A method and system for constructing an evaluation index model of preventive control measures

Abstract

The application discloses a kind of methods and systems for constructing preventive control measure effectiveness evaluation index model, belong to power system analysis and control technical field.The method of the present application comprises: sampling or calculating each type of stability index before and after the preventive control measure of power system;According to the various stability indexes obtained by sampling or calculation, the effectiveness index of the preventive control measure of power system is constructed;According to the effectiveness of each type of stability index, the fitting parameters in the calculation model of the preventive control measure effectiveness index are adjusted, and the effectiveness evaluation index model of the preventive control measure for power system optimization is obtained.The weight parameters of the stability preventive control measure effectiveness index evaluation can be adjusted to obtain the evaluation index of the prepared preventive control measure, and the system frequency and voltage control capability can be mastered in order to strengthen the frequency and voltage control means in time.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power system analysis and control, and more particularly, to a method and system for constructing an effectiveness evaluation index model of preventive control measures. BACKGROUND

[0002] The dynamic security analysis of traditional power grid is driven by knowledge model. Researchers understand the problem mechanism based on causal logic, extract stability rules, and make decision judgments. The knowledge-driven method corresponds to clear physical concepts, and the results are relatively accurate and reliable, but it depends on accurate physical modeling and rigorous formula derivation. With the expansion of power grid scale, the diversification of operation mode, and the diversification of electrical information, for the complex system problems caused by the interaction of multiple types of equipment and multiple time scale controls, the traditional research method has the shortcomings of difficulty in constructing electronic model, high dependence on expert experience, long calculation time, etc., and it is difficult to meet the development needs of the current power grid.

[0003] In terms of difficulty in constructing electronic model: with the rapid development of the application field of power electronics technology, the complexity of power electronic system is further increased, which brings great challenges to the efficient and reliable operation of the system. The power electronic system is a typical multi-time scale nonlinear system, which involves electromagnetic characteristics and interaction mechanism under different time scales. Accurate modeling and simulation are generally electromagnetic simulation.

[0004] In terms of high dependence on expert experience: it is necessary to support a large amount of analysis and calculation work to screen out feasible schemes that meet various constraints from a large number of possible operation schemes. However, the current preventive control measures are calculated by experienced professionals based on long-term work experience and limited operation mode data. It is impossible to fully estimate and give preventive control measures for extreme operation modes that have not occurred in dispatching operation.

[0005] In terms of long calculation time: the electromagnetic simulation calculation tool for a large number of power electronic devices in new power systems takes a long time; for online mechanical and electrical simulation of power grid, the simulation time is 3-5 minutes. During this process, the power grid has already changed, and the existing traditional power grid online simulation cannot meet the time efficiency of giving preventive control strategies after power grid failure.

[0006] Although the neural network based on Euclidean data has achieved certain results in power flow analysis problems, it ignores the essence of power system power flow data as graph structure data. Some key calculations defined in the Euclidean domain, such as convolution calculation, are not suitable for processing irregular graph structure data. Therefore, the graph neural network (GNN) that can directly accept graph structure data of any node scale and any topology as input and realize end-to-end learning is more suitable for the needs of power flow analysis problems. SUMMARY

[0007] To solve the above problems, the application provides a method for constructing an effectiveness evaluation index model of preventive control measures, comprising the following steps:

[0008] sampling or calculating various stability indexes before and after the preventive control measures of the power system;

[0009] constructing an effectiveness index of the preventive control measures of the power system according to the various stability indexes obtained by sampling or calculation;

[0010] adjusting fitting parameters in a calculation model of the effectiveness index of the preventive control measures according to the effectiveness of the various stability indexes, to obtain an effectiveness evaluation index model of the preventive control measures for optimization of the power system.

[0011] Optionally, the various stability indexes obtained by sampling or calculation include a new energy short-circuit ratio stability index, a moment of inertia margin index, a transient frequency analysis index, a transient power angle analysis index, a transient voltage support capability index and a static voltage support index.

[0012] Optionally, the effectiveness index of the preventive control measures of the power system is constructed by:

[0013] classifying the sampled effectiveness indexes into stability positive indexes and stability negative indexes;

[0014] determining sub-item indexes of the effectiveness of the stability positive indexes and the stability negative indexes before and after the preventive control measures are implemented in the power grid fault;

[0015] determining the effectiveness evaluation index of the preventive control measures of the power system according to the sub-item indexes.

[0016] Optionally, the sub-item indexes include a short-circuit ratio preventive control effectiveness index, a transient frequency analysis index effectiveness index, a transient power angle stability margin effectiveness index, a transient voltage support capability preventive control measure effectiveness index and a static voltage support preventive control measure effectiveness index.

[0017] Optionally, a calculation formula of the effectiveness index is as follows:

[0018] Ei_cm=(η_scr+η_J+η_nfd+η_ηA+η_T+η_v) / 6

[0019] Wherein, Ei cm is the effectiveness index, η scr, η J, η nfd, η ηA, η T and η v are the short-circuit ratio prevention control effectiveness index, the short-circuit ratio prevention control effectiveness index, the transient frequency analysis index effectiveness index, the transient power angle stability margin effectiveness index, the transient voltage support capability prevention control measure effectiveness index and the static voltage support prevention control measure effectiveness index respectively.

