row and j-th column The calculation formula is , Let be the reactive power regulation of the j-th energy storage unit; The input variables are calculated using the following formula. Sensitivity coefficients of each variable: In the formula Let be the sensitivity coefficient of the j-th input variable; This represents the change in the static voltage stability index. Let j be the j-th input variable; Training the constructed mapping model: using a preprocessed sample set The mapping model is trained; the output of the mapping model is the energy storage regulation amount. , represented as During training, the weighted MSE loss function is used, and the Adam optimizer is employed for training.

6. The performance evaluation method for improving static voltage stability index of energy storage power stations according to claim 5, characterized in that... Step S4, based on the mapping model obtained in step S3, divides the state of the target power system and performs corresponding control, specifically including the following steps: Based on the input of the mapping model obtained in step S3, obtain the data information of the target power system at the current moment; The acquired data is input into the mapping model obtained in step S3 to obtain the benchmark value of the voltage stability index of the target power system at the current moment, expressed as: In the formula This is the baseline value for active power margin; This represents the initial active power margin without energy storage regulation. The active power margin improvement is the amount corresponding to the baseline adjustment output by the mapping model. This is the reference value for voltage deviation rate; This represents the initial voltage deviation rate without energy storage regulation. The voltage deviation rate improvement corresponding to the baseline adjustment output of the mapping model; Based on the obtained voltage stability index benchmark value, the state of the target power system is classified as follows: like and If so, the target power system is determined to be operating in a safe zone; like or If so, the target power system is determined to be operating in the warning zone; like or If so, the target power system is determined to be operating in the critical region; Based on the state division results of the target power system, corresponding control measures are implemented: Controls located in the safe zone: Maintain the current control method unchanged; if The rate of change is lower than the set value or If the rate of change is higher than the set value, the control will switch to the control in the warning zone; Control measures located in the warning zone: The following formula is used for calculation: In the formula Let be the energy storage reactive power control adjustment amount at time t; This is the rated reactive power capacity for energy storage; It is a symbolic function; The target reactive power regulation; Let be the energy storage active power control adjustment at time t; This is the rated active capacity of the energy storage. The target active power adjustment; Real-time calculation and And perform corresponding control on the energy storage system; Control in the critical region: The energy storage system's regulation output is maximized, and the reactive power regulation is determined based on the direction of the deviation between the node voltage and the rated voltage; until... and The conditions for the warning zone are met, and the area within the warning zone is placed under control.

7. The performance evaluation method for improving static voltage stability index of energy storage power stations according to claim 6, characterized in that... Step S5, which involves determining the performance of the energy storage power station in improving static voltage stability based on the control results obtained in step S4, specifically includes the following steps: Based on the control results obtained in step S4, data information of the target power system before and after control is acquired. The absolute improvement in static voltage stability is calculated using the following formula: In the formula This represents the increase in active power margin. This refers to the active power margin after control. This refers to the active power margin before control is implemented. This represents the reduction in voltage deviation rate. The voltage deviation rate before control; The controlled voltage deviation rate; The relative improvement rate of static voltage stability index is calculated using the following formula: In the formula The relative improvement rate of active power margin; The relative improvement rate of voltage deviation rate; The energy storage input-output ratio is calculated using the following formula: In the formula The overall benefits brought about by improved voltage stability; The energy consumption cost of the energy storage regulation process; Performance evaluation of energy storage power stations to improve static voltage stability: Power grid dispatch support: or In areas where energy storage capacity is increased; Energy storage configuration optimization: During the planning phase, optimize the energy storage capacity and power configuration to ensure... Greater than the set value; and according to Determine the lower limit of the reactive power regulation capacity of the energy storage system; Operational performance evaluation: Monitoring the indicators of each energy storage system; Energy storage systems that have been below the set value for several consecutive months, or For energy storage systems that are below the set value, maintenance or optimization should be carried out.

8. A system for determining the performance of an energy storage power station that improves static voltage stability as described in any one of claims 1 to 7, characterized in that... It includes a data acquisition module, a data processing module, an indicator mapping module, a system control module, and a performance determination module; the data acquisition module, data processing module, indicator mapping module, system control module, and performance determination module are connected in series; the data acquisition module is used to acquire data information from the target power system and energy storage power station, and upload the data information to the data processing module; The data processing module is used to preprocess the acquired data information based on the received data information, and to screen out the key controllable variables of energy storage by combining the causal decomposition method, and then upload the data information to the indicator mapping module. The index mapping module is used to construct and train a mapping model based on the received data information and a convolutional network model to obtain the mapping relationship between energy storage regulation amount and index improvement amount, and upload the data information to the system control module. The system control module is used to classify the state of the target power system based on the received data information and the obtained mapping model, and to perform corresponding control, and then upload the data information to the performance judgment module. The performance determination module is used to determine the performance of the energy storage power station in improving static voltage stability based on the received data and the obtained control results.

Images