A power emergency online deduction optimization method and system based on EMD method

Abstract

The present invention relates to an optimization method and system for online power emergency simulation based on an EMD method. The method comprises: constructing a short-time scale evaluation index, a medium- and long-time scale evaluation index, and a long-time scale evaluation index; obtaining index data for online power emergency simulation and performing stabilization processing using a modal decomposition method; using an adaptive weight model to determine the weight of each index, and introducing a state variable weight factor for adjustment; combining the index weight and index data, using a Topsis model to evaluate the online power emergency simulation, and optimizing according to the evaluation results. Compared with the prior art, the present invention constructs a dynamic evaluation model for online power emergency plan simulation under multiple time scales from a global perspective. By analyzing the evaluation results, it is possible to discover the deficiencies of online power emergency simulation based on a dynamic irregular mechanism and make timely improvements, thereby making the power emergency plan have higher execution efficiency when handling actual emergencies.

Description

Technical Field

[0001] The present invention relates to the field of emergency drill evaluation and optimization of power grid enterprises, and in particular to an EMD-based power emergency online deduction optimization method and system. Background Art

[0002] Like all large-scale production activities, power generation carries potential safety risks. Power emergency drills, or power accident response exercises, are a crucial component of power emergency preparedness and a crucial activity in the prevention phase. These drills serve an irreplaceable assessment function. They assess the grid's ability to respond to accidents, showcasing its advanced technologies and methods. They also serve as an assessment of the response capabilities of grid personnel.

[0003] Most existing emergency simulations are conducted under static scripts, and the high cost of actual simulations makes it impossible to normalize emergency simulations. Therefore, the simulation evaluation models designed for such power emergency simulations cannot truly provide guiding improvement suggestions for the simulations themselves. Online simulations have the advantages of low overall cost, minimal impact on the actual operation of the power grid, and easy normalization, and therefore are widely used. In addition, the evaluation of power emergency simulations is the most effective way to improve and optimize emergency simulations. A reasonable simulation evaluation model can identify deficiencies in the emergency response process and also verify the effectiveness of the corresponding emergency plan. Therefore, the simulation evaluation model of power emergency plan should be the focus of power emergency research. At present, there is no good evaluation model to evaluate online emergency simulations, which makes it impossible to timely identify and improve the deficiencies of the corresponding power emergency plan. Summary of the Invention

[0004] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide an electric power emergency online deduction optimization method and system based on the EMD method.

[0005] The purpose of the present invention can be achieved by the following technical solutions:

[0006] An EMD-based power emergency online deduction optimization method includes the following steps:

[0007] S1. Construct short-time scale evaluation indicators, medium- and long-time scale evaluation indicators, and long-time scale evaluation indicators;

[0008] S2. Obtain the index data of the power emergency online simulation and use the modal decomposition method to stabilize the index data;

[0009] S3, using an adaptive weight model to determine the weight of each indicator, and introducing a state variable weight factor to adjust the weight;

[0010] S4. Combine the indicator weights and indicator data, use the Topsis model to evaluate the online simulation of power emergency, and improve the online simulation of power emergency based on the evaluation results.

[0011] Preferably, the short-time scale evaluation indicator is an elasticity index, which has the following specific meanings:

[0012]

[0013] Where K(t) represents the elasticity index of the power system at time t, t0 is a moment when the power system is not disturbed, that is, the moment when the power system operates normally; t d is the moment when the power system is in the worst operating state after being disturbed; ΔP R (t) represents the degree of imbalance between supply and demand of the power system at time t, and:

[0014]

[0015] in, is the power generation of the thermal power generating unit in the power system at time t; is the power generation of the wind turbine in the power system at time t; is the power generation of the photovoltaic device in the power system at time t; is the power generation of the storage device in the power system at time t; is the total load of the power system at time t;

[0016] The medium- and long-term evaluation indicator is the dispatch coefficient. The dispatch resources of the power system include the transmission lines between the upper and lower power grids, controllable loads, and energy storage devices in the microgrid. The specific meaning of the dispatch coefficient is as follows:

[0017]

[0018] Among them, A(t) represents the dispatch coefficient of the power system at time t, ΔP MGi (t) represents the dispatch output of microgrid i in the power system at time t, λ MGi (t) represents the dispatch coefficient of microgrid i in the power system at time t; ΔP CL (t) represents the dispatch output of controllable loads in the power system at time t, λ CL (t) represents the dispatch coefficient of controllable load in the power system at time t; ΔP trans (t) represents the exchange dispatch output of the distribution network and the upper power grid in the power system at time t, λ trans represents the exchange scheduling coefficient of the power system, that is, the scheduling coefficient of the transmission line resources, μ trans(t) indicates whether the transmission line switch between the upper and lower power grids in the power system is closed at time t, and takes values ​​of 0 and 1; P cj (t) is the impact load change value of the power system at time t;

[0019] The long-term evaluation index is a static calculation index fitted by multiple factors through the DEA method. The specific meaning is as follows:

[0020]

[0021] Among them, ρ represents the static calculation index under long time scale, c1 and c2 represent the operation cost and organization cost of power emergency online simulation respectively, and c sum is the total budget cost, e loss is the equipment loss, e total is the total number of electrical equipment, m loss is the consumption of emergency supplies, m total is the total amount of emergency supplies, p use is the actual number of personnel mobilized for the power emergency online simulation, p total is the total number of personnel participating in the power emergency online simulation, and η is the system recovery percentage.