[0020] Optionally, according to the effectiveness of various stability indexes, the fitting parameters in the calculation model of the prevention control measure effectiveness index are adjusted to obtain an effectiveness evaluation index model of the prevention control measure for power system optimization, including:

[0021] According to the change amount of the various stability indexes before and after the measure and the loss function of the power grid calculation model of the GNN constructed in advance, the weights of the power grid calculation model of the GNN are updated using the gradient descent method according to the calculated loss function, and the power grid calculation model of the GNN with updated weights is repeatedly trained;

[0022] According to the change amount of the various stability indexes of the prevention control measure determined by the power grid calculation model of the GNN after each training, the loss amount of the power grid calculation model of the GNN after each training is obtained;

[0023] According to the loss amount, the hyperparameters of the power grid calculation model of the GNN are adjusted to obtain an optimized calculation model of the power grid stability evaluation index effectiveness of the GNN, and the weight parameters of the calculation model of the prevention control measure effectiveness evaluation index are adjusted based on the optimized calculation model of the power grid stability evaluation index effectiveness of the GNN to obtain an effectiveness evaluation index model of the prevention control measure effectiveness evaluation index for power system optimization;

[0024] The effectiveness evaluation index model is used to determine the effectiveness evaluation index of the prevention control measure for power system optimization.

[0025] In still another aspect, the present application further provides a system for constructing a prevention control measure effectiveness evaluation index model, comprising:

[0026] A sampling and calculation unit is configured to sample or calculate various stability indexes before and after the prevention control measure of the power system;

[0027] A construction index unit is configured to construct a power system prevention control measure effectiveness index according to the various stability indexes obtained by sampling or calculation;

[0028] The model building unit is configured to adjust fitting parameters in a calculation model of the effectiveness index of the preventive control measure according to effectiveness of various stability indexes, so as to obtain an effectiveness evaluation index model of the preventive control measure for power system optimization.

[0029] Optionally, the sampled various stability indexes include a new energy short-circuit ratio stability index, a moment of inertia margin index, a transient frequency analysis index, a transient power angle analysis index, a transient voltage support capability index, and a static voltage support index.

[0030] Optionally, the effectiveness index of the preventive control measure of the power system includes:

[0031] The stability indexes are classified into stability positive indexes and stability negative indexes;

[0032] The effectiveness of the stability positive indexes and the stability negative indexes before and after the preventive control measure is implemented in the power grid fault is determined;

[0033] The effectiveness index of the preventive control measure of the power system is determined according to the sub-indexes.

[0034] Optionally, the sub-indexes include a short-circuit ratio preventive control effectiveness index, a transient frequency analysis index effectiveness index, a transient power angle stability margin effectiveness index, a transient voltage support capability preventive control measure effectiveness index, and a static voltage support preventive control measure effectiveness index.

[0035] Optionally, a calculation formula of the effectiveness index is as follows:

[0036] Ei_cm=(η_scr+η_J+η_nTd+η_ηA+η_T+η_v) / 6

[0037] wherein, Ei_cm is the effectiveness index, η_scr, η_J, η_nfd, η_ηA, η_T, and η_v are respectively the short-circuit ratio preventive control effectiveness index, the short-circuit ratio preventive control effectiveness index, the transient frequency analysis index effectiveness index, the transient power angle stability margin effectiveness index, the transient voltage support capability preventive control measure effectiveness index, and the static voltage support preventive control measure effectiveness index.

[0038] Optionally, the fitting parameters in the calculation model of the effectiveness index of the preventive control measure are adjusted according to effectiveness of various stability indexes, so as to obtain an effectiveness evaluation index model of the preventive control measure for power system optimization, including:

[0039] According to the change amount of the preventive control measures of each type of stability index before and after the measure, and the loss function calculated by the power grid calculation model of the GNN, the weight of the power grid calculation model of the GNN is updated by using the gradient descent method according to the calculated loss function, and the power grid calculation model of the GNN with the updated weight is repeatedly trained;

[0040] According to the change amount of the preventive control measures of each type of stability index determined by the power grid calculation model of the GNN after each training, the loss amount of the power grid calculation model of the GNN after each training is obtained;

[0041] According to the loss amount, the hyperparameters of the power grid calculation model of the GNN are adjusted to obtain an optimized power grid stability evaluation index effectiveness calculation model of the GNN, the weight parameters of the calculation model of the preventive control measure effectiveness evaluation index are adjusted based on the optimized power grid stability evaluation index effectiveness calculation model of the GNN, and an effectiveness evaluation index model of the preventive control measure effectiveness evaluation index for power system optimization is obtained.

[0042] The effectiveness evaluation index model is used to determine the effectiveness evaluation index of the preventive control measure for power system optimization.

[0043] In still another aspect, the present application also provides a computing device, comprising: one or more processors;

[0044] The processor is used to execute one or more programs;

[0045] When the one or more programs are executed by the one or more processors, the method as described above is implemented.

[0046] In still another aspect, the present application also provides a computer readable storage medium, which has a computer program stored thereon, and the computer program is executed to implement the method as described above.

