Virtual power plant scheduling system and method

Abstract

The invention discloses a virtual power plant dispatching system and method, and particularly relates to the technical field of power dispatching. By monitoring the output characteristics of the distributed energy, whether nonlinear interference is generated or not is judged. And further analyzing the disturbance of the distributed energy on the power flow inversion path, judging the abnormal change of the coupling correlation degree between potential hidden resources, and evaluating the influence on the scheduling precision. And in combination with a node differentiation response mode of a distributed energy storage aggregation region and a flexible load community, analyzing the influence of a local grouping cooperation effect on the overall output boundary identification precision of the virtual power plant. Potential erosion of scheduling errors caused by demand side multi-dimensional elastic cooperation to scheduling precision is evaluated by analyzing spectral domain characteristics of demand side elastic gradients and transient disturbance under different spatial and temporal scales. Through comprehensive analysis, the internal regulation and control scheme of the virtual power plant is adjusted, and the overall scheduling efficiency and stability of the virtual power plant are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electric power dispatching, and more specifically, to a virtual power plant dispatching system and method. Background Art

[0002] As an important platform for integrating distributed energy resources and optimizing scheduling, virtual power plants can achieve efficient utilization and flexible regulation of energy. However, virtual power plants currently face many challenges in actual operation, especially under high-frequency power market price fluctuations and transient fluctuations in distributed energy output, which can easily seriously affect scheduling accuracy and system stability. Traditional scheduling methods often rely on linear models, which are difficult to effectively deal with complex nonlinear interference and changes in coupling correlation between implicit resources, resulting in significant errors and deviations in the execution of scheduling instructions. In addition, the dynamic changes of the multi-dimensional synergistic effects of demand-side elastic response at different time and space scales further aggravate the potential erosion of scheduling errors and restrict the optimization capabilities of virtual power plants in highly dynamic environments.

[0003] In order to solve the above problems, a virtual power plant scheduling system and method are provided. Summary of the invention

[0004] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a virtual power plant scheduling system and method to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A virtual power plant scheduling method comprises the following steps:

[0007] By monitoring the output characteristics of distributed energy resources connected to virtual power plants under the disturbance of high-frequency power market price signals, it is determined whether the transient fluctuations of distributed energy resources in the short-cycle scheduling process will cause nonlinear interference in the real-time identification of the comprehensive control strategy;

[0008] When the transient fluctuations of distributed energy resources in the short-cycle dispatch process produce nonlinear interference in the real-time identification of the comprehensive control strategy, by analyzing the disturbance of the distributed energy within the virtual power plant to the power flow inversion path, it is determined whether there is an abnormality in the impact of abnormal changes in the coupling correlation between potential hidden resources on the dispatch accuracy;

[0009] When there is an abnormal impact of abnormal changes in the coupling correlation between potential hidden resources on the dispatching accuracy: through fine-grained analysis of the node-differentiated response patterns between the distributed energy storage aggregation area and the flexible load community, the influence of the local group coordination effect on the overall output boundary identification accuracy of the virtual power plant is evaluated;

[0010] By analyzing the spectral characteristics of demand-side elasticity gradients and transient disturbances at different time and space scales, the potential erosion of dispatching error caused by multi-dimensional elasticity coordination on the dispatching instruction execution accuracy is evaluated.

[0011] By comprehensively analyzing the influence of local group coordination effect on the overall output boundary identification accuracy of the virtual power plant and the impact of scheduling errors caused by multi-dimensional elastic coordination on the demand side on the execution accuracy of scheduling instructions, the internal control plan of the virtual power plant is adjusted.

[0012] In a preferred embodiment, by monitoring the output characteristics of distributed energy resources connected to the virtual power plant under the disturbance of high-frequency power market price signals, it is determined whether the transient fluctuations of distributed energy resources in the short-cycle scheduling process produce nonlinear interference in the real-time identification of the comprehensive control strategy, specifically:

[0013] Monitor the distributed energy resources connected to the virtual power plant and record the changing characteristics of output power under the influence of high-frequency power market price fluctuations;

[0014] During short-cycle scheduling, the output fluctuations of distributed energy are monitored in real time to identify the amplitude, frequency and time characteristics of transient fluctuations;

[0015] Analyze the relationship between transient fluctuations and real-time identification of virtual power plant dispatch strategies, and evaluate whether there is nonlinear interference: A multivariate regression model is used to analyze the relationship between transient fluctuations and real-time identification of dispatch strategies. The mathematical form is:

[0016] Among them, R t represents the real-time identification result of the scheduling strategy; γ0 represents the constant term of the regression model; γ i represents the regression coefficient of transient fluctuation amplitude; γ′ represents the regression coefficient of transient fluctuation frequency; γ i ″ represents the regression coefficient of transient fluctuation duration; δ t represents the model error term; N represents the number of distributed energy resources;

[0017] The multivariate regression model is trained with the training data set to evaluate the impact of transient fluctuation characteristics on the identification of scheduling strategies; nonlinear terms are introduced to capture nonlinear interference:

[0018]

[0019] Among them, γ i,j A represents the nonlinear interaction regression coefficient between transient fluctuation amplitude and transient fluctuation frequency; i,t Represents distributed energy E i The transient fluctuation amplitude at time t; F j,t Represents distributed energy E jThe transient fluctuation frequency at time t; T i,t Represents distributed energy E i Duration of transient fluctuation at time t;

[0020] By testing the nonlinear interaction coefficient γ in the regression model i,j The statistical significance of is used to determine whether there is nonlinear interference:

[0021] For the nonlinear interaction coefficient γ i,j Perform statistical tests when the nonlinear interaction coefficient γ i,j When the value of is greater than the critical value of the statistical test, the nonlinear interaction coefficient γ i,j If the coefficient of the nonlinear interaction term is significantly not zero, it is determined that the transient fluctuation of distributed energy resources has produced nonlinear interference in the real-time identification of the comprehensive control strategy.

[0022] In a preferred embodiment, when the transient fluctuation of distributed energy resources in the short-cycle scheduling process produces nonlinear interference in the real-time identification of the comprehensive control strategy, by analyzing the disturbance of the distributed energy in the virtual power plant to the power flow inversion path, it is judged whether there is an abnormality in the influence of abnormal changes in the coupling correlation between potential hidden resources on the scheduling accuracy, specifically:

[0023] Extract disturbance parameters of the distributed energy power flow inversion path within the virtual power plant;

[0024] Analyze the changing rules of coupling correlation between potential hidden resources;

[0025] Determine whether the abnormal coupling correlation has an abnormal effect on the dispatching accuracy: establish the following judgment formula: D(m,n)=R(m,n)*S(m,n)-Q(m,n); where D(m,n) is the abnormality judgment index; R(m,n) is the correlation change parameter; S(m,n) represents the dispatching accuracy sensitivity parameter; Q(m,n) represents the dispatching accuracy error correction parameter;

[0026] If D(m,n) is greater than 0, it means that the abnormal change of coupling correlation between potential implicit resources has an abnormal impact on scheduling accuracy;

[0027] If D(m,n) is less than or equal to 0, it means that there is no abnormal impact of abnormal changes in the coupling correlation between potential implicit resources on scheduling accuracy.

