Test method and test system of power system

Abstract

The invention relates to the technical field of electric power system testing, in particular to an electric power system testing method and system, and the method comprises the steps: determining a voltage stability vulnerability testing target and range, collecting and processing the multi-dimensional data of an electric power system, and constructing a digital twinborn model; carrying out power flow calculation by utilizing a power flow algorithm of an improved Newton-Raphson method, and training the digital twin model through historical data in combination with the dynamic load model to enable the digital twin model to converge; and designing normal operation, single equipment fault and extreme weather multi-working-condition test scenes, determining node voltage amplitude, voltage phase angle, voltage stability margin and reactive power distribution test indexes, establishing a vulnerability evaluation model, and evaluating the vulnerability of the nodes of the power system. According to the method, accurate evaluation of the node vulnerability of the power system is realized, and the problem of subjective evaluation index weight of the existing power system is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of power system testing, and in particular to a power system testing method and a testing system. Background Art

[0002] In power system testing, accurate assessment of voltage stability vulnerability is crucial to system safety. Traditional methods have problems such as insufficient characterization of load dynamic characteristics and low model accuracy. The weights of evaluation indicators often rely on subjective settings and lack a data-driven optimization mechanism. At the same time, test scenarios are mostly limited to normal operating conditions, making it difficult to cover complex operating conditions such as single equipment failures and extreme weather. This leads to insufficient accuracy in node vulnerability assessment and an inability to effectively respond to multiple risks in actual operation. Therefore, there is an urgent need to solve this problem by constructing a vulnerability assessment method for power system nodes that combines dynamic load models with data-driven optimization algorithms. Summary of the Invention

[0003] The present invention aims to solve the problems existing in the background technology and to provide a test method and a test system for an electric power system.

[0004] The technical solution of the present invention is a method for testing an electric power system, comprising the following steps: Determine the objectives and scope of voltage stability vulnerability testing, collect and process multi-dimensional data of the power system, and build a digital twin model; The construction of the digital twin model involves using the improved Newton-Raphson method to calculate the power flow, combining it with the dynamic load model and training the digital twin model with historical data to achieve convergence. Design test scenarios for normal operation, single equipment failure and extreme weather conditions, determine the node voltage amplitude, voltage phase angle, voltage stability margin and reactive power distribution test indicators, establish a vulnerability assessment model based on PSO optimized AHP, and evaluate the vulnerability value of power system nodes based on the vulnerability assessment model.

[0005] Preferably, the dynamic load model divides the load into constant power, constant impedance and other power system components, and its expression is as follows: ; Where, 、 are dynamic load active power and reactive power respectively; 、 are the constant power load active power and load reactive power respectively; 、 are the active power of constant resistance load and the reactive power of load respectively; 、 are the load active power and load reactive power based on the rest of the power system components, respectively.

[0006] Preferably, the power flow algorithm of the Newton-Raphson method is improved by combining the dynamic load model to obtain an improved power flow calculation equation. The expression of the improved power flow calculation equation is as follows: ; Where, 、 are the unbalanced amounts caused by changes in the total load active power and reactive power of the node; 、 are the injected active power and reactive power of node i respectively; 、 are the voltage amplitudes of nodes i and j respectively; 、 are the real and imaginary parts of the node admittance matrix elements respectively; , is the voltage phase angle difference between node i and node j; n is the total number of system nodes.

[0007] Preferably, a digital twin model is constructed based on the Jacobian matrix and integrated with the dynamic load model. The expression of the digital twin model is as follows: ; Where, 、 are the voltage phase angle increment and voltage amplitude increment of node i respectively; is the Jacobian matrix; 、 are the unbalanced quantities caused by the changes in the active power and reactive power of the dynamic load, respectively, and are calculated by the dynamic load model; k is the number of iterations, k is an integer, k≥0; where, 、 are the initial phase angle and voltage values ​​of node i respectively; Update the node voltage amplitude and phase angle as follows: .

[0008] Preferably, whether the convergence condition is met is determined, and the judgment expression of the convergence condition is as follows: ; If the conditional expression is established, the iteration ends and the voltage amplitude and phase angle of each node are output; otherwise, the digital twin model is returned to continue the iteration; where, It is the preset convergence accuracy, obtained based on fitting of historical big data.