[0047] Compared with the prior art, the present application has the following beneficial effects:

[0048] The present application provides an effectiveness evaluation index and evaluation method for optimizing preventive control measures of a power system, comprising: sampling or calculating various stability indexes before and after preventive control measures of the power system; constructing a preventive control measure effectiveness index of the power system according to the various stability indexes obtained by sampling or calculation; and adjusting fitting parameters in a calculation model of the preventive control measure effectiveness index according to the effectiveness of the various stability indexes, to obtain an effectiveness evaluation index model of the preventive control measures for optimization of the power system. By adjusting the weight parameters of the effectiveness evaluation model of the stability preventive control measures, the present application can obtain a prepared preventive control measure evaluation model, and thus can grasp the frequency and voltage control capability of the system, so as to timely strengthen the frequency and voltage control means. BRIEF DESCRIPTION OF DRAWINGS

[0049] Figure 1 is a flowchart of the method of the present application;

[0050] Figure 2 is a flowchart of the method embodiment of the present application;

[0051] Figure 3 is a schematic diagram of the model trained by the method embodiment of the present application;

[0052] Figure 4 is a structural diagram of the system of the present application. DETAILED DESCRIPTION

[0053] Reference will now be made to the drawings to describe the exemplary embodiments of the present application in detail. The present application can be implemented in various forms, and is not limited to the embodiments described herein, which are provided to fully and completely disclose the present application and to fully convey the scope of the present application to those skilled in the art. The terms used in the exemplary embodiments represented in the drawings are not limitations of the present application. In the drawings, the same elements / elements are denoted by the same reference numerals.

[0054] Unless otherwise defined, the terms used herein (including technical terms) have meanings commonly understood by those skilled in the art. In addition, it is to be understood that the terms defined in commonly used dictionaries are to be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art, and should not be interpreted in an idealized or overly formal sense.

[0055] Example 1:

[0056] The present application also proposes a method for constructing a preventive control measure effectiveness evaluation index model, as shown in Figure 1 , comprising:

[0057] Step 1, sampling or calculating various stability indexes before and after preventive control measures of the power system;

[0058] Step 2, according to the sampling or calculation of various stability indicators, the effectiveness of the power system preventive control measures index is constructed;

[0059] Step 3, according to the effectiveness of various stability indicators, the fitting parameters in the calculation model of the effectiveness of the preventive control measures are adjusted, and the effectiveness evaluation index model of the preventive control measures for power system optimization is obtained.

[0060] Among them, the sampling or calculation of various stability indicators includes: new energy short-circuit ratio stability index, moment of inertia margin index, transient frequency analysis index, transient power angle analysis index, transient voltage support capability index and static voltage support index.

[0061] Among them, the effectiveness index of the preventive control measures of the power system is constructed, including:

[0062] The effectiveness index is classified into: stability positive index and stability negative index;

[0063] The sub-item index of the effectiveness of the stability positive index and the stability negative index before and after the implementation of the preventive control measures in the power grid is determined;

[0064] According to the sub-item index, the effectiveness evaluation index of the preventive control measures of the power system is determined.

[0065] Among them, the sub-item index includes: short-circuit ratio preventive control effectiveness index, transient frequency analysis index effectiveness index, transient power angle stability margin effectiveness index, transient voltage support capability preventive control measures effectiveness index and static voltage support preventive control measures effectiveness index.

[0066] Among them, the calculation formula of the effectiveness index is as follows:

[0067] Ei_cm=(η_scr+η_J+η_nfd+η_ηA+η_T+η_v) / 6

[0068] Among them, Ei_cm is the effectiveness index, η_scr, η_J, η_nfd, η_ηA, η_T and η_v are respectively the short-circuit ratio preventive control effectiveness index, the short-circuit ratio preventive control effectiveness index, the transient frequency analysis index effectiveness index, the transient power angle stability margin effectiveness index, the transient voltage support capability preventive control measures effectiveness index and the static voltage support preventive control measures effectiveness index.

[0069] The fitting parameters in the calculation model of the effectiveness evaluation index of the preventive control measures are adjusted according to the effectiveness of various stability indexes, and a calculation model of the effectiveness evaluation index of the preventive control measures for power system optimization is obtained, including:

[0070] According to the change amount of the various stability indexes before and after the measures and the loss function of the power grid calculation model of the GNN constructed in advance, the weights of the power grid calculation model of the GNN are updated using the gradient descent method according to the calculated loss function, and the power grid calculation model of the GNN with updated weights is repeatedly trained.

[0071] The loss amount of the power grid calculation model of the GNN after each training is obtained according to the change amount of the various stability indexes of the preventive control measures determined by the power grid calculation model of the GNN after each training.

[0072] According to the loss amount, the hyperparameters of the power grid calculation model of the GNN are adjusted to obtain a calculation model of the effectiveness of the GNN-based power grid stability evaluation index, and the weight parameters of the calculation model of the effectiveness evaluation index of the preventive control measures are adjusted based on the calculation model of the effectiveness of the GNN-based power grid stability evaluation index to obtain a calculation model of the effectiveness evaluation index of the preventive control measures for power system optimization.

[0073] The effectiveness evaluation index model is used to determine the effectiveness evaluation index of the preventive control measures for power system optimization.

[0074] The application will be further described below in combination with specific embodiments:

[0075] The embodiment steps are as shown in Figure 4 Specifically, the embodiment steps include:

[0076] Step one, "sampling or calculating various stability indexes of the power grid", refers to calculating stability indexes of factors affecting the stability of the power grid, including new energy short-circuit ratio stability indexes, moment of inertia margin indexes, transient frequency division indexes, transient power angle analysis indexes, transient voltage support capability indexes, and static voltage support capability indexes.

[0077] Step two, "constructing the effectiveness index of the power grid preventive control measures", is the core of the application, and a power system stability preventive control measure evaluation index construction method is proposed. The objective function of the index can be used to online calculate and optimize the effectiveness index of the preventive control measures of the power system.

[0078] Step three "online calculation of evaluation and optimization of preventive control measures effectiveness" the method adopted in the application is to use artificial intelligence graph neural network to construct power grid calculation model, and use log function as loss function to train the optimization model of preventive control measures.