[0022] Preferably, the specific meaning of the dispatch coefficient of microgrid i in the power system at time t is as follows:

[0023]

[0024] Among them, ES i (t) represents the energy storage capacity of the energy storage device in microgrid i at time t, ES imax Indicates the maximum power storage capacity of the energy storage device in microgrid i, ES imin represents the minimum power reserve of the energy storage device in microgrid i, λ MGi,min represents the minimum dispatch coefficient allocated to microgrid i;

[0025] Controllable load dispatch coefficient λ CL (t) and the exchange scheduling coefficient λ trans Take a constant, and λ MGi,min >λ CL (t)>λ trans .

[0026] Preferably, in step S2, the modal decomposition method is used to perform stabilization processing on the indicator data of the short-time scale evaluation index and the medium- and long-time scale evaluation index, comprising the following steps:

[0027] S21. Obtain a dynamic indicator evaluation value and set it as the signal sequence X(t);

[0028] S22. Find all extreme points of the signal sequence X(t), use spline interpolation to form a lower envelope for the minimum points and an upper envelope for the maximum points, and calculate the average envelope m(t) of the upper and lower envelopes;

[0029] S23, subtracting the average envelope sequence m(t) from the signal sequence X(t) to obtain h(t), and determining whether h(t) satisfies the first preset condition. If so, proceed to step S24; otherwise, take h(t) as the new signal sequence X(t) and proceed to step S22;

[0030] S24, taking h(t) as the IMF component of the signal sequence X(t), denotes that the signal sequence X(t) is subtracted from h(t) to obtain the residue R(t), and taking the residue R(t) as the new signal sequence X(t), executing step S22 until R(t) meets the second preset condition, and the signal sequence X(t) in step S21 is decomposed into the residue R(t) and multiple IMF components;

[0031] S25. Calculate the average value of the sequence R(t), and use the average value as the index data after stabilization processing.

[0032] Preferably, step S3 is specifically as follows:

[0033] The adaptive weight model is used to assign weights to the three indicators, and the initial weights ω'1, ω'2, and ω'3 of the short-time scale evaluation indicator, the medium- and long-time scale evaluation indicator, and the long-time scale evaluation indicator are obtained. The state-variable weight factor of each indicator is calculated, and the initial weight is modified using the state-variable weight factor to obtain the weights ω1, ω2, and ω3 of each indicator.

[0034] Among them, ω i =S i ×ω' i , S i is the state variable weight factor of the i-th indicator, and its specific meaning is as follows:

[0035]

[0036] Among them, max i Indicates the maximum reasonable value of the i-th indicator, min i Indicates the minimum reasonable value of the i-th indicator, a i represents the indicator data of the i-th indicator, and t represents the preset penalty factor.

[0037] Preferably, in step S4, the evaluation of the power emergency online simulation using the Topsis model is specifically as follows:

[0038] Calculate the dynamic evaluation closeness C of the power emergency online simulation:

[0039]

[0040]

[0041]

[0042] Among them, C represents the dynamic evaluation closeness of the power emergency online simulation, represents the indicator data of the i-th indicator, represents the positive ideal solution of the i-th index, represents the negative ideal solution of the i-th index, ω i represents the weight of the i-th indicator.

[0043] An online power emergency deduction and optimization system based on the EMD method includes an indicator construction module, a data processing module, a weight determination module and an evaluation and optimization module;

[0044] The indicator construction module is used to construct short-time scale evaluation indicators, medium- and long-time scale evaluation indicators, and long-time scale evaluation indicators;

[0045] The data processing module is used to obtain the index data of the power emergency online deduction and use the modal decomposition method to perform a stabilization process on the index data;

[0046] The weight determination module adopts an adaptive weight model to determine the weight of each indicator and introduces a state variable weight factor to adjust the weight;

[0047] The evaluation and optimization module combines the weights and index data of the indicators, uses the Topsis model to evaluate the online power emergency simulation, and optimizes the online power emergency simulation based on the evaluation results.