[0009] Preferably, the method for determining the node voltage amplitude, voltage phase angle, voltage stability margin and reactive power distribution test index includes: The voltage stability margin is calculated using a calculation method based on singular value decomposition; The node voltage amplitude, voltage phase angle, voltage stability margin based on singular value decomposition and reactive power distribution are used as the first-level evaluation indicators and recorded as , each first-level evaluation indicator is subdivided into several second-level evaluation indicators; Based on the multiple secondary indicators of each primary evaluation indicator, the overall evaluation indicator system is constructed as follows: ; In the formula, n1, n2, n3 and n4 are the total number of secondary indicators of each first-level indicator.

[0010] Preferably, each group of indicators is regarded as a particle, and the secondary indicator statistics of the overall evaluation indicator system are performed. The total number of secondary indicators is calculated by the formula N=n1+n2+n3+n4, and is used as the dimension N of each particle; Particle h can be expressed as ;in, is the weight of the zth secondary index in particle h, and satisfies .

[0011] Preferably, historical fault cases are obtained, and the Spearman correlation coefficient between the vulnerability ranking in the historical fault cases and the actual fault severity is used as the objective function f(X) to measure the consistency between the evaluation results and the actual situation; The expression of the objective function is ; Where m is the number of historical failure cases; is the difference between the vulnerability ranking obtained based on the evaluation model and the actual fault severity ranking in the cth case; The objective function is optimized based on the PSO algorithm. The particle h is iteratively optimized and the speed of the particle h at the t+1 iteration is obtained using the update formula. and location , the update formula is as follows: ; ; Where, is the inertia weight, which is used to balance the global search and local search capabilities; 、 is the learning factor; 、 is a random number uniformly distributed in the interval [0,1]; is the individual optimal position of particle h at the tth iteration; is the global optimal position of the entire particle swarm at the tth iteration; The particle speed and position are continuously updated iteratively until the preset stop condition is met. The global optimal position is obtained at this time. This is the optimal indicator weight combination obtained by optimization; the preset stopping condition is reaching the maximum number of iterations or the convergence of the objective function value.

[0012] Preferably, for the indicator weight combination, the vulnerability value CZ of each node is calculated by weighted summation, and the calculation formula is: Where, is the weight of the node in the e-th secondary index; is the normalized value of the node under the e-th secondary index; Sort the nodes in descending order according to their vulnerability values ​​CZ to obtain the vulnerability ranking of each node based on the evaluation model under each fault case. The actual fault severity is ranked by inviting experts in the power system field to score based on historical fault cases and sorting the scores in descending order.

[0013] The present invention also discloses a power system test system, which applies the above-mentioned power system test method, specifically comprising: Data acquisition and processing module, used to determine the voltage stability vulnerability test objectives and scope, collect and process multi-dimensional data of the power system, and build a digital twin model; The construction of the digital twin model involves using the improved Newton-Raphson method to calculate the power flow, combining it with the dynamic load model and training the digital twin model with historical data to achieve convergence. The indicator determination and evaluation module is used to design multi-condition test scenarios for normal operation, single equipment failure and extreme weather, determine the node voltage amplitude, voltage phase angle, voltage stability margin and reactive power distribution test indicators, establish a vulnerability assessment model based on PSO optimized AHP, and evaluate the vulnerability value of power system nodes based on the vulnerability assessment model.

[0014] Compared with the prior art, the above technical solution of the present invention has the following beneficial technical effects: The present invention constructs a digital twin model that integrates a dynamic load model and an improved Newton-Raphson power flow algorithm, and improves the model accuracy by training historical data until convergence. It designs multi-operating condition test scenarios, combines the PSO-optimized AHP vulnerability assessment model, and optimizes the indicator weights using the Spearman correlation coefficient as the objective function to achieve accurate assessment of the vulnerability of power system nodes, solving the problem of subjective evaluation indicator weights in existing power systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1This is a layer structure diagram of the first embodiment of the present invention. DETAILED DESCRIPTION