[0079] The following describes steps one to three:

[0080] Step one: sampling or calculating various stability indicators of the power grid, including:

[0081] New energy short-circuit ratio stability indicator:

[0082] The short-circuit ratio (SCR) of the power grid is an important indicator of the strength of the power grid, and the distance between it and the minimum short-circuit ratio (i.e. critical short-circuit ratio, CSCR) required by new energy equipment reflects the stability margin of the system. Short-circuit ratio calculation is simple and has clear physical meaning, providing a simple and intuitive method for stability analysis of new energy grid-connected systems.

[0083] The application gives the indicators that affect the stability of the power grid, and a five-dimensional power grid operation intelligent evaluation system is constructed, which includes new energy short-circuit ratio, minimum inertia demand after expected disturbance, rotational inertia margin, transient frequency, and transient voltage support capability

[0084] Short-circuit ratio refers to the ratio of voltage to current between two power supply terminals, used to describe the short-circuit resistance capability between new energy power generation stations. Short-circuit ratio is one of the indicators that measures the change, abnormality and short-circuit of load without affecting the normal work of new energy power generation.

[0085] Short-circuit ratio, also known as short-circuit capacity coefficient, is also the load short-circuit ratio or short-circuit resistance ratio of multi-field power stations, which represents the maximum current output by the unit under short-circuit conditions and the ratio of the total load current, usually represented by Kg:

[0086] SCR = Kg (short-circuit current / system active power) (2)

[0087] Rotational inertia margin indicator:

[0088] The system can realize panoramic monitoring of the rotational inertia of the synchronous power grid system by real-time sensing the operating state (grid-connected / shutdown) of the conventional units in the grid and combining the unit dynamic parameter library (unit inertia time constant). According to the start-up state of the conventional units in SCADA, the system rotational inertia can be calculated according to the following formula.

[0089]

[0090] In the formula: J ∑ is the system rotational inertia of the synchronous power grid; P Gi is the rated capacity of unit i; Inertia time constant of the machine i.

[0091] Transient frequency analysis index:

[0092] Transient frequency deviation acceptability margin ηfd is defined as:

[0093] η fd = [f ext - (f cr - kT cr )] x 100% (4)

[0094] where f cr and T cr are the frequency deviation threshold and the allowed duration of the bus, respectively, f ext is the extreme value of the bus frequency during the transient process, and k is the conversion factor from the critical frequency deviation duration to the frequency. ηfd is positive (or negative) value indicating that the frequency deviation is acceptable (or not acceptable).

[0095] For each binary table, the corresponding transient frequency deviation acceptability margin can be calculated. The transient frequency safety margin of a power grid needs to consider all the constraints in the binary table group, and the minimum value of the margin in all set disturbances is selected as the frequency deviation acceptability margin of the power grid.

[0096] Transient power angle analysis index:

[0097] 1) The maximum power angle difference of the computer group is determined by the computer, including:

[0098] The maximum power angle difference is defined as the maximum value between the generator power angle curves during the simulation process:

[0099]

[0100] where N F is the number of expected faults, δ tas.i is the maximum power angle difference under fault i, and δ tas.max is the maximum value of the maximum power angle difference required by the system. Where δ tas.max is an empirical value, and the power angle instability criterion in the current online system is set to 500°, which needs to be discussed.

[0101] 2) The transient power angle stability margin criterion includes:

[0102] Specifically, under a certain fault, the transient power angle stability margin formula is determined according to the properties of each swing. When the trajectory encounters the farthest point (FEP), the formula is When the trajectory encounters the dynamic saddle point (DSP), the formula is:

[0103]

[0104] where A inc is the area of the current swing kinetic energy increase; A dec is the area of the current swing kinetic energy decrease; A dec.pot is the virtual deceleration area of the stable swing. The trajectory stability margin of the original high-dimensional system is obtained by taking the minimum value of the trajectory stability margin of each swing in the candidate grouping mode and taking the minimum value of the trajectory stability margin of all candidate groupings.

[0105] The transient power angle stability margin is the transient power angle stability margin of the fault with the minimum margin among all faults.

[0106] The transient voltage support capability index:

[0107] The transient voltage recovery index of bus j after fault i is defined as:

[0108]

[0109] The transient voltage recovery index describes the area of bus voltage below 0.9 p.u. within 10s after fault clearance, which can quantify the severity of bus transient voltage drop. Under M expected faults, the transient voltage recovery index of bus j is defined as:

[0110]

[0111] To analyze the transient voltage stability of the whole network, the system transient voltage stability index is defined as follows:

[0112]

[0113] N is the number of key bus nodes, and ωj is the weight. The larger T is, the lower the level of system transient voltage stability is.

[0114] Static voltage support index:

[0115] The static voltage support index mainly includes the sensitivity index, singular value index, impedance module index, and load margin index.

[0116] The voltage sensitivity index takes the derivative of voltage with respect to power as the parameter of voltage sensitivity.

[0117] 1) Take the derivative of the load node voltage with respect to the power node voltage. If the load node voltage increases with the increase of the power node voltage, it indicates that the system is in a voltage stable state.

[0118] 2) Take the derivative of the load node voltage with respect to the reactive power demand of the node. If the load node voltage increases with the increase of the reactive power demand of the node, it indicates that the system is in a stable state.

[0119] 3) Take the derivative of the load node voltage to the active demand of the node, if the load node voltage increases with the decrease of the reactive demand of the node, it means that the system is in a stable state.