[0048] Preferably, among the indicators constructed by the indicator construction module, the short-time scale evaluation indicator is the elasticity index, which has the following specific meanings:

[0049]

[0050] Where K(t) represents the elasticity index of the power system at time t, t0 is a moment when the power system is not disturbed, that is, the moment when the power system operates normally; t d is the moment when the power system is in the worst operating state after being disturbed; ΔP R (t) represents the degree of imbalance between supply and demand of the power system at time t, and:

[0051]

[0052] in, is the power generation of the thermal power generating unit in the power system at time t; is the power generation of the wind turbine in the power system at time t; is the power generation of the photovoltaic device in the power system at time t; is the power generation of the storage device in the power system at time t; is the total load of the power system at time t;

[0053] The medium- and long-term evaluation indicator is the dispatch coefficient. The dispatch resources of the power system include the transmission lines between the upper and lower power grids, controllable loads, and energy storage devices in the microgrid. The specific meaning of the dispatch coefficient is as follows:

[0054]

[0055] Among them, A(t) represents the dispatch coefficient of the power system at time t, ΔP MGi (t) represents the dispatch output of microgrid i in the power system at time t, λ MGi (t) represents the dispatch coefficient of microgrid i in the power system at time t; ΔP CL (t) represents the dispatch output of controllable loads in the power system at time t, λ CL (t) represents the dispatch coefficient of controllable load in the power system at time t; ΔP trans (t) represents the exchange dispatch output of the distribution network and the upper power grid in the power system at time t, λ trans represents the exchange scheduling coefficient of the power system, that is, the scheduling coefficient of the transmission line resources, μ trans (t) indicates whether the transmission line switch between the upper and lower power grids in the power system is closed at time t, and takes values ​​of 0 and 1; P cj (t) is the impact load change value of the power system at time t;

[0056] The long-term evaluation index is a static calculation index fitted by multiple factors through the DEA method. The specific meaning is as follows:

[0057]

[0058] Among them, ρ represents the static calculation index under long time scale, c1 and c2 represent the operation cost and organization cost of power emergency online simulation respectively, and c sum is the total budget cost, e loss is the equipment loss, e total is the total number of electrical equipment, m loss is the consumption of emergency supplies, m total is the total amount of emergency supplies, p use is the actual number of personnel mobilized for the power emergency online simulation, p totalis the total number of personnel participating in the power emergency online simulation, and η is the system recovery percentage.

[0059] Preferably, the weight determination module works specifically as follows:

[0060] The adaptive weight model is used to assign weights to the three indicators, and the initial weights ω'1, ω'2, and ω'3 of the short-time scale evaluation indicator, the medium- and long-time scale evaluation indicator, and the long-time scale evaluation indicator are obtained. The state-variable weight factor of each indicator is calculated, and the initial weight is modified using the state-variable weight factor to obtain the weights ω1, ω2, and ω3 of each indicator.

[0061] Among them, ω i =S i ×ω' i , S i is the state variable weight factor of the i-th indicator, and its specific meaning is as follows:

[0062]

[0063] Among them, max i Indicates the maximum reasonable value of the i-th indicator, min i Indicates the minimum reasonable value of the i-th indicator, a i represents the indicator data of the i-th indicator, and t represents the preset penalty factor.

[0064] Preferably, in the evaluation and optimization module, the Topsis model is used to evaluate the power emergency online simulation as follows:

[0065] Calculate the dynamic evaluation closeness C of the power emergency online simulation:

[0066]

[0067]

[0068]

[0069] Among them, C represents the dynamic evaluation closeness of the power emergency online simulation, represents the indicator data of the i-th indicator, represents the positive ideal solution of the i-th index, represents the negative ideal solution of the i-th index, ω i represents the weight of the i-th indicator.

[0070] Compared with the prior art, the present invention has the following beneficial effects:

[0071] From a global perspective, a dynamic evaluation model for online simulation of power emergency plans under multiple time scales is constructed. By analyzing the evaluation results, the shortcomings of online simulation of power emergencies based on the dynamic irregular mechanism can be discovered and improved in a timely manner, thereby making the power emergency plan more efficient in handling actual emergencies. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 Schematic diagram of the evaluation indicators;

[0073] Figure 2 It is a diagram of the supply and demand balance changes of the power system;

[0074] Figure 3 Schematic diagram of EMD frequency reduction of elasticity index. DETAILED DESCRIPTION

[0075] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0076] In the drawings, components with identical structures are denoted by the same reference numerals, and components with similar structures or functions are denoted by similar reference numerals. The size and thickness of each component shown in the drawings are arbitrary and are not limited by the present invention. For clarity, some components in the drawings are exaggerated.

[0077] Example 1:

[0078] An EMD-based power emergency online deduction optimization method includes the following steps:

[0079] S1, such as Figure 1 As shown, short-time scale evaluation indicators, medium- and long-time scale evaluation indicators, and long-time scale evaluation indicators are constructed;

[0080] S2. Obtain the index data of the power emergency online simulation and use the modal decomposition method to stabilize the index data;

[0081] S3, using an adaptive weight model to determine the weight of each indicator, and introducing a state variable weight factor to adjust the weight;

[0082] S4. Combine the indicator weights and indicator data, use the Topsis model to evaluate the online simulation of power emergency, and improve the online simulation of power emergency based on the evaluation results.

[0083] This application first establishes an evaluation index system for power emergency simulation based on different time scales, and constructs three emergency plan simulation evaluation indicators at different time scales, namely short time scale evaluation index, medium and long time scale evaluation index and long time scale evaluation index, to evaluate the dynamic online emergency simulation based on irregular rules, such as Figure 1 shown.