[0028] In a preferred embodiment, when the abnormal change of coupling correlation between potential hidden resources has an abnormal impact on the dispatching accuracy: by performing a fine-grained analysis on the node differentiation response mode between the distributed energy storage aggregation area and the flexible load group, the influence of the local group coordination effect on the overall output boundary identification accuracy of the virtual power plant is evaluated, specifically:

[0029] Collect the real-time power, frequency and voltage responses of each node in the distributed energy storage aggregation area and flexible load group;

[0030] Extracting dynamic response characteristics between nodes from the collected data, including power offset rate, time delay and voltage disturbance;

[0031] Based on the dynamic response characteristics between nodes, a fine-grained response model of distributed energy storage and flexible load communities is constructed to quantify the differentiation mode;

[0032] Combined with fine-grained models, the coordinated response of distributed energy storage aggregation areas and flexible load groups is evaluated;

[0033] Comprehensive group coordination effect, analyze the impact on the recognition accuracy of the overall output boundary of the virtual power plant: Definition: Among them, RAI is the recognition accuracy index; H syg (c) represents the quantitative value of the coordination response effect at time point c; σ op (c) represents the standard deviation of the overall output of the virtual power plant at time point c.

[0034] In a preferred embodiment, by analyzing the spectral characteristics of demand-side elastic gradient and transient disturbance at different time and space scales, the potential erosion degree of scheduling error caused by multi-dimensional elastic coordination on the scheduling instruction execution accuracy is evaluated, specifically:

[0035] Collect demand-side elasticity data and transient disturbance characteristics at different time and space scales;

[0036] Partition the demand-side elasticity data and construct an elasticity gradient model at spatiotemporal scales;

[0037] Extract the spectral characteristics of transient disturbances and quantify the dynamic change trend of disturbances;

[0038] Establish a collaborative model of elastic gradient and disturbance characteristics to analyze the nonlinear impact of collaborative effects on scheduling;

[0039] Calculate the scheduling error caused by synergy and assess the potential erosion of scheduling accuracy: Defining Scheduling Error Among them, E d represents the scheduling error; f1 and f2 represent the lower and upper limits of the frequency range involved in the spectrum analysis respectively; W f Represents the weight coefficient at a given frequency f; G p,q,f is the synergy index;

[0040] Define the scheduling accuracy erosion index Among them, SAEI is the scheduling accuracy erosion index; E d represents the scheduling error; T dIndicates the scheduling accuracy tolerance threshold.

[0041] In a preferred embodiment, by comprehensively analyzing the influence of the local group coordination effect on the overall output boundary recognition accuracy of the virtual power plant and the influence of the dispatch error caused by the multi-dimensional elastic coordination on the demand side on the dispatch instruction execution accuracy, the internal control scheme of the virtual power plant is adjusted, specifically:

[0042] The recognition accuracy index corresponding to the degree of influence of the local group coordination effect on the overall output boundary recognition accuracy of the virtual power plant and the dispatch accuracy erosion index corresponding to the influence of the dispatch error caused by the multi-dimensional elastic coordination on the dispatch instruction execution accuracy are normalized, and the normalized recognition accuracy index and dispatch accuracy erosion index are calculated to obtain the comprehensive performance adjustment coefficient. The calculation formula is: Among them, CPA represents the comprehensive performance adjustment coefficient; SAEI is the scheduling accuracy erosion index; RAI is the recognition accuracy index; μ is the adjustment coefficient;

[0043] A comprehensive performance adjustment coefficient threshold is preset, and the comprehensive performance adjustment coefficient is compared with the comprehensive performance adjustment coefficient threshold:

[0044] When the comprehensive performance adjustment coefficient is greater than or equal to the comprehensive performance adjustment coefficient threshold, the virtual power plant dispatching system needs to adjust the internal control plan;

[0045] When the comprehensive performance adjustment coefficient is greater than or equal to the comprehensive performance adjustment coefficient threshold, the virtual power plant dispatching system does not need to adjust the internal control plan.

[0046] On the other hand, the present invention provides a virtual power plant dispatching system, including an energy monitoring module, a disturbance analysis module, an identification accuracy assessment module, an erosion degree assessment module, and a control module;

[0047] Energy monitoring module: By monitoring the output characteristics of distributed energy resources connected to the virtual power plant under the disturbance of high-frequency power market price signals, it is determined whether the transient fluctuations of distributed energy resources in the short-cycle scheduling process will cause nonlinear interference to the real-time identification of the comprehensive control strategy;

[0048] Disturbance analysis module: When the transient fluctuations of distributed energy resources in the short-cycle scheduling process produce nonlinear interference on the real-time identification of the comprehensive control strategy, by analyzing the disturbance of the distributed energy in the virtual power plant to the power flow inversion path, it is determined whether there is an abnormality in the impact of abnormal changes in the coupling correlation between potential hidden resources on the scheduling accuracy;

[0049] Identification accuracy assessment module: When the abnormal change of coupling correlation between potential hidden resources has an abnormal impact on the dispatching accuracy: through fine-grained analysis of the node differentiation response mode between the distributed energy storage aggregation area and the flexible load community, the influence of the local group coordination effect on the overall output boundary identification accuracy of the virtual power plant is evaluated;

[0050] Erosion degree assessment module: By analyzing the spectral characteristics of demand-side elasticity gradient and transient disturbance at different time and space scales, the potential erosion degree of scheduling error caused by multi-dimensional elasticity coordination on the execution accuracy of scheduling instructions is assessed;

[0051] Control module: Adjust the internal control plan of the virtual power plant by conducting a comprehensive analysis of the impact of local group coordination effect on the overall output boundary identification accuracy of the virtual power plant and the impact of the dispatch error caused by multi-dimensional elastic coordination on the demand side on the execution accuracy of dispatch instructions.

[0052] Technical effects and advantages of a virtual power plant dispatching system and method of the present invention:

[0053] 1. By real-time monitoring of the output characteristics of distributed energy under high-frequency electricity market price fluctuations, the nonlinear interference of transient fluctuations on the dispatching strategy can be effectively judged, thereby reducing the negative impact of price fluctuations on the overall dispatching accuracy of the virtual power plant. When nonlinear interference occurs, the abnormal changes in the coupling correlation between potential hidden resources can be accurately evaluated through the disturbance analysis of the distributed energy flow inversion path within the virtual power plant, further improving the dispatching accuracy. Fine-grained analysis of the local group coordination effect helps to optimize the overall output boundary identification accuracy of the virtual power plant, thereby reducing the dispatching deviation caused by local resource mismatch.