[0120] 4) Take the derivative of the reactive power generated by the power supply to the reactive demand of the load node, if the reactive power generated by the power supply increases with the increase of the reactive demand of the load, it means that the system is in a stable state.

[0121] 5) Use the minimum eigenvalue of the Jacobian matrix to judge the stability of the current operating state, or judge the distance between the current operating state and the critical state according to the difference between the eigenvalue and 0.

[0122] Step two: Build the effectiveness index of power grid preventive control measures, including:

[0123] Traditional control methods include primary frequency control, secondary frequency control, excitation system control, etc., to maintain the stability of the power grid. Primary frequency control uses reactive power regulation and frequency deviation adjustment to stabilize the grid frequency. Secondary frequency control adjusts the generator output to regulate the grid frequency and power flow, achieving regional grid balance. Excitation system control adjusts the generator excitation current to control the terminal voltage, reactive power output and steady-state stability.

[0124] In addition to the above traditional stability control methods, control measures also include generator tripping, load shedding and other control methods under different operating modes and power flow constraints after a fault.

[0125] The invention of this application proposes a method for constructing an evaluation index of power system stability preventive control measures, which can also be used to optimize the preventive control measures of the power system online.

[0126] Step one: Analyze the sub-indexes of the system, including:

[0127] The sub-indexes of system stability evaluation in step one are positive evaluation indexes and negative evaluation indexes. The positive evaluation index means that the larger the index, the higher the system stability. The negative evaluation index means that the larger the index, the worse the system stability.

[0128] 1) The new energy short circuit ratio stability index SCR represents the maximum short circuit current that the system can allow. The larger the index, the better the system stability, which is a positive stability index.

[0129] 2) The moment of inertia margin index J represents the system moment of inertia. The larger the index, the better the system stability, which is a positive stability index.

[0130] 3) Transient frequency analysis index ηTd is positive (or negative) value, which indicates that frequency deviation is acceptable (or unacceptable), so the index participates in the comprehensive stability evaluation index in the form of taking the sign of the data, and the expression is: sgin(ηTd).

[0131] 4) The greater the transient power angle stability margin, the higher the system stability, and the transient power angle stability margin η is a positive index of stability;

[0132] 5) The transient voltage support capability index T represents the work area required for the system to recover to the voltage level before the fault, and the greater the index, the worse the system stability, so the index is a reverse index of system stability;

[0133] 6) The static voltage support index is the sensitivity index of several voltage to system power, and the index is positive, which means that the system static voltage increases with the increase of power, and it is a positive index of stability.

[0134] In summary, the sub-index of the system, whether it is a positive index of stability or a reverse index of stability, needs to be larger.

[0135] The effectiveness index of the grid preventive control measure = the comprehensive stability evaluation index before the control measure after the fault - the comprehensive stability evaluation index after the control measure after the fault (12);

[0136] The so-called comprehensive stability evaluation index before the control measure after the fault means that after the power grid is disturbed, the comprehensive stability evaluation index of system stability is evaluated before the preventive control measure.

[0137] The so-called comprehensive stability evaluation index after the control measure after the fault means that after the power grid is disturbed, the comprehensive stability evaluation index of system stability is evaluated after the preventive control measure.

[0138] Based on this conclusion, the sub-index of the effectiveness of the preventive control measure is proposed:

[0139] 1) The preventive control effectiveness index of short-circuit ratio:

[0140] η_scr=(scr e -scr s ) / scr s (13)

[0141] 2) The preventive control effectiveness index of moment of inertia margin:

[0142] η_J=(Je-Js) / Js (14)

[0143] 3) The effectiveness index of transient frequency analysis index ηfd:

[0144] η_nfd = (η_nfd e -η_nfd s ) / η_nfd S (15)

[0145] 4) Transient power angle stability margin effectiveness index:

[0146] η_ηA = (η_ηA s -η_ηA e ) / η_ηA s (16)

[0147] 5) Effectiveness index of preventive control measures for transient voltage support capability:

[0148] η_T = (η_T e -η_T s ) / η_T s (17)

[0149] 6) Effectiveness index of preventive control measures for static voltage support:

[0150] η_v = (η_v e -η_v s ) / η_v s (18)

[0151] Step two: Based on the comprehensive evaluation index of system stability evaluation, the following effectiveness index of power system preventive control measures is proposed:

[0152] Ei_cm = (η_scr + η_J + η_nfd + η_ηA + η_T + η_v) / 6 (19)

[0153] Step three: Online construction of power grid calculation model of GNN:

[0154] (1) Based on the unsupervised dynamic edge convolutional graph neural network, the power flow optimization algorithm is realized, including:

[0155] 1) Data preprocessing;

[0156] The following steps are needed to vectorize the power flow data:

[0157] a Read the power flow data file;

[0158] b Establish the mapping of node name and node number, considering that the original node number in different batches of data will change, the number needs to be unified according to the mapping;

[0159] c Read the node file, and splice the node number, active power, reactive power and other data into the node feature vector V;

[0160] d Read the line file, establish the adjacency matrix A according to the bus number on both sides and the mapping table, and splice the resistance, reactance and other data into the edge feature vector E;

[0161] e G=(V, A, E) is the generated graph data.

[0162] Since it is unsupervised learning, there is no labeled data, and the loss function will be calculated by other methods.