[0084] In the evaluation of the emergency plan simulation, the modal decomposition method (EMD) was used to standardize the data. The values ​​of the short-term and medium- to long-term evaluation indicators can be considered as non-stationary discrete data that changes at different time points. Therefore, the EMD method was used to stabilize the data of these two evaluation indicators, which were then used as the final simulation evaluation indicators.

[0085] On this basis, a comprehensive evaluation model for online simulation of power emergency plans was established. An adaptive weight model was used to weight the evaluation indicators at three different time scales. A variable weight correction coefficient was introduced to modify the weights of the evaluation indicators. Finally, the Topsis model was used to conduct a comprehensive evaluation of power emergency simulations.

[0086] 1. Short-term evaluation indicators:

[0087] When the power system faces an emergency, the system performance will change. In the power emergency online simulation, the supply and demand balance relationship of the power grid is used as a short-time scale evaluation indicator to measure the simulation quality. This is because the system performance changes rapidly and the change cycle is short. Therefore, this application selects the elasticity index of the power system as the short-time scale simulation evaluation indicator. The definition and analysis of the elasticity index are as follows: Figure 2 As shown:

[0088] At time t0, the system's supply and demand are balanced. When the system receives an unexpected disturbance, the system's supply and demand imbalance decreases. At t d The lowest point at t is the lowest point. When effective emergency response measures are taken, the imbalance between supply and demand in the system gradually increases. f Achieve stability at all times.

[0089] The short-time scale evaluation index in this application is the elasticity index, and its specific meaning is as follows: The short-time scale evaluation index is the elasticity index, and its specific meaning is as follows:

[0090]

[0091] Where K(t) represents the elasticity index of the power system at time t, t0 is a moment when the power system is not disturbed, that is, the moment when the power system operates normally; t d is the moment when the power system is in the worst operating state after being disturbed; ΔP R(t) represents the degree of imbalance between supply and demand of the power system at time t, and:

[0092]

[0093] in, is the power generation of the thermal power generating units in the power system at time t, which is understood as the sum of the power generation of all thermal power generating units in the power system; is the power generation of the wind turbine generator set in the power system at time t, which is understood as the sum of the power generation of all wind turbine generator sets in the power system; is the power generation of the photovoltaic device in the power system at time t, which is understood as the sum of the power generation of all photovoltaic devices in the power system; is the power generation of the storage device in the power system at time t, taking into account the charge and discharge of the storage device; is the total load of the power system at time t, which is understood as the sum of all loads in the power system;

[0094] The elasticity coefficient is a broadly defined indicator that includes the evaluation content of multiple power systems. It can be dynamically modified according to different operating modes. Relevant practitioners can understand it, so I will not elaborate on it here.

[0095] 2. Medium- and long-term evaluation indicators

[0096] The load of the power system changes in real time, but the system's dispatching capacity can be considered unchanged in a short period of time. Therefore, this application selects the system's dispatching coefficient as the medium- and long-term scale evaluation indicator for emergency simulation.

[0097] The grid's dispatchable resources include transmission lines between upper and lower layers, controllable loads, and energy storage devices within microgrids. A dispatch coefficient, A(t), is established to assess the rationality of the grid's emergency dispatch response to load changes caused by emergencies. From the dispatcher's perspective, priority is given to dispatching microgrid and load resources within the distribution network to ensure power matching for load changes, before considering increasing or decreasing power on transmission lines between upper and lower layers.

[0098] The dispatch coefficient is a virtual coefficient that can be given by the dispatcher himself based on the actual operation of the power grid. The larger the dispatch coefficient, the greater the possibility that the dispatchable resources corresponding to the dispatch coefficient will be selected in the power emergency simulation and will produce more power to cope with the impact load caused by the emergency.

[0099] In this application, the medium- and long-term evaluation indicator is the dispatch coefficient. The dispatch resources of the power system include the transmission lines between the upper and lower grids, controllable loads, and energy storage devices in the microgrid. The specific meaning of the dispatch coefficient is as follows:

[0100]

[0101] Among them, A(t) represents the dispatch coefficient of the power system at time t, ΔP MGi (t) represents the dispatch output of microgrid i in the power system at time t, λ MGi (t) represents the dispatch coefficient of microgrid i in the power system at time t; ΔP CL (t) represents the dispatch output of controllable loads in the power system at time t, λ CL (t) represents the dispatch coefficient of controllable load in the power system at time t; ΔP trans (t) represents the exchange dispatch output of the distribution network and the upper power grid in the power system at time t, λ trans It represents the exchange scheduling coefficient of the power system, that is, the scheduling coefficient of the transmission line resources. It is generally a constant and is not affected by the time scale. trans (t) indicates whether the transmission line switch between the upper and lower power grids in the power system is closed at time t, and takes 0 and 1 to represent the open and closed states respectively; P cj (t) is the impact load change value of the power system at time t; the above parameters are all parameters defined in the prior art, which can be understood by relevant practitioners and will not be described in detail here;

[0102] The specific meaning of the dispatch coefficient of microgrid i in the power system at time t is as follows:

[0103]

[0104] Among them, ES i (t) represents the energy storage capacity of the energy storage device in microgrid i at time t, ES imax Indicates the maximum power storage capacity of the energy storage device in microgrid i, ES imin represents the minimum power reserve of the energy storage device in microgrid i, λ MGi,min represents the minimum dispatch coefficient allocated to microgrid i;

[0105] Controllable load dispatch coefficient λ CL (t) and the exchange scheduling coefficient λ trans Take a constant, and generally λ MGi,min >λ CL (t)>λ trans .