[0016] Example 1, as Figure 1 As shown, the present invention proposes a method for testing a power system, comprising the following steps: Determine the objectives and scope of voltage stability vulnerability testing, collect and process multi-dimensional data of the power system, and build a digital twin model; Data collection includes power system connection structure, equipment parameters, load characteristics, historical operation and equipment failure history data. Data processing includes cleaning, screening, missing data completion, classification and normalization. The construction of the digital twin model involves using the improved Newton-Raphson method to calculate the power flow, combining it with the dynamic load model and training the digital twin model with historical data to achieve convergence. The dynamic load model divides the load into constant power, constant impedance and other power system components, and its expression is as follows: ; Where, 、 are dynamic load active power and reactive power respectively; 、 are the constant power load active power and load reactive power respectively; 、 are the active power of constant resistance load and the reactive power of load respectively; 、 are the load active power and load reactive power based on the rest of the power system components, respectively; The Newton-Raphson method's power flow algorithm is improved by combining it with the dynamic load model, and an improved power flow calculation equation is obtained. The expression of the improved power flow calculation equation is as follows; ; Where, 、 are the unbalanced amounts caused by changes in the total load active power and reactive power of the node; 、 are the injected active power and reactive power of node i respectively; 、 are the voltage amplitudes of nodes i and j respectively; 、 are the real and imaginary parts of the node admittance matrix elements respectively; , is the voltage phase angle difference between node i and node j; n is the total number of system nodes; The digital twin model is constructed based on the Jacobian matrix and integrated with the dynamic load model. The expression of the digital twin model is as follows: ; Where, 、 are the voltage phase angle increment and voltage amplitude increment of node i respectively; is the Jacobian matrix; 、 are the unbalanced quantities caused by the changes in the active power and reactive power of the dynamic load, respectively, and are calculated by the dynamic load model; k is the number of iterations, k is an integer, k≥0; where, 、 are the initial phase angle and voltage values ​​of node i respectively; Update the node voltage amplitude and phase angle as follows: ; Determine whether the convergence condition is met. The judgment expression of the convergence condition is as follows: ; If the conditional expression is established, the iteration ends and the voltage amplitude and phase angle of each node are output; otherwise, the digital twin model is returned to continue the iteration; where, It is the preset convergence accuracy, obtained based on historical big data fitting; Design test scenarios for multiple operating conditions, including normal operation, single equipment failure, and extreme weather conditions; determine the node voltage amplitude, voltage phase angle, voltage stability margin, and reactive power distribution test indicators; establish a vulnerability assessment model based on PSO optimized AHP; and evaluate the vulnerability values ​​of power system nodes based on the vulnerability assessment model; For example, single equipment failure conditions may include generator tripping, generator excitation system failure, transformer winding short circuit, transformer bushing damage, single-phase ground short circuit of transmission line, three-phase short circuit of transmission line fault types, etc.; extreme weather conditions may include high temperature weather, strong wind weather, heavy rain weather, etc.; Methods for determining node voltage amplitude, voltage phase angle, voltage stability margin, and reactive power distribution test indicators include: The voltage stability margin is calculated using a calculation method based on singular value decomposition. Specifically, the Jacobian matrix is ​​subjected to singular value decomposition to obtain the minimum singular value on the diagonal of the singular matrix. The voltage stability margin is calculated by calculating the ratio of the minimum singular value to the norm of the Jacobian matrix. It should be noted that singular value decomposition and norm calculation are existing technical means and will not be elaborated on here. The node voltage amplitude, voltage phase angle, voltage stability margin based on singular value decomposition and reactive power distribution are used as the first-level evaluation indicators and recorded as , each first-level evaluation index is subdivided into several second-level evaluation indexes; for example, for the voltage amplitude , which is subdivided into the average voltage amplitude , standard deviation , minimum value etc.; For voltage stability margin , which is subdivided into the minimum value of the voltage stability margin of each node , rate of change wait; Based on the multiple secondary indicators of each primary evaluation indicator, the overall evaluation indicator system is constructed as follows: ; In the formula, n1, n2, n3 and n4 are the total number of secondary indicators of each primary indicator; The method of establishing a vulnerability assessment model based on PSO optimized AHP specifically includes the following steps: Take each group of indicators as a particle, and perform secondary indicator statistics on the overall evaluation indicator system. The total number of secondary indicators is calculated using the formula N=n1+n2+n3+n4, and used as the dimension N of each particle. Particle h can be expressed as ;in, is the weight of the zth secondary index in particle h, and satisfies ; Obtain historical failure cases and use the Spearman correlation coefficient between the vulnerability ranking in the historical failure cases and the actual failure severity as the objective function f(X) to measure the consistency between the evaluation results and the actual situation; The expression of the objective function is ; Where m is the number of historical failure cases; is the difference between the vulnerability ranking obtained based on the evaluation model and the actual fault severity ranking in the cth case; The objective function is optimized based on the PSO algorithm. The particle h is iteratively optimized and the speed of the particle h at the t+1 iteration is obtained using the update formula. and location , the update formula is as follows: ; ; Where, is the inertia weight, which is used to balance the global search and local search capabilities; 、 is the learning factor, which is usually a constant greater than 0; 、 is a random number uniformly distributed in the interval [0,1]; is the individual optimal position of particle h at the tth iteration; is the global optimal position of the entire particle swarm at the tth iteration; The particle speed and position are continuously updated iteratively until the preset stop condition is met. The global optimal position is obtained at this time. This is the optimal indicator weight combination obtained by optimization; the preset stopping condition is reaching the maximum number of iterations or the convergence of the objective function value; For the indicator weight combination, the vulnerability value CZ of each node is calculated by weighted summation. The calculation formula is: Where, is the weight of the node in the e-th secondary index; is the normalized value of the node under the e-th secondary index; Sort the nodes in descending order according to their vulnerability values ​​CZ to obtain the vulnerability ranking of each node based on the evaluation model under each fault case. The actual fault severity is ranked by inviting experts in the power system field to score based on historical fault cases and sorting the scores in descending order. The scoring is achieved using a 9-scale method. The objective function provides the direction for the iterative updates of the PSO algorithm. The PSO algorithm searches for the optimal solution by maximizing the objective function. In the power system voltage stability vulnerability assessment model, the objective function is the Spearman correlation coefficient between the vulnerability ranking in historical fault cases and the actual fault severity. The PSO algorithm iteratively updates the particle speed and position, attempting to find the indicator weight combination that maximizes this correlation coefficient, which is the optimal solution. The vulnerability assessment model based on PSO optimized AHP achieves objective and accurate assessment and testing of power system voltage stability vulnerability through scientific weight determination and comprehensive evaluation.