[0163] 2) The structure of the graph model, the training schematic is shown in Figure 2 ;

[0164] Want to output the optimized power flow, can consider the model output optimization result compared with the change of input power flow AX, then the optimization result X' can be expressed as follows:

[0165] ΔX=F(X;W) (20)

[0166] X'=X+ΔX (21)

[0167] Where F(X) is the mapping from input to output of the model, W is the parameter of the model, that is, the optimization power flow calculation result is:

[0168] X'=X+F(X;W) (22)

[0169] The model consists of an input layer, an output layer and multiple ECC layers, as shown in the following figure:

[0170] Where N is the number of intermediate ECC layers. The input and output mapping of a single ECC layer can be expressed as:

[0171] f(X)=f(V,A,E;w) (23)

[0172] Where w is the parameter of the ECC layer, and the input and output mapping of the whole model is:

[0173] F(X;W)=F(V,A,E;W)=output(f N (V,A,E;w N )) (24)

[0174] Where output is the output layer, f n is the input and output mapping of the nth ECC layer, that is:

[0175] f n (V,A,E;w n ) = f(f n-1 (V,A,E;w n-1 );w n ) (25)

[0176] In particular, when n = 1, V, A, E is the output of the input layer, that is:

[0177] f0(V, A, E; w0) = input(V, A, E) (26)

[0178] 3) loss function;

[0179] The power flow optimization problem can be expressed in the following form:

[0180]

[0181] where C(X) is the objective function, G i (X) and H i (X) are constraint conditions. The traditional algorithm directly solves the optimization problem according to the constraints by mathematical method. In order to enable the neural network model to fit this nonlinear relationship, it is necessary to convert the objective function and constraints into loss functions to update the weights of the model. The loss function can be expressed in the following form:

[0182]

[0183] where X' is the result of the optimization power flow calculation, λ i , μ i ≥ 0 are weight parameters, τ i (X) and can be expressed in the following form:

[0184]

[0185] where t > 0 is a parameter, the higher the value, the better the fitting effect.

[0186] Step four: online calculation of preventive control measures for the optimized power system

[0187] In each batch of training process, according to the output of the model, the optimized power flow data is obtained, the loss function is calculated using formula (28) for the power flow data, and the model weight is updated using gradient descent method. Every time a generation is trained, the evaluation data set is input into the model to obtain the evaluation set power flow optimization result, which is calculated using formula (19) as the performance index of the model. Observe the performance index, manually adjust the hyperparameters of the model, and optimize the model.

[0188] The existing dispatching control system has large online monitoring data and many online analysis functions, but lacks scientific, reasonable and effective summary and extraction for the safety and stability of the AC-DC hybrid large power grid and the new energy consumption problem, cannot comprehensively evaluate the safety and stability of the power electronic AC-DC hybrid large power grid online, and lacks a scientific, systematic and reasonable power grid operation key indicator for real-time dispatching operation of the large power grid. The present application serves to build an evaluation system for the prevention and control measures of the large power grid dispatching operation stability based on the power grid operation data, comprehensively utilizes the multi-source data and various analysis results in the dispatching control system, develops a large power grid dispatching operation indicator evaluation software based on the power grid operation data, realizes the acquisition, calculation and visual display of the key dispatching operation indicators of the large power grid, so as to provide integrated one-stop operation display and decision support for the power grid dispatching operation personnel. The present application mainly solves the following key problems:

[0189] Firstly, through the effectiveness evaluation of the stability prevention and control measures, the global influence of the fault on the power grid can be prevented. The present application comprehensively evaluates the effectiveness of the prevention and control measures by using the stability indicators of transient state, steady state, frequency and voltage, gives exact evaluation indicators for the complex power system stability form, and further constructs the effectiveness evaluation indicators of the prevention and control measures.

[0190] Secondly, the short-circuit ratio index in the present application considers the units of the power electronic control device of the wind farm. A large number of power electronic equipment access makes the power electronic degree of the power system increase day by day, the stability form after system failure is more complex, the influence range is greatly expanded, and the risk is continuously increased. Therefore, the present application studies the stability indicators and control measure effectiveness indicators of the power electronic AC-DC hybrid large power grid safety and stability online evaluation, so that the dispatching operation personnel can accurately master the safety and stability level of the large power grid, identify the weak link of the power grid safety and stability, and master the system frequency and voltage control capability, so as to timely strengthen the frequency and voltage control means.

[0191] Thirdly, the present application uses the GNN graph neural network algorithm, improves the graph attention mechanism combined with the physical characteristics of the power system in view of the defect that the classical graph convolution and graph pooling method do not consider edge information, proposes a double-view graph attention network (DGAT), and applies it to the graph pooling calculation, and proposes a graph self-attention aggregation pooling method (GSAPool).

[0192] Embodiment 2:

[0193] The present application also proposes a system 200 for constructing a prevention and control measure effectiveness evaluation index model, as shown in Figure 4 ​

[0194] a sampling and calculation unit 201 configured to sample or calculate various stability indexes before and after a preventive control measure of a power system;

[0195] a construction index unit 202 configured to construct a preventive control measure effectiveness index of the power system according to the various stability indexes sampled or calculated;

[0196] a model building unit 203 configured to adjust fitting parameters in a calculation model of the preventive control measure effectiveness index according to effectiveness of the various stability indexes, to obtain an effectiveness evaluation index model of the preventive control measure for power system optimization.

[0197] The various stability indexes sampled include a new energy short-circuit ratio stability index, a moment of inertia margin index, a transient frequency analysis index, a transient power angle analysis index, a transient voltage support capability index, and a static voltage support index.

[0198] The preventive control measure effectiveness index of the power system includes:

[0199] The stability indexes sampled are classified into stability positive indexes and stability negative indexes.