[0106] Because energy storage in a microgrid is a fast-responding dispatching resource, if the microgrid has sufficient storage capacity, prioritizing its use to address load surges is crucial. When addressing load surges, prioritizing dispatching the local grid's energy storage capacity is crucial, followed by flexible load response, and finally, adjusting the transmission power on transmission lines between upstream and downstream grids.

[0107] 3. Long-term evaluation indicators

[0108] The purpose of power emergency simulation is to enable power companies and emergency personnel to effectively reduce the impact of emergencies on the power grid when emergencies occur. Therefore, power companies need to conduct emergency simulations regularly, which requires that the comprehensive cost of emergency simulations should not be too high. The comprehensive cost of power emergency simulations includes simulation operation costs and simulation organization costs. At the same time, all emergency plans hope to minimize the overall losses caused by emergencies, and the response losses in power emergency plans include system equipment losses, emergency material consumption, and the number of personnel involved. Finally, the emergency plan also hopes that after the emergency response, the power system can be restored to the state before the emergency occurred. In summary, this application provides long-term evaluation indicators for power emergency simulations, including: simulation operation costs, simulation organization costs, equipment losses, emergency material consumption, number of personnel involved, and system recovery status. This application uses the Data Envelopment Analysis (DEA) method to fit these long-term evaluation indicators of power emergency simulations into a static calculation indicator.

[0109] The long-term evaluation index is a static calculation index that is fitted by multiple factors through the DEA method. The specific meanings are as follows:

[0110]

[0111] Among them, ρ represents the static calculation index under long time scale, c1 and c2 represent the operation cost and organization cost of power emergency online simulation respectively, and c sum is the total budget cost, e loss is the equipment loss, e total is the total number of electrical equipment, m loss is the consumption of emergency supplies, m total is the total amount of emergency supplies, p use is the actual number of personnel mobilized for the power emergency online simulation, p total is the total number of personnel participating in the online power emergency simulation, η is the system recovery percentage, and the above parameters are relevant data of the emergency simulation, which can be understood by relevant practitioners and will not be elaborated here.

[0112] For the short-time scale evaluation indicators and the medium- and long-time scale evaluation indicators, their indicator values ​​can be considered as non-stationary discrete data that changes with different time nodes. Using the EMD method to down-convert the acquired data can obtain a stable integrated indicator value, thereby standardizing the evaluation indicators of different time scales. In step S2, the indicator data of the short-time scale evaluation indicators and the medium- and long-time scale evaluation indicators are stabilized using the modal decomposition method, including the following steps:

[0113] S21. Obtain a dynamic indicator evaluation value and set it as the signal sequence X(t);

[0114] S22. Find all extreme points of the signal sequence X(t), use spline interpolation to form a lower envelope for the minimum points and an upper envelope for the maximum points, and calculate the average envelope m(t) of the upper and lower envelopes;

[0115] S23, subtracting the average envelope sequence m(t) from the signal sequence X(t) to obtain h(t), and determining whether h(t) satisfies the first preset condition. If so, proceed to step S24; otherwise, take h(t) as the new signal sequence X(t) and proceed to step S22;

[0116] S24, taking h(t) as the IMF component of the signal sequence X(t), denotes that the signal sequence X(t) is subtracted from h(t) to obtain the residue R(t), and taking the residue R(t) as the new signal sequence X(t), executing step S22 until R(t) meets the second preset condition, and the signal sequence X(t) in step S21 is decomposed into the residue R(t) and multiple IMF components;

[0117] S25. Calculate the average value of the sequence R(t), and use the average value as the index data after stabilization processing.

[0118] Empirical Mode Decomposition (EMD) decomposes signals based on the time-scale characteristics of the data itself. It can linearize and stabilize data from nonlinear and non-stationary processes. The decomposed functions are orthogonal and theoretically uncorrelated, thus preserving as many of the basic characteristics of the original data as possible. EMD decomposition of the original signal yields n IMF components, ranging in frequency from high to low, and a residual, also known as the residual term. Since each IMF component represents a data sequence with a set of characteristic scales (frequencies), EMD decomposition effectively decomposes the original data sequence into a superposition of various characteristic fluctuations.

[0119] The spline interpolation method can use a third-order spline function, obtain the upper and lower envelopes, and then calculate the average of the two as the average envelope. Regarding the first and second preset conditions, relevant practitioners can set them according to conventional understanding. For example, the first preset condition can be "calculate the standard deviation and determine whether the standard deviation SD is within the threshold range. The threshold is usually 0.2 to 0.3." It can also be "the number of extreme points and the number of zero-crossing points are equal or differ by at most one and m(t) is equal to 0." The second preset condition can be "the residual R(t) is less than a certain set value" or "the residual R(t) becomes a monotonic function."