[0017] In a second embodiment, the present invention provides a power system test system, which is applied to a power system test method provided in the first embodiment, and specifically includes: Data acquisition and processing module, used to determine the voltage stability vulnerability test objectives and scope, collect and process multi-dimensional data of the power system, and build a digital twin model; The construction of the digital twin model involves using the improved Newton-Raphson method to calculate the power flow, combining it with the dynamic load model and training the digital twin model with historical data to achieve convergence. The indicator determination and evaluation module is used to design multi-condition test scenarios for normal operation, single equipment failure and extreme weather, determine the node voltage amplitude, voltage phase angle, voltage stability margin and reactive power distribution test indicators, establish a vulnerability assessment model based on PSO optimized AHP, and evaluate the vulnerability value of power system nodes based on the vulnerability assessment model.

[0018] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A method for testing a power system, characterized in that: The following steps are involved: Determine the objectives and scope of voltage stability vulnerability testing, collect and process multi-dimensional data of the power system, and build a digital twin model; The construction of the digital twin model involves using the improved Newton-Raphson method to calculate the power flow, combining it with the dynamic load model and training the digital twin model with historical data to achieve convergence. Design test scenarios for normal operation, single equipment failure and extreme weather conditions, determine the node voltage amplitude, voltage phase angle, voltage stability margin and reactive power distribution test indicators, establish a vulnerability assessment model based on PSO optimized AHP, and evaluate the vulnerability value of power system nodes based on the vulnerability assessment model.

2. A method for testing a power system according to claim 1, characterized in that: The dynamic load model divides the load into constant power, constant impedance and other power system components, and its expression is as follows: ； Where, 、 are dynamic load active power and reactive power respectively; 、 are the constant power load active power and load reactive power respectively; 、 are the active power of constant resistance load and the reactive power of load respectively; 、 are the load active power and load reactive power based on the rest of the power system components, respectively.