[0054] 2. Through comprehensive analysis of the dispatching errors of multi-dimensional elastic coordination on the demand side, it is possible to effectively identify potential dispatching accuracy erosion, adjust the dispatching plan in a timely manner, improve the dispatching accuracy and stability of virtual power plants in high-frequency disturbance environments, and achieve more efficient and reliable energy management, so that virtual power plants can better cope with complex electricity market dynamics and provide important technical support for the stable operation of future smart grids and energy optimization management. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A schematic diagram of a virtual power plant scheduling method of the present invention;

[0056] Figure 2 It is a structural schematic diagram of a virtual power plant dispatching system of the present invention. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] Example 1

[0059] Figure 1 The present invention provides a virtual power plant scheduling method, which includes the following steps:

[0060] By monitoring the output characteristics of distributed energy resources connected to virtual power plants under the disturbance of high-frequency power market price signals, it is determined whether the transient fluctuations of distributed energy resources in the short-cycle scheduling process will cause nonlinear interference in the real-time identification of the comprehensive control strategy;

[0061] When the transient fluctuations of distributed energy resources in the short-cycle dispatch process produce nonlinear interference in the real-time identification of the comprehensive control strategy, by analyzing the disturbance of the distributed energy within the virtual power plant to the power flow inversion path, it is determined whether there is an abnormality in the impact of abnormal changes in the coupling correlation between potential hidden resources on the dispatch accuracy;

[0062] When there is an abnormal impact of abnormal changes in the coupling correlation between potential hidden resources on the dispatching accuracy: through fine-grained analysis of the node-differentiated response patterns between the distributed energy storage aggregation area and the flexible load community, the influence of the local group coordination effect on the overall output boundary identification accuracy of the virtual power plant is evaluated;

[0063] By analyzing the spectral characteristics of demand-side elasticity gradients and transient disturbances at different time and space scales, the potential erosion of dispatching error caused by multi-dimensional elasticity coordination on the dispatching instruction execution accuracy is evaluated.

[0064] By comprehensively analyzing the influence of local group coordination effect on the overall output boundary identification accuracy of the virtual power plant and the impact of scheduling errors caused by multi-dimensional elastic coordination on the demand side on the execution accuracy of scheduling instructions, the internal optimization and control plan of the virtual power plant is adjusted.

[0065] Specifically, by monitoring the output characteristics of distributed energy resources connected to the virtual power plant under the disturbance of high-frequency power market price signals, it is determined whether the transient fluctuations of distributed energy resources in the short-cycle scheduling process will cause nonlinear interference to the real-time identification of the comprehensive control strategy, including:

[0066] Monitor the distributed energy connected to the virtual power plant and record the changing characteristics of the output power under the influence of high-frequency power market price fluctuations: The virtual power plant dispatching system first obtains the high-frequency power market price signal through the power market interface module. The power market interface module is connected to the power market trading platform and receives the price signal data in real time. The power market price signal is expressed in the form of a time series and is defined as: P t , where P represents the electricity market price and t represents the time point. The electricity market interface module ensures that the price signal P at each time point t are accurately recorded and sampled at regular time intervals.

[0067] The output power of each distributed energy source (such as photovoltaic power generation, wind power generation, energy storage system, etc.) in the virtual power plant is monitored in real time through the monitoring module. i The output power at time point t is defined as: i,t , where O represents the output power, i represents the i-th distributed energy source, and t represents the time point. The monitoring module ensures that the output power of all distributed energy sources is O i,t It is recorded in real time and stored in the distributed energy database.

[0068] The dispatching system uses the correlation analysis module to convert the high-frequency power market price signal P obtained by the power market interface module into t The distributed energy output power O recorded by the monitoring module i,t Conduct correlation analysis. Correlation analysis aims to identify the impact of price signal changes on distributed energy output power. The specific steps include:

[0069] Data synchronization: ensure the synchronization of price signals and output power at time t to avoid data errors caused by time deviation;

[0070] Change rate calculation: Calculate the change rate ΔP of the price signal and output power between time t and t-1 t , ΔO i,t ;

[0071] Correlation analysis: using statistical methods (such as Pearson correlation coefficient) to calculate the rate of change of price signals ΔP t and the output power change rate ΔO i,t The correlation coefficient between The calculation formula is: Among them, Cov(ΔP,ΔO i ) represents the covariance between the rate of change of the price signal and the rate of change of the output power; σΔP represents the standard deviation of the rate of change of the price signal; σΔO i Indicates the standard deviation of the output power change rate.

[0072] Record the output power change characteristics of each distributed energy source under the disturbance of high-frequency price signals, including parameters such as change amplitude, frequency and time delay.

[0073] In the short-cycle scheduling process, the output fluctuations of distributed energy are monitored in real time to identify the amplitude, frequency and time characteristics of transient fluctuations: a low-pass filter is used to remove the high-frequency noise of the output power of distributed energy. Indicates the output power after filtering.

[0074] Output power after denoising Perform transient fluctuation detection. Transient fluctuation refers to power changes that occur in a short period of time and is defined as: in, For distributed energy i The transient fluctuation amplitude at time t; τ is the time window for transient fluctuation detection. Set the threshold θ trs ,when The value is greater than the threshold θ trs When , it is determined that a transient fluctuation is detected.

[0075] For the detected transient fluctuation, extract its amplitude A i,t , frequency F i,t and duration T i,t : The duration of transient fluctuation is defined as the time interval between the start time and the end time of the fluctuation.

[0076] Analyze the relationship between transient fluctuations and real-time identification of virtual power plant dispatch strategies, and evaluate whether there is nonlinear interference: A multivariate regression model is used to analyze the relationship between transient fluctuations and real-time identification of dispatch strategies. Its mathematical form is:

[0077] Among them, R t represents the real-time identification result of the scheduling strategy; γ0 represents the constant term of the regression model; γ i represents the regression coefficient of transient fluctuation amplitude; γ′ represents the regression coefficient of transient fluctuation frequency; γ i ″ represents the regression coefficient of transient fluctuation duration; δ t represents the model error term; N Represents the number of distributed energy resources.