[0120] Step S3 is specifically as follows:

[0121] The adaptive weight model is used to assign weights to the three indicators, and the initial weights ω'1, ω'2, and ω'3 of the short-time scale evaluation indicator, the medium- and long-time scale evaluation indicator, and the long-time scale evaluation indicator are obtained. The state-variable weight factor of each indicator is calculated, and the initial weight is modified using the state-variable weight factor to obtain the weights ω1, ω2, and ω3 of each indicator.

[0122] Among them, ω i =S i ×ω' i , S i is the state variable weight factor of the i-th indicator, and its specific meaning is as follows:

[0123]

[0124] Among them, max i Indicates the maximum reasonable value of the i-th indicator, min i Indicates the minimum reasonable value of the i-th indicator, a i represents the indicator data of the i-th indicator, and t represents the preset penalty factor.

[0125] In step S4, the Topsis model is used to evaluate the power emergency online simulation. Specifically, the dynamic evaluation closeness C of the power emergency online simulation is calculated:

[0126]

[0127]

[0128]

[0129] Among them, C represents the dynamic evaluation closeness of the power emergency online simulation, represents the indicator data of the i-th indicator, represents the positive ideal solution of the i-th index, represents the negative ideal solution of the i-th index, ω i represents the weight of the i-th indicator.

[0130] as well as It represents the positive and negative ideal solutions of the power emergency simulation evaluation index set after EMD processing. It is a difference indicator and also serves as the optimization basis of the simulation algorithm. It is a dynamic acquisition quantity and is generally established by the expected simulation goal.

[0131] The Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), also known as the distance method, ranks the evaluation object by measuring its distance from the optimal and worst solutions. If the evaluation object is closest to the optimal solution and farthest from the worst solution, it is considered the best; otherwise, it is not the best. The optimal solution has all the optimal values ​​for each evaluation metric, while the worst solution has all the worst values ​​for each evaluation metric.

[0132] After down-conversion processing of short-time scale evaluation indicators and medium- and long-time scale evaluation indicators using the EMD method, the index data are obtained. The values ​​of the long-time scale evaluation indicators are directly used as the index data. Combined with the index weights, the dynamic evaluation closeness C of the power emergency online simulation is calculated using the Topsis method. The power emergency online simulation is evaluated and improved based on the dynamic evaluation closeness.

[0133] Example 2:

[0134] This application also protects an electric power emergency online deduction optimization system based on the EMD method, which is based on the electric power emergency online deduction optimization method based on the EMD method in Example 1, including an indicator construction module, a data processing module, a weight determination module and an evaluation optimization module;

[0135] The indicator construction module is used to construct short-time scale evaluation indicators, medium- and long-time scale evaluation indicators, and long-time scale evaluation indicators;

[0136] The data processing module is used to obtain the index data of the power emergency online simulation and use the modal decomposition method to stabilize the index data;

[0137] The weight determination module uses an adaptive weight model to determine the weight of each indicator and introduces a state variable weight factor to adjust the weight;

[0138] The evaluation and optimization module combines the indicator weights and indicator data, uses the Topsis model to evaluate the power emergency online simulation, and optimizes the power emergency online simulation based on the evaluation results.

[0139] Example 3:

[0140] An online simulation of a large-scale power outage emergency held by a power grid company was selected to verify this application. The simulation was an electric power emergency simulation based on a dynamic irregular model. The cause of the emergency was a 220 kV substation failure that triggered a chain failure of the transmission line. The simulation lasted for 2 hours.

[0141] First, the short-time scale evaluation index based on the elasticity index is used, with a time step of 2 minutes. The grid elasticity index under emergencies can be obtained as shown in Table 1.

[0142] Table 1 Dynamic changes in short-time scale indicators based on resilience index under emergencies

[0143]

[0144]

[0145] Then, the EMD method is used to reduce the frequency of the elasticity index value, and the following can be obtained: Figure 3 The curve shown in the figure shows the elasticity index value as the original data, and the integrated index value as the residual R(t) obtained after EMD decomposition. In this embodiment, the original signal sequence is decomposed into the residual R(t) and two IMF components. Finally, the residual is averaged to obtain the short-time scale evaluation index value based on the elasticity index.

[0146] Then, the medium and long time scale indicators based on the dispatch coefficient are presented. The time step is selected as 15 minutes, and the dispatch coefficient indicator values ​​of the system under emergency events can be obtained as shown in Table 2.