[0200] The effectiveness of the stability positive indexes and the stability negative indexes before and after the preventive control measure is implemented in the power grid fault is determined.

[0201] The preventive control measure effectiveness index of the power system is determined according to the sub-indexes.

[0202] The sub-indexes include a short-circuit ratio preventive control effectiveness index, a transient frequency analysis index effectiveness index, a transient power angle stability margin effectiveness index, a transient voltage support capability preventive control measure effectiveness index, and a static voltage support preventive control measure effectiveness index.

[0203] The calculation formula of the effectiveness index is as follows:

[0204] Ei_cm=(η_scr+η_J+η_nfd+η_ηA+η_T+η_v) / 6

[0205] Ei_cm is the effectiveness index, and η_scr, η_J, η_nfd, η_ηA, η_T, and η_v are respectively the short-circuit ratio preventive control effectiveness index, the short-circuit ratio preventive control effectiveness index, the transient frequency analysis index effectiveness index, the transient power angle stability margin effectiveness index, the transient voltage support capability preventive control measure effectiveness index, and the static voltage support preventive control measure effectiveness index.

[0206] The fitting parameters in the calculation model of the effectiveness evaluation index of the preventive control measures are adjusted according to the effectiveness of various stability indexes, and an effectiveness evaluation index model of the preventive control measures for power system optimization is obtained, including:

[0207] According to the change amount of the various stability indexes before and after the measures and the loss function of the power grid calculation model of the GNN, the weights of the power grid calculation model of the GNN are updated using the gradient descent method according to the calculated loss function, and the power grid calculation model of the GNN with updated weights is repeatedly trained.

[0208] The loss amount of the power grid calculation model of the GNN after each training is obtained according to the change amount of the various stability indexes of the preventive control measures determined by the power grid calculation model of the GNN after each training.

[0209] According to the loss amount, the hyperparameters of the power grid calculation model of the GNN are adjusted to obtain an optimized calculation model of the effectiveness of the power grid stability evaluation index of the GNN, and the weight parameters of the calculation model of the effectiveness evaluation index of the preventive control measures are adjusted based on the optimized calculation model of the effectiveness of the power grid stability evaluation index of the GNN to obtain an effectiveness evaluation index model of the effectiveness evaluation index of the preventive control measures for power system optimization.

[0210] The effectiveness evaluation index model is used to determine the effectiveness evaluation index of the preventive control measures for power system optimization.

[0211] By adjusting the weight parameters of the effectiveness index, the prepared preventive control measures can be obtained, and the system frequency and voltage regulation capability can be mastered in order to timely strengthen the frequency and voltage control means.

[0212] Embodiment 3:

[0213] Based on the same inventive concept, the present application further provides a computer device, which comprises a processor and a memory, the memory is used to store a computer program, the computer program comprises program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., which are the computing core and control core of the terminal, and are suitable for implementing one or more instructions, and are specifically suitable for loading and executing one or more instructions in the computer storage medium to implement a corresponding method flow or a corresponding function, so as to implement the steps of the method in the above embodiments.

[0214] Embodiment 4:

[0215] Based on the same inventive concept, the present application further provides a storage medium, specifically a computer readable storage medium (Memory), which is a memory device in the computer device, and is used to store programs and data. It can be understood that the computer readable storage medium herein can include the built-in storage medium in the computer device, and of course can also include the expansion storage medium supported by the computer device. The computer readable storage medium provides a storage space, and the storage space stores the operating system of the terminal. Moreover, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and the instructions can be one or more computer programs (including program codes). It should be noted that the computer readable storage medium herein can be a high-speed RAM memory, or a non-volatile memory such as at least one disk memory. One or more instructions stored in the computer readable storage medium can be loaded and executed by the processor to implement the steps of the method in the above embodiments.

[0216] The present application is described in reference to the flowchart illustrations and / or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart illustrations and / or block diagrams block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 one or more flowcharts and / or blocks

[0217] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the flowchart illustrations and / or block diagrams block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 one or more flowcharts and / or blocks

[0218] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the flowchart illustrations and / or block diagrams block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 one or more flowcharts and / or blocks

[0219] While the preferred embodiments of the application have been described, additional variations and modifications can be made to the embodiments by those of skill in the art once they have the benefit of the present disclosure without departing from the spirit and scope of the application. Accordingly, the attached claims are intended to embrace all such variations and modifications as fall within the scope of the application.

[0220] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described.