[0078] The multivariate regression model is trained with a training data set (including price signals at different time points and distributed energy output fluctuation characteristics) to evaluate the impact of transient fluctuation characteristics on the identification of dispatch strategies. Nonlinear terms (such as interaction terms) are introduced to capture nonlinear interference:

[0079] Among them, γ i,jA represents the nonlinear interaction regression coefficient between transient fluctuation amplitude and transient fluctuation frequency; i,t Represents distributed energy E i The transient fluctuation amplitude at time t; F j,t Represents distributed energy E j The transient fluctuation frequency at time t; T i,t Represents distributed energy E i The duration of transient fluctuation at time t; N represents the number of distributed energy resources.

[0080] By testing the nonlinear interaction coefficient γ in the regression model i,j The statistical significance of the results can be used to determine whether there is nonlinear interference. The specific determination methods include:

[0081] Significance test: Nonlinear interaction coefficient γ i,j Perform statistical tests when the nonlinear interaction coefficient γ i,j When the value of is greater than the critical value of the statistical test, the nonlinear interaction coefficient γ i,j Significantly not zero.

[0082] Nonlinear interference determination: If the coefficient of the nonlinear interaction term is significantly not zero, it is determined that the transient fluctuation of distributed energy resources has caused nonlinear interference in the real-time identification of the comprehensive control strategy.

[0083] Specifically, when the transient fluctuations of distributed energy resources in the short-cycle scheduling process produce nonlinear interference in the real-time identification of the comprehensive control strategy, by analyzing the disturbance of the distributed energy in the virtual power plant to the power flow inversion path, it is judged whether there is any abnormality in the influence of abnormal changes in the coupling correlation between potential hidden resources on the scheduling accuracy, including:

[0084] Extraction of disturbance parameters of the inversion path of distributed energy power flow in the virtual power plant: In the virtual power plant, different distributed energy sources will be coupled with each other through the power network. This coupling relationship can be characterized by analyzing the inversion path of the power flow. The power flow inversion path refers to a transient change process in which the power flow deviates, reverses or complexly wraps from the original path under a specific topological structure due to the rapid change in the output power of distributed energy.

[0085] When transient fluctuations of distributed energy are detected in the short-term dispatch process and cause nonlinear interference to the identification of the comprehensive control strategy, it is necessary to extract disturbance parameters of the power flow inversion path in the virtual power plant, including:

[0086] The quantified value U(m,n) of the disturbance of the power flow inversion path between the mth node and the nth node in the node pair is used to measure the disturbance intensity of the power flow inversion path between the node m and the node n. The larger the value, the more severe the disturbance.

[0087] The voltage correlation offset parameter V(m,n) from node m to node n is used to characterize the change characteristics of voltage distribution in the power flow inversion path, indicating the small change amplitude of voltage correlation;

[0088] The current phase offset parameter W(m,n) from node m to node n is used to characterize the degree of change of the current phase in the power flow inversion path under disturbance. It can be positive or negative. The value indicates the amplitude of the phase change, and the positive or negative value indicates the direction of the phase offset.

[0089] Among them, m and n are positive integer indexes used to identify different electrical nodes within the virtual power plant.

[0090] After high-frequency sampling and data preprocessing of the internal electrical data of the virtual power plant, the following mathematical relationship can be constructed to further quantify the disturbance parameters: define the comprehensive disturbance parameter X(m,n)=U(m,n)+α*V(m,n)+β*W(m,n): where X(m,n) represents the comprehensive disturbance parameter of the node pair (m,n), which is used to characterize the overall disturbance level of the power flow inversion path under the influence of transient fluctuations; α is a real number parameter used to adjust the contribution of the voltage-related offset parameter in the comprehensive disturbance; β is a real number parameter used to adjust the contribution of the current phase offset parameter in the comprehensive disturbance.

[0091] The larger the value of the comprehensive disturbance parameter X(m,n), the stronger the disturbance to the power flow inversion path between nodes m and n.

[0092] Analyze the changing rules of coupling correlation between potential implicit resources: Implicit resources refer to those resources or nodes that are not directly concerned in daily scheduling but still have an impact on the state of the power grid. There may be complex nonlinear coupling between these implicit resources.

[0093] In order to analyze the coupling correlation, the following parameters are introduced:

[0094] Y(m,n) represents the coupling metric between potential implicit resources between nodes m and n. By comparing historical and current data through statistical methods or data-driven methods, a quantitative evaluation of the degree of coupling is obtained. The larger the value of Y(m,n), the stronger the implicit coupling between nodes m and n.

[0095] Z(m,n) represents the benchmark coupling measurement value between node m and node n under the reference state, which can be obtained through long-term statistical results under no disturbance or slight disturbance conditions; it is used as a comparison benchmark for Y(m,n).

[0096] By analyzing the relationship between X(m,n) and Y(m,n) and Z(m,n), the following formula can be constructed to quantitatively evaluate the change in coupling correlation:

[0097] Define the correlation change parameters Among them, R(m,n) is the correlation change parameter, which is used to quantitatively measure the deviation of the implicit coupling degree Y(m,n) in the current state relative to the benchmark coupling degree Z(m,n) after considering the disturbance X(m,n) of the power flow inversion path. The size and positive and negative of the R(m,n) value will be used to determine whether there is an abnormal change trend in the coupling correlation.

[0098] A coupling correlation threshold is preset. When the coupling correlation is greater than the coupling correlation threshold, it is considered that the coupling correlation has an abnormal change trend.

[0099] Determine whether the abnormal coupling correlation affects the scheduling accuracy abnormally: To determine the impact of abnormal coupling correlation on scheduling accuracy, the following parameters are introduced:

[0100] S(m,n) represents the scheduling accuracy sensitivity parameter, which is calibrated based on the scheduling optimization model and actual test data. The higher the value, the more significant the change in the coupling degree between the node pairs is on the scheduling accuracy.

[0101] Q(m,n) represents the scheduling accuracy error correction parameter, which is used as the strategy correction reference value applied when responding to abnormal changes. It is obtained through historical scheduling results and sensitivity analysis methods. It can be understood as the strategy fine-tuning amount that needs to be applied to maintain scheduling accuracy when abnormal changes occur in the coupling correlation.

[0102] The following judgment formula D(m,n) is established: D(m,n)=R(m,n)*S(m,n)-Q(m,n); wherein D(m,n) is the abnormal judgment index, which is used to comprehensively evaluate the impact of abnormal changes in coupling correlation on scheduling accuracy; R(m,n) is the correlation change parameter, which indicates the degree of coupling deviation; S(m,n) is the scheduling accuracy sensitivity parameter, which indicates the sensitivity of scheduling accuracy to changes in coupling; Q(m,n) is the scheduling accuracy error correction parameter, which is used as a reference value for strategy correction applied when responding to abnormal changes.

[0103] If D(m,n) is greater than 0, it means that the abnormal change of coupling correlation still exceeds the error correction capability after superimposing the scheduling accuracy sensitivity, and the impact of the abnormal change of coupling correlation between potential hidden resources on scheduling accuracy is abnormal.