[0147] Table 2 Dynamic changes of short-time scale indicators based on scheduling coefficients under emergencies

[0148] time A(t) Integrated Value IMF1 IMF2 15 0.131 0.606 -0.333 -0.143 30 0.193 0.636 -0.31 -0.133 45 0.361 0.586 -0.158 -0.068 60 0.586 0.601 -0.0311 -0.017 75 0.731 0.588 0.1001 0.043 90 0.886 0.603 0.198 0.085 105 0.961 0.586 0.263 0.113 120 1 0.603 0.278 0.119

[0149] The EMD method is also used to perform frequency reduction operation on the scheduling coefficient index value. As shown in the table above, it is decomposed into the residual R(t) and two IMF components. The residual is the integrated value in the table. Then the residual R(t) is averaged to obtain the medium and long-term scale evaluation index value based on the scheduling coefficient.

[0150] Next, the static indicator values ​​of this power emergency online simulation and the calculated static fitting index sizes under long time scales are shown in Table 3.

[0151] Table 3 Dynamic changes of static long-time scale indicators under emergencies

[0152]

[0153] The adaptive weight partitioning model is used to assign weights to the evaluation indicators at different time scales, and the initial weights ω'1, ω'2, ω'3 = (0.4011, 0.3316, 0.2673) at different time scales can be obtained.

[0154] However, due to the particularity of power emergency simulation and evaluation, the same indicators are not comparable, so the indicator similarity in this application is defined as:

[0155]

[0156] Where, d i Represents the similarity of the i-th indicator, Represents the indicator data of the i-th indicator. For the static fitting indicator of a long time scale, That is ρ.

[0157] In actual application, it is easy to find that during the evaluation process, some indicators are not static in their impact on the final evaluation results. When the "quantitative change" of these indicators exceeds a certain standard, it will have a "qualitative change" effect on the final evaluation results. If the values ​​of these indicators drop to a certain level, even if the values ​​of other indicators in the emergency simulation evaluation indicators are very high, the final evaluation results should deteriorate sharply. The fixed-weight model is difficult to reflect the above-mentioned phenomenon of "quantitative change leading to qualitative change". Therefore, this application introduces a state variable weight factor to modify the initial weight of the evaluation indicator to obtain the weights of each indicator ω1, ω2, ω3;

[0158] Among them, ω i =S i ×ω' i , S i is the state variable weight factor of the i-th indicator, and its specific meaning is as follows:

[0159]

[0160] Among them, max i Indicates the maximum reasonable value of the i-th indicator, min i Indicates the minimum reasonable value of the i-th indicator, a i It represents the indicator data of the i-th indicator, and t represents the preset penalty factor. The larger t is, the greater the penalty is.

[0161] Finally, the TOPSIS model is used to calculate the Euclidean distance between each evaluation indicator of this power emergency simulation and the positive ideal solution:

[0162]

[0163]

[0164] The Euclidean distance between each evaluation indicator and the negative ideal solution is:

[0165]

[0166]

[0167] The dynamic evaluation closeness C of the power emergency simulation is calculated as follows:

[0168]

[0169] The evaluation results show that C < 0.5, indicating that the evaluation index is far from the ideal solution and that power emergency simulation needs to be improved. The evaluation index values ​​for the three different time scales are all below 0.5, which also shows that the current irregular dynamic power emergency simulation model has significant room for improvement. Based on the evaluation results of the power emergency simulation using the evaluation indicators at three different time scales, the power emergency simulation model is improved.

[0170] The above describes in detail the preferred embodiments of the present invention. It should be understood that those skilled in the art can make numerous modifications and variations based on the concepts of the present invention without inventive effort. Therefore, any technical solutions that can be derived by those skilled in the art through logical analysis, reasoning, or limited experimentation based on the concepts of the present invention and the prior art should be within the scope of protection defined by the claims.