Claims

1. A method of constructing a model of an effectiveness evaluation index of preventive control measures, characterized by, The method comprises the following steps: sampling or calculating various stability indexes before and after the preventive control measures of the power system; constructing the effectiveness index of the preventive control measures of the power system according to the various stability indexes obtained by sampling or calculation; adjusting the fitting parameters in the calculation model of the effectiveness index of the preventive control measures according to the effectiveness of the various stability indexes, to obtain the effectiveness evaluation index model of the preventive control measures for the optimization of the power system; the construction of the effectiveness index of the preventive control measures of the power system comprises: classifying the sampled stability indexes into stability positive indexes and stability negative indexes; determining the sub-indexes of the stability positive indexes and the stability negative indexes before and after the preventive control measures are implemented in the power grid fault; determining the effectiveness evaluation index of the preventive control measures of the power system according to the sub-indexes; the sub-indexes comprise the preventive control effectiveness index of short-circuit ratio, the effectiveness index of transient frequency analysis index, the effectiveness index of transient power angle stability margin, the effectiveness index of preventive control measures of transient voltage support capability and the effectiveness index of preventive control measures of static voltage support; the calculation formula of the effectiveness index is as follows: Ei_cm=(η_scr+η_J+η_nfd+η_ηA+η_T+η_v) / 6 wherein Ei_cm is the effectiveness index, η_scr, η_J, η_nfd, η_ηA, η_T and η_v are respectively the preventive control effectiveness index of short-circuit ratio, the effectiveness index of transient frequency analysis index, the effectiveness index of transient power angle stability margin, the effectiveness index of preventive control measures of transient voltage support capability and the effectiveness index of preventive control measures of static voltage support; the adjustment of the fitting parameters in the calculation model of the effectiveness index of the preventive control measures according to the effectiveness of the various stability indexes to obtain the effectiveness evaluation index model of the preventive control measures for the optimization of the power system comprises: according to the change amount of the various stability indexes before and after the measures and the loss function of the power grid calculation model of the GNN constructed in advance, the weight of the power grid calculation model of the GNN is updated by using the gradient descent method according to the calculated loss function, and the power grid calculation model of the GNN with the updated weight is repeatedly trained; the loss amount of the power grid calculation model of the GNN after each training is obtained according to the change amount of the various stability indexes of the preventive control measures determined by the power grid calculation model of the GNN after each training; the hyperparameters of the power grid calculation model of the GNN are adjusted according to the loss amount, to obtain the calculation model of the effectiveness of the power grid stability evaluation index of the optimized GNN, the weight parameters of the calculation model of the effectiveness evaluation index of the preventive control measures are adjusted based on the calculation model of the effectiveness of the power grid stability evaluation index of the optimized GNN, to obtain the effectiveness evaluation index model of the effectiveness evaluation index of the preventive control measures for the optimization of the power system; the effectiveness evaluation index model is used to determine the effectiveness evaluation index of the preventive control measures for the optimization of the power system.

2. The method of claim 1, wherein, The sampling or calculation of each type of stability index includes: a new energy short-circuit ratio stability index, a rotational inertia margin index, a transient frequency analysis index, a transient power angle analysis index, a transient voltage support capability index, and a static voltage support index.

3. A system for constructing a model of an effectiveness evaluation index of preventive control measures, characterized by, The method comprises: a sampling and calculation unit configured to sample or calculate each type of stability index before and after the preventive control measure of the power system; a construction index unit configured to construct a preventive control measure effectiveness index of the power system according to each type of stability index sampled or calculated; a model building unit configured to adjust fitting parameters in a calculation model of the preventive control measure effectiveness index according to the effectiveness of each type of stability index, and obtain an effectiveness evaluation index model of the preventive control measure for power system optimization; The preventive control measure effectiveness index of the power system comprises: classifying the sampled stability index into a stability positive index and a stability negative index; determining sub-item indexes of the stability positive index and the stability negative index before and after the preventive control measure is implemented in the power grid fault; determining the preventive control measure effectiveness index of the power system according to the sub-item indexes; The sub-item indexes include: a short-circuit ratio preventive control effectiveness index, a transient frequency analysis index effectiveness index, a transient power angle stability margin effectiveness index, a transient voltage support capability preventive control measure effectiveness index, and a static voltage support preventive control measure effectiveness index; The calculation formula of the effectiveness index is as follows: Ei_cm=(η_scr+η_J+η_nfd+η_ηA+η_T+η_v) / 6 wherein Ei_cm is the effectiveness index, η_scr, η_J, η_nfd, η_ηA, η_T, and η_v are respectively the short-circuit ratio preventive control effectiveness index, the transient frequency analysis index effectiveness index, the transient power angle stability margin effectiveness index, the transient voltage support capability preventive control measure effectiveness index, and the static voltage support preventive control measure effectiveness index; The method of adjusting the fitting parameters in the calculation model of the preventive control measure effectiveness index according to the effectiveness of each type of stability index, and obtaining the effectiveness evaluation index model of the preventive control measure for power system optimization comprises: According to the change amount of each type of stability index before and after the measure and the loss function of the power grid calculation model of the GNN constructed in advance, the weight of the power grid calculation model of the GNN is updated using the gradient descent method according to the calculated loss function, and the power grid calculation model of the GNN with the updated weight is repeatedly trained; According to the change amount of each type of stability index of the preventive control measure determined by the power grid calculation model of the GNN after each training, the loss amount of the power grid calculation model of the GNN after each training is obtained. According to the loss amount, the hyperparameters of the power grid calculation model of the GNN are adjusted to obtain an optimized GNN power grid stability evaluation index effectiveness calculation model, and the weight parameters of the preventive control measure effectiveness evaluation index calculation model are adjusted based on the optimized GNN power grid stability evaluation index effectiveness calculation model to obtain an effectiveness evaluation index model of the preventive control measure effectiveness evaluation index for power system optimization. The effectiveness evaluation index model is used to determine the effectiveness evaluation index of the preventive control measure for power system optimization.

4. The system of claim 3, wherein, The sampled various stability indexes include a new energy short-circuit ratio stability index, a rotational inertia margin index, a transient frequency analysis index, a transient power angle analysis index, a transient voltage support capability index, and a static voltage support index.

5. A computer device, comprising: The method comprises the following steps: one or more processors; a processor configured to execute one or more programs; when the one or more programs are executed by the one or more processors, the method as claimed in any one of claims 1-2 is implemented.

6. A computer-readable storage medium, characterized in that, a computer program is stored thereon, and the computer program is executed to implement the method as claimed in any one of claims 1-2.

Images