[0104] If D(m,n) is less than or equal to 0, it means that through comprehensive consideration of the coupling correlation change and scheduling sensitivity, as well as the existence of error correction parameters, the scheduling accuracy can still be maintained within a reasonable range, and there is no abnormal impact of abnormal changes in the coupling correlation between potential hidden resources on the scheduling accuracy.

[0105] Specifically, when the abnormal change of coupling correlation between potential hidden resources has an abnormal impact on the dispatching accuracy: through fine-grained analysis of the node differentiation response mode between distributed energy storage aggregation area and flexible load group, the influence of local group coordination effect on the overall output boundary identification accuracy of virtual power plant is evaluated, including:

[0106] Collect the real-time power, frequency and voltage response of each node in the distributed energy storage aggregation area and flexible load group: Use high-frequency sensors to monitor the power, frequency and voltage of each node in the distributed energy storage aggregation area and flexible load group in real time. Each node collects data through multi-channel sensors to ensure that its timing accuracy can reflect small fluctuations and disturbances.

[0107] Power response: Measure the output power of each node through sensors, record its changes at different time points, and calibrate the power fluctuation range.

[0108] Frequency response: Use frequency sensors to capture power frequency changes at each node.

[0109] Voltage response: Voltage response monitoring ensures that the coordination status of the energy storage device and the flexible load can be reflected, especially under load fluctuations and power output disturbances.

[0110] Extract dynamic response characteristics between nodes from the collected data, including power deviation rate, time delay and voltage disturbance:

[0111] The power deviation rate is used to measure the power variation of a node during a disturbance event. It is defined as follows:

[0112] P of (c) = P ms (c)-P bl (c); where P of (c) represents the power deviation rate at time point c, in kilowatts; P ms (c) represents the actual measured power value at time point c, in kilowatts; P bl (c) represents the reference power value at time point c, in kilowatts.

[0113] Time delay is used to measure the time difference between two nodes in responding to a disturbance. It is defined as follows:

[0114] Δt dl (a,b)=t pe (a)-t pe (b); where Δt dl (a,b) represents the time delay between node a and node b, in seconds; t pe (a) represents the peak time point of the response of node a, in seconds; t pe(b) represents the peak time point of node b’s response, in seconds.

[0115] Voltage disturbance is used to measure the voltage change amplitude of a node during a disturbance event. It is defined as follows:

[0116] V dis (c) = V ms (c)-V bl (c); where V dis (c) represents the voltage disturbance at time point c, in volts; V ms (c) represents the actual measured voltage value at time point c, in volts; V bl (c) represents the reference voltage value at time point c, in volts.

[0117] Through the above calculations, the dynamic response characteristics of each node in the disturbance event, namely power deviation rate, time delay and voltage disturbance, can be extracted. These characteristics are used to quantify the differentiated response patterns between nodes.

[0118] Based on the dynamic response characteristics between nodes, a fine-grained response model of distributed energy storage and flexible load communities is constructed to quantify the differentiated patterns: A fine-grained response model is constructed to describe the differentiated response patterns between nodes in the distributed energy storage and flexible load communities. The model is used to quantify the response characteristics of each node and their relationships. The definitions are as follows:

[0119] M rp (a,b,c)=ρ1*P of (a,t)+ρ2*Δt dl (a,b)+ρ3*V dis (b, t); where M rp (a, b, c) represents the fine-grained response value of node a to node b at time point c; ρ1, ρ2, ρ3 are the weight coefficients of power offset rate, time delay and voltage disturbance respectively; P of (a,t) represents the power offset rate of node a at time point c; Δt dl (a,b) represents the time delay between node a and node b; V dis (b,t) represents the voltage disturbance at node b at time point c.

[0120] The weight coefficients ρ1, ρ2, and ρ3 are estimated through historical data and regression analysis methods to ensure that the fine-grained response model can accurately reflect the impact of each response feature on the overall response.

[0121] Combined with fine-grained models, the coordinated response of distributed energy storage aggregation areas and flexible load groups is evaluated: the coordinated response effect is quantified by summarizing and weighting the response values ​​of the fine-grained models. The definition is as follows: Among them, Hsyg (c) represents the quantitative value of the coordination response effect at time point c; A and B represent the number of nodes in the distributed energy storage aggregation area and the number of nodes in the flexible load group, respectively; θ a represents the weight coefficient of the ath node in the distributed energy storage aggregation area; φ b represents the weight coefficient of the bth node in the flexible load community; M rp (a,b,c) represents the fine-grained response value of node a to node b at time point c.

[0122] Weight coefficient θ a and φ b , which reflects the contribution of each node to the overall coordination response effect. It is determined through historical scheduling data and optimization algorithms (such as principal component analysis or weight optimization method) to ensure that the distribution of weights can truly reflect the importance and influence of each node.

[0123] Through the fine-grained response model M rp (a, b, c) can describe the differentiated response modes between nodes in the distributed energy storage and flexible load community:

[0124] When the fine-grained response value M rp When (a, b, c) is high, it means that node a has a stronger coordination response to node b;

[0125] When the fine-grained response value M rp When (a, b, c) is low, it means that the coordination response of node a to node b is weak.

[0126] By introducing a fine-grained response model, the dynamic interaction between distributed energy storage and each node in the flexible load community can be accurately captured.

[0127] Comprehensive group coordination effect, analyze the impact on the recognition accuracy of the overall output boundary of the virtual power plant: Comprehensive group coordination effect, analyze its impact on the recognition accuracy of the overall output boundary of the virtual power plant, defined as follows:

[0128] Among them, RAI is the recognition accuracy index, which indicates the influence of group coordination effect on the output boundary recognition accuracy; H syg (c) represents the quantitative value of the coordination response effect at time point c; σ op (c) represents the standard deviation of the overall output of the virtual power plant at time point c.

[0129] The larger the identification accuracy index, the greater the impact of the local group coordination effect on the overall output boundary identification accuracy of the virtual power plant, that is, the higher the overall output boundary identification accuracy. The increase in the identification accuracy index indicates that the virtual power plant system can more finely capture the response differences of each distributed energy source and flexible load group in the scheduling, thereby improving the ability to identify dynamic changes in resources. It means that when dealing with complex scheduling tasks, it can more effectively deal with influencing factors such as load fluctuations and energy storage response, ensure that scheduling goals can be accurately executed, and reduce scheduling errors caused by improper local resource coordination. The increase in the identification accuracy index also means that the virtual power plant can better understand and predict the synergy between nodes, especially in a complex scheduling environment, and can promptly identify and adjust factors that may cause power output boundary deviations.