Claims

1. A power emergency online deduction optimization method based on EMD method, characterized in that: The following steps are involved: S1. Construct short-time scale evaluation indicators, medium- and long-time scale evaluation indicators, and long-time scale evaluation indicators; S2. Obtain the index data of the power emergency online simulation and use the modal decomposition method to stabilize the index data; S3. Adopting an adaptive weight model to determine the weight of each indicator, and introducing a state variable weight factor to adjust the weight; S4. Combine the indicator weights and indicator data, use the Topsis model to evaluate the power emergency online simulation, and improve the power emergency online simulation based on the evaluation results; The short-time scale evaluation indicator is the elasticity index, which has the following specific meanings: Where K(t) represents the elasticity index of the power system at time t, t0 is a moment when the power system is not disturbed, that is, the moment when the power system operates normally; t d is the moment when the power system is in the worst operating state after being disturbed; ΔP R (t) represents the degree of imbalance between supply and demand of the power system at time t, and: in, is the power generation of the thermal power generating unit in the power system at time t; is the power generation of the wind turbine in the power system at time t; is the power generation of the photovoltaic device in the power system at time t; is the power generation of the storage device in the power system at time t; is the total load of the power system at time t; The medium- and long-term evaluation indicator is the dispatch coefficient. The dispatch resources of the power system include the transmission lines between the upper and lower power grids, controllable loads, and energy storage devices in the microgrid. The specific meaning of the dispatch coefficient is as follows: Among them, A(t) represents the dispatch coefficient of the power system at time t, ΔP MGi (t) represents the dispatch output of microgrid i in the power system at time t, λ MGi (t) represents the dispatch coefficient of microgrid i in the power system at time t; ΔP CL (t) represents the dispatch output of controllable loads in the power system at time t, λ CL (t) represents the dispatch coefficient of controllable load in the power system at time t; ΔP trans (t) represents the exchange dispatch output of the distribution network and the upper power grid in the power system at time t, λ trans represents the exchange scheduling coefficient of the power system, that is, the scheduling coefficient of the transmission line resources, μ trans (t) indicates whether the transmission line switch between the upper and lower power grids in the power system is closed at time t, and takes values ​​of 0 and 1; P cj (t) is the impact load change value of the power system at time t; The long-term evaluation index is a static calculation index fitted by multiple factors through the DEA method. The specific meaning is as follows: Among them, ρ represents the static calculation index under long time scale, c1 and c2 represent the operation cost and organization cost of power emergency online simulation respectively, and c sum is the total budget cost, e loss is the equipment loss, e total is the total number of electrical equipment, m loss is the consumption of emergency supplies, m total is the total amount of emergency supplies, p use is the actual number of personnel mobilized for the power emergency online simulation, p total is the total number of personnel participating in the power emergency online simulation, η is the system recovery percentage; The specific meaning of the dispatch coefficient of microgrid i in the power system at time t is as follows: Among them, ES i (t) represents the energy storage capacity of the energy storage device in microgrid i at time t, ES imax Indicates the maximum power storage capacity of the energy storage device in microgrid i, ES imin represents the minimum power reserve of the energy storage device in microgrid i, λ MGi,min represents the minimum dispatch coefficient allocated to microgrid i; Controllable load dispatch coefficient λ CL (t) and the exchange scheduling coefficient λ trans Take a constant, and λ MGi,min >λ CL (t)>λ trans ; In step S2, the modal decomposition method is used to stabilize the indicator data of the short-time scale evaluation indicator and the medium- and long-time scale evaluation indicator, including the following steps: S21. Obtain a dynamic indicator evaluation value and set it as the signal sequence X(t); S22. Find all extreme points of the signal sequence X(t), use spline interpolation to form a lower envelope for the minimum points and an upper envelope for the maximum points, and calculate the average envelope m(t) of the upper and lower envelopes; S23, subtracting the average envelope sequence m(t) from the signal sequence X(t) to obtain h(t), and determining whether h(t) satisfies the first preset condition. If so, proceed to step S24; otherwise, take h(t) as the new signal sequence X(t) and proceed to step S22; S24, taking h(t) as the IMF component of the signal sequence X(t), denotes that the signal sequence X(t) is subtracted from h(t) to obtain the residue R(t), and taking the residue R(t) as the new signal sequence X(t), executing step S22 until R(t) meets the second preset condition, and the signal sequence X(t) in step S21 is decomposed into the residue R(t) and multiple IMF components; S25. Calculate the average value of the sequence R(t), and use the average value as the index data after stabilization processing.

2. The power emergency online deduction optimization method based on the EMD method according to claim 1 is characterized in that: Step S3 is specifically as follows: The adaptive weight model is used to weight the three indicators and obtain the initial weights ω1', ω'2, ω3' of the short-time scale evaluation indicator, the medium- and long-time scale evaluation indicator, and the long-time scale evaluation indicator. The state-variable weight factor of each indicator is calculated and the initial weight is modified using the state-variable weight factor to obtain the weights ω1, ω2, ω3 of each indicator. Among them, ω i =S i ×ω i ', S i is the state variable weight factor of the i-th indicator, and its specific meaning is as follows: Among them, max i Indicates the maximum reasonable value of the i-th indicator, min i Indicates the minimum reasonable value of the i-th indicator, a i represents the indicator data of the i-th indicator, and t represents the preset penalty factor.

3. The power emergency online deduction optimization method based on the EMD method according to claim 1 is characterized in that: In step S4, the Topsis model is used to evaluate the power emergency online simulation as follows: Calculate the dynamic evaluation closeness C of the power emergency online simulation: Among them, C represents the dynamic evaluation closeness of the power emergency online simulation, Represents the indicator data of the i-th indicator, represents the positive ideal solution of the i-th index, represents the negative ideal solution of the i-th index, ω i represents the weight of the i-th indicator.

4. An EMD-based power emergency online deduction optimization system, characterized by: An electric power emergency online deduction optimization method based on the EMD method as described in any one of claims 1 to 3, comprising an indicator construction module, a data processing module, a weight determination module and an evaluation optimization module; The indicator construction module is used to construct short-time scale evaluation indicators, medium- and long-time scale evaluation indicators, and long-time scale evaluation indicators; The data processing module is used to obtain the index data of the power emergency online deduction and use the modal decomposition method to perform a stabilization process on the index data; The weight determination module adopts an adaptive weight model to determine the weight of each indicator and introduces a state variable weight factor to adjust the weight; The evaluation and optimization module combines the weights and index data of the indicators, uses the Topsis model to evaluate the online power emergency simulation, and optimizes the online power emergency simulation based on the evaluation results.

Images