[0130] Specifically, by analyzing the spectral characteristics of demand-side elasticity gradients and transient disturbances at different time and space scales, the potential erosion of the dispatching error caused by the multi-dimensional elasticity synergy on the dispatching instruction execution accuracy is evaluated, including:

[0131] Collect demand-side elasticity data and transient disturbance characteristics at different time and space scales: In the dispatch management of virtual power plants, demand-side elasticity is manifested as the user's ability to dynamically adjust electricity demand, while transient disturbances refer to short-term, unstable fluctuations caused by external factors. In order to analyze the synergistic effect of multi-dimensional elasticity on the demand side, it is first necessary to collect these data, including elasticity characteristics and disturbance characteristics at different time scales (seconds, minutes, hours, etc.) and spatial ranges (regions, nodes, user groups, etc.).

[0132] Among them, D p,h represents the demand side load data of the pth region at time point h; E p,h represents the elastic response of the pth region at time point h, and represents the actual adjustment value of the demand-side user to the external excitation signal; U p,h represents the transient disturbance of the pth region at time point h; the elastic response is calculated as: E p,h =α p *D p,h ; α p is the elasticity coefficient of the pth region, ranging from [0,1], indicating the strength of demand elasticity in the region.

[0133] The intelligent metering terminal is used to collect the demand side load data in real time, and the elastic response is calculated by combining the historical response data to deduce the elastic coefficient. The transient disturbance is recorded by the disturbance monitoring device (such as high-frequency sampler) to reflect the dynamic changes of the demand side affected by external factors.

[0134] The demand-side elasticity data is partitioned and a spatiotemporal elasticity gradient model is constructed: the coverage of the virtual power plant is divided into M regions according to regional characteristics (such as user type, load composition, etc.), each region is represented by p, satisfying p=1,2,3..M.

[0135] Elastic gradient G p,q It is used to describe the elastic difference between the pth region and the qth region. The calculation formula is: G p,q =|α p -α q |; The larger the value of the elasticity gradient, the more significant the difference in elasticity levels between regions.

[0136] By recording the changing trend of elastic gradient at different time points, the inter-regional elastic gradient matrix G is constructed, whose elements are G p,q .

[0137] Extract the spectral characteristics of transient disturbances and quantify the dynamic change trend of disturbances: Use transient disturbance data U p,h Construct time series {U p,h}, and extract its spectral characteristics through fast Fourier transform.

[0138] Defining spectral components Among them, S p,f It represents the energy component of the disturbance signal in the pth region at frequency f; f is the frequency variable; h1 and h2 are the start time and end time of the spectrum analysis time interval respectively; h is the time variable.

[0139] Define the spectrum variation in, Represents the spectrum components at the current moment; Represents the spectral components under the reference state.

[0140] Establish a synergistic model of elasticity gradient and disturbance characteristics to analyze the nonlinear impact of synergy on scheduling: The synergy model describes the coupling between elasticity and disturbance by quantifying the interaction between elasticity gradient and spectrum variation. Define the synergy index G p,q,f : G p,q,f =G p,q *ΔS p,f Among them, G p,q,f is the synergistic effect index, which indicates the synergistic effect strength between the p-th region and the q-th region at frequency f.

[0141] By analyzing the synergy index G p,q,f The distribution law of the disturbance spectrum is used to identify the potential nonlinear impact of the synergistic effect of elastic gradient and disturbance spectrum on scheduling errors.

[0142] Calculate the scheduling error caused by synergy and assess the potential erosion of scheduling accuracy: Defining Scheduling Error Among them, E d represents the scheduling error; f1 and f2 represent the lower and upper limits of the frequency range involved in the spectrum analysis respectively; W f Represents the weight coefficient at a given frequency f; G p,q,f is the synergistic effect index, which indicates the synergistic effect strength between the p-th region and the q-th region at frequency f.

[0143] Define the scheduling accuracy erosion index Among them, SAEI is the scheduling accuracy erosion index; E d represents the scheduling error; T d Indicates the scheduling accuracy tolerance threshold.

[0144] The larger the dispatch accuracy erosion index, the greater the potential erosion of dispatch error caused by multi-dimensional elastic coordination on the demand side on the dispatch instruction execution accuracy. This shows that under the influence of demand side elastic changes and transient disturbances, the dynamic response and disturbance characteristics of the demand side may lead to execution deviations of the virtual power plant dispatch system. When the dispatch accuracy erosion index is large, the error of the dispatch system during execution may increase, resulting in the inability to accurately execute the dispatch plan, which in turn affects the stability and economy of power supply.

[0145] Specifically, by comprehensively analyzing the influence of the local group coordination effect on the overall output boundary recognition accuracy of the virtual power plant and the influence of the dispatch error caused by the multi-dimensional elastic coordination on the demand side on the dispatch instruction execution accuracy, the internal optimization control plan of the virtual power plant is adjusted, including:

[0146] The recognition accuracy index corresponding to the degree of influence of the local group coordination effect on the overall output boundary recognition accuracy of the virtual power plant and the dispatch accuracy erosion index corresponding to the influence of the dispatch error caused by the multi-dimensional elastic coordination on the dispatch instruction execution accuracy are normalized, and the normalized recognition accuracy index and dispatch accuracy erosion index are calculated to obtain the comprehensive performance adjustment coefficient. According to the comprehensive performance adjustment coefficient, the internal control plan of the virtual power plant is adjusted.

[0147] The calculation formula of the comprehensive performance adjustment coefficient is: Among them, CPA represents the comprehensive performance adjustment coefficient; SAEI is the scheduling accuracy erosion index; RAI is the recognition accuracy index; μ is the adjustment coefficient, which is used to adjust the relationship between the recognition accuracy index and the scheduling accuracy erosion index.

[0148] A comprehensive performance adjustment coefficient threshold is preset, and the comprehensive performance adjustment coefficient is compared with the comprehensive performance adjustment coefficient threshold:

[0149] When the comprehensive performance adjustment coefficient is greater than or equal to the comprehensive performance adjustment coefficient threshold, there may be large scheduling errors and inaccurate identification problems, and the virtual power plant scheduling system needs to adjust the internal control plan;

[0150] When the comprehensive performance adjustment coefficient is greater than or equal to the comprehensive performance adjustment coefficient threshold, the virtual power plant dispatching system operates smoothly and there is no need to adjust the internal control plan;

[0151] The setting of the comprehensive performance adjustment coefficient threshold is based on multiple factors such as the actual operation of the virtual power plant, the equipment load characteristics and the responsiveness of its dispatching system. It is usually determined by empirical data, historical dispatching results and the performance evaluation of the optimization algorithm. In order to ensure the robustness and flexibility of the virtual power plant operation, the setting of the threshold should take into account a certain safety margin to cope with sudden demand fluctuations or system disturbances.

[0152] Adjusting the internal control plan includes immediately launching the optimization control mechanism, including but not limited to adjusting the sensitivity of the dispatch instruction, reallocating load response resources or introducing more advanced control algorithms to quickly slow down the spread of dispatch errors and restore the dispatch accuracy and stability of the dispatch system. Specific measures include dynamically adjusting the output strategy of distributed energy, optimizing the charging and discharging plan of energy storage equipment, and coordinating demand response measures to cope with sudden market price fluctuations and demand changes.

[0153] Example 2

[0154] The difference between Example 2 of the present invention and Example 1 is that this example introduces a virtual power plant dispatching system.

[0155] Figure 2 A structural schematic diagram of a virtual power plant dispatching system of the present invention is given, and a virtual power plant dispatching system includes an energy monitoring module, a disturbance analysis module, an identification accuracy assessment module, a corrosion degree assessment module and a control module;

[0156] Energy monitoring module: By monitoring the output characteristics of distributed energy resources connected to the virtual power plant under the disturbance of high-frequency power market price signals, it is determined whether the transient fluctuations of distributed energy resources in the short-cycle scheduling process will cause nonlinear interference to the real-time identification of the comprehensive control strategy;

[0157] Disturbance analysis module: When the transient fluctuations of distributed energy resources in the short-cycle scheduling process produce nonlinear interference on the real-time identification of the comprehensive control strategy, by analyzing the disturbance of the distributed energy in the virtual power plant to the power flow inversion path, it is determined whether there is an abnormality in the impact of abnormal changes in the coupling correlation between potential hidden resources on the scheduling accuracy;

[0158] Identification accuracy assessment module: When the abnormal change of coupling correlation between potential hidden resources has an abnormal impact on the dispatching accuracy: through fine-grained analysis of the node differentiation response mode between the distributed energy storage aggregation area and the flexible load community, the influence of the local group coordination effect on the overall output boundary identification accuracy of the virtual power plant is evaluated;

[0159] Erosion degree assessment module: By analyzing the spectral characteristics of demand-side elasticity gradient and transient disturbance at different time and space scales, the potential erosion degree of scheduling error caused by multi-dimensional elasticity coordination on the execution accuracy of scheduling instructions is assessed;

[0160] Control module: Adjust the internal control plan of the virtual power plant by conducting a comprehensive analysis of the impact of local group coordination effect on the overall output boundary identification accuracy of the virtual power plant and the impact of the dispatch error caused by multi-dimensional elastic coordination on the demand side on the execution accuracy of dispatch instructions.

[0161] The above formulas are dimensionless and numerical calculations. The formula is a formula obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions. The parts not described in the present invention are applicable to the prior art.

Claims

1. A virtual power plant scheduling method, characterized in that: The steps include: By monitoring the output characteristics of distributed energy resources connected to virtual power plants under the disturbance of high-frequency power market price signals, it is determined whether the transient fluctuations of distributed energy resources in the short-cycle scheduling process will cause nonlinear interference in the real-time identification of the comprehensive control strategy; When the transient fluctuations of distributed energy resources in the short-cycle dispatch process produce nonlinear interference in the real-time identification of the comprehensive control strategy, by analyzing the disturbance of the distributed energy within the virtual power plant to the power flow inversion path, it is determined whether there is an abnormality in the impact of abnormal changes in the coupling correlation between potential hidden resources on the dispatch accuracy; When there is an abnormal impact of abnormal changes in the coupling correlation between potential hidden resources on the dispatching accuracy: through fine-grained analysis of the node-differentiated response patterns between the distributed energy storage aggregation area and the flexible load community, the influence of the local group coordination effect on the overall output boundary identification accuracy of the virtual power plant is evaluated; By analyzing the spectral characteristics of demand-side elasticity gradients and transient disturbances at different time and space scales, the potential erosion of dispatching error caused by multi-dimensional elasticity coordination on the dispatching instruction execution accuracy is evaluated. By comprehensively analyzing the influence of local group coordination effect on the overall output boundary identification accuracy of the virtual power plant and the impact of scheduling errors caused by multi-dimensional elastic coordination on the demand side on the execution accuracy of scheduling instructions, the internal control plan of the virtual power plant is adjusted.

2. A virtual power plant scheduling method according to claim 1, characterized in that: By monitoring the output characteristics of distributed energy resources connected to virtual power plants under the disturbance of high-frequency power market price signals, it is determined whether the transient fluctuations of distributed energy resources in the short-term scheduling process will cause nonlinear interference in the real-time identification of the comprehensive control strategy. Specifically: Monitor the distributed energy resources connected to the virtual power plant and record the changing characteristics of output power under the influence of high-frequency power market price fluctuations; During short-cycle scheduling, the output fluctuations of distributed energy are monitored in real time to identify the amplitude, frequency and time characteristics of transient fluctuations; Analyze the relationship between transient fluctuations and real-time identification of virtual power plant dispatch strategies, and evaluate whether there is nonlinear interference: A multivariate regression model is used to analyze the relationship between transient fluctuations and real-time identification of dispatch strategies. The mathematical form is: Among them, R t represents the real-time identification result of the scheduling strategy; γ0 represents the constant term of the regression model; γ i represents the regression coefficient of transient fluctuation amplitude; γ′ represents the regression coefficient of transient fluctuation frequency; γ i ″ represents the regression coefficient of transient fluctuation duration; δ t represents the model error term; N represents the number of distributed energy resources; The multivariate regression model is trained with the training data set to evaluate the impact of transient fluctuation characteristics on the identification of scheduling strategies; nonlinear terms are introduced to capture nonlinear interference: Among them, γ i,j A represents the nonlinear interaction regression coefficient between transient fluctuation amplitude and transient fluctuation frequency; i,t Represents distributed energy E i The transient fluctuation amplitude at time t; F j,t Represents distributed energy E j The transient fluctuation frequency at time t; T i,t Represents distributed energy E i Duration of transient fluctuation at time t; By testing the nonlinear interaction coefficient γ in the regression model i,j The statistical significance of is used to determine whether there is nonlinear interference: For the nonlinear interaction coefficient γ i,j Perform statistical tests when the nonlinear interaction coefficient γ i,j When the value of is greater than the critical value of the statistical test, the nonlinear interaction coefficient γ i,j If the coefficient of the nonlinear interaction term is significantly not zero, it is determined that the transient fluctuation of distributed energy resources has produced nonlinear interference in the real-time identification of the comprehensive control strategy.

3. A virtual power plant scheduling method according to claim 2, characterized in that: When the transient fluctuations of distributed energy resources in the short-cycle scheduling process produce nonlinear interference in the real-time identification of the comprehensive control strategy, by analyzing the disturbance of the distributed energy in the virtual power plant to the power flow inversion path, it is judged whether there is an abnormality in the influence of abnormal changes in the coupling correlation between potential hidden resources on the scheduling accuracy, specifically: Extract disturbance parameters of the distributed energy power flow inversion path within the virtual power plant; Analyze the changing rules of coupling correlation between potential hidden resources; Determine whether the abnormal coupling correlation has an abnormal effect on the dispatching accuracy: establish the following judgment formula: D(m,n)=R(m,n)*S(m,n)-Q(m,n); where D(m,n) is the abnormality judgment index; R(m,n) is the correlation change parameter; S(m,n) represents the dispatching accuracy sensitivity parameter; Q(m,n) represents the dispatching accuracy error correction parameter; If D(m,n) is greater than 0, it means that the abnormal change of coupling correlation between potential implicit resources has an abnormal impact on scheduling accuracy; If D(m,n) is less than or equal to 0, it means that there is no abnormal impact of abnormal changes in the coupling correlation between potential implicit resources on scheduling accuracy.

4. A virtual power plant scheduling method according to claim 3, characterized in that: When the abnormal change of coupling correlation between potential hidden resources has an abnormal impact on dispatch accuracy: through fine-grained analysis of the node differentiation response mode between distributed energy storage aggregation area and flexible load group, the influence of local group coordination effect on the overall output boundary identification accuracy of virtual power plant is evaluated, specifically: Collect the real-time power, frequency and voltage responses of each node in the distributed energy storage aggregation area and flexible load group; Extracting dynamic response characteristics between nodes from the collected data, including power offset rate, time delay and voltage disturbance; Based on the dynamic response characteristics between nodes, a fine-grained response model of distributed energy storage and flexible load communities is constructed to quantify the differentiation mode; Combined with fine-grained models, the coordinated response of distributed energy storage aggregation areas and flexible load groups is evaluated; Comprehensive group coordination effect, analyze the impact on the recognition accuracy of the overall output boundary of the virtual power plant: Definition: Among them, RAI is the recognition accuracy index; H syg (c) represents the quantitative value of the coordination response effect at time point c; σ op (c) represents the standard deviation of the overall output of the virtual power plant at time point c.

5. A virtual power plant scheduling method according to claim 4, characterized in that: By analyzing the spectral characteristics of demand-side elastic gradients and transient disturbances at different time and space scales, the potential erosion of the dispatch error caused by the multi-dimensional elastic coordination on the demand side on the execution accuracy of dispatch instructions is evaluated, specifically: Collect demand-side elasticity data and transient disturbance characteristics at different time and space scales; Partition the demand-side elasticity data and construct an elasticity gradient model at spatiotemporal scales; Extract the spectral characteristics of transient disturbances and quantify the dynamic change trend of disturbances; Establish a collaborative model of elastic gradient and disturbance characteristics to analyze the nonlinear impact of collaborative effects on scheduling; Calculate the scheduling error caused by the synergy effect and evaluate the potential erosion of scheduling accuracy: Define the scheduling error E d : Among them, E d represents the scheduling error; f1 and f2 represent the lower and upper limits of the frequency range involved in the spectrum analysis respectively; W f Represents the weight coefficient at a given frequency f; G p,q,f is the synergy index; Define the scheduling accuracy erosion index SAEI: Among them, SAEI is the scheduling accuracy erosion index; E d represents the scheduling error; T d Indicates the scheduling accuracy tolerance threshold.

6. A virtual power plant scheduling method according to claim 5, characterized in that: Through a comprehensive analysis of the impact of local group coordination effect on the overall output boundary recognition accuracy of the virtual power plant and the impact of the dispatch error caused by the multi-dimensional elastic coordination on the dispatch instruction execution accuracy, the internal control plan of the virtual power plant is adjusted, specifically: The recognition accuracy index corresponding to the degree of influence of the local group coordination effect on the overall output boundary recognition accuracy of the virtual power plant and the dispatch accuracy erosion index corresponding to the influence of the dispatch error caused by the multi-dimensional elastic coordination on the dispatch instruction execution accuracy are normalized, and the normalized recognition accuracy index and dispatch accuracy erosion index are calculated to obtain the comprehensive performance adjustment coefficient. The calculation formula is: Among them, CPA represents the comprehensive performance adjustment coefficient; SAEI is the scheduling accuracy erosion index; RAI is the recognition accuracy index; μ is the adjustment coefficient; A comprehensive performance adjustment coefficient threshold is preset, and the comprehensive performance adjustment coefficient is compared with the comprehensive performance adjustment coefficient threshold: When the comprehensive performance adjustment coefficient is greater than or equal to the comprehensive performance adjustment coefficient threshold, the virtual power plant dispatching system needs to adjust the internal control plan; When the comprehensive performance adjustment coefficient is greater than or equal to the comprehensive performance adjustment coefficient threshold, the virtual power plant dispatching system does not need to adjust the internal control plan.

7. A virtual power plant dispatching system, used to implement a virtual power plant dispatching method according to any one of claims 1 to 6, characterized in that: It includes energy monitoring module, disturbance analysis module, identification accuracy assessment module, erosion degree assessment module and control module; Energy monitoring module: By monitoring the output characteristics of distributed energy resources connected to the virtual power plant under the disturbance of high-frequency power market price signals, it is determined whether the transient fluctuations of distributed energy resources in the short-cycle scheduling process will cause nonlinear interference to the real-time identification of the comprehensive control strategy; Disturbance analysis module: When the transient fluctuations of distributed energy resources in the short-cycle scheduling process produce nonlinear interference on the real-time identification of the comprehensive control strategy, by analyzing the disturbance of the distributed energy in the virtual power plant to the power flow inversion path, it is determined whether there is an abnormality in the impact of abnormal changes in the coupling correlation between potential hidden resources on the scheduling accuracy; Identification accuracy assessment module: When the abnormal change of coupling correlation between potential hidden resources has an abnormal impact on the dispatching accuracy: through fine-grained analysis of the node differentiation response mode between the distributed energy storage aggregation area and the flexible load community, the influence of the local group coordination effect on the overall output boundary identification accuracy of the virtual power plant is evaluated; Erosion degree assessment module: By analyzing the spectral characteristics of demand-side elasticity gradient and transient disturbance at different time and space scales, the potential erosion degree of scheduling error caused by multi-dimensional elasticity coordination on the execution accuracy of scheduling instructions is assessed; Control module: Adjust the internal control plan of the virtual power plant by conducting a comprehensive analysis of the impact of local group coordination effect on the overall output boundary identification accuracy of the virtual power plant and the impact of the dispatch error caused by multi-dimensional elastic coordination on the demand side on the execution accuracy of dispatch instructions.

Images