3. A method for testing an electric power system according to claim 2, characterized in that: The Newton-Raphson method's power flow algorithm is improved by combining it with the dynamic load model, and an improved power flow calculation equation is obtained. The expression of the improved power flow calculation equation is as follows; ； Where, 、 are the unbalanced amounts caused by changes in the total load active power and reactive power of the node; 、 are the injected active power and reactive power of node i respectively; 、 are the voltage amplitudes of nodes i and j respectively; 、 are the real and imaginary parts of the node admittance matrix elements respectively; , is the voltage phase angle difference between node i and node j; n is the total number of system nodes.

4. A method for testing a power system according to claim 3, characterized in that: The digital twin model is constructed based on the Jacobian matrix and integrated with the dynamic load model. The expression of the digital twin model is as follows: ； Where, 、 are the voltage phase angle increment and voltage amplitude increment of node i respectively; is the Jacobian matrix; 、 are the unbalanced quantities caused by the changes in the active power and reactive power of the dynamic load, respectively, and are calculated by the dynamic load model; k is the number of iterations, k is an integer, k≥0; where, 、 are the initial phase angle and voltage values ​​of node i respectively; Update the node voltage amplitude and phase angle as follows: 。 5. A method for testing a power system according to claim 4, characterized in that: Determine whether the convergence condition is met. The judgment expression of the convergence condition is as follows: ; If the conditional expression is established, the iteration ends and the voltage amplitude and phase angle of each node are output; otherwise, the digital twin model is returned to continue the iteration; where, It is the preset convergence accuracy, obtained based on fitting of historical big data.

6. A method for testing an electric power system according to claim 2, characterized in that: Methods for determining node voltage amplitude, voltage phase angle, voltage stability margin, and reactive power distribution test indicators include: The voltage stability margin is calculated using a calculation method based on singular value decomposition; The node voltage amplitude, voltage phase angle, voltage stability margin based on singular value decomposition and reactive power distribution are used as the first-level evaluation indicators and recorded as , each first-level evaluation indicator is subdivided into several second-level evaluation indicators; Based on the multiple secondary indicators of each primary evaluation indicator, the overall evaluation indicator system is constructed as follows: ; In the formula, n1, n2, n3 and n4 are the total number of secondary indicators of each first-level indicator.

7. A method for testing an electric power system according to claim 5, characterized in that: Take each group of indicators as a particle, and perform secondary indicator statistics on the overall evaluation indicator system. The total number of secondary indicators is calculated using the formula N=n1+n2+n3+n4, and used as the dimension N of each particle. Particle h can be expressed as ;in, is the weight of the zth secondary index in particle h, and satisfies .

8. A method for testing an electric power system according to claim 7, characterized in that: Obtain historical failure cases and use the Spearman correlation coefficient between the vulnerability ranking in the historical failure cases and the actual failure severity as the objective function f(X) to measure the consistency between the evaluation results and the actual situation; The expression of the objective function is ; Where m is the number of historical failure cases; is the difference between the vulnerability ranking obtained based on the evaluation model and the actual fault severity ranking in the cth case; The objective function is optimized based on the PSO algorithm. The particle h is iteratively optimized and the speed of the particle h at the t+1 iteration is obtained using the update formula. and location , the update formula is as follows: ； ； Where, is the inertia weight, which is used to balance the global search and local search capabilities; 、 is the learning factor; 、 is a random number uniformly distributed in the interval [0,1]; is the individual optimal position of particle h at the tth iteration; is the global optimal position of the entire particle swarm at the tth iteration; The particle speed and position are continuously updated iteratively until the preset stop condition is met. The global optimal position is obtained at this time. This is the optimal indicator weight combination obtained by optimization; the preset stopping condition is reaching the maximum number of iterations or the convergence of the objective function value.

9. A method for testing a power system according to claim 8, characterized in that: For the indicator weight combination, the vulnerability value CZ of each node is calculated by weighted summation. The calculation formula is: Where, is the weight of the node in the e-th secondary index; is the normalized value of the node under the e-th secondary index; Sort the nodes in descending order according to their vulnerability values ​​CZ to obtain the vulnerability ranking of each node based on the evaluation model under each fault case. The actual fault severity is ranked by inviting experts in the power system field to score based on historical fault cases and sorting the scores in descending order.

10. A power system testing system, applied to a power system testing method according to any one of claims 1 to 9, characterized in that: Specifically include: Data acquisition and processing module, used to determine the voltage stability vulnerability test objectives and scope, collect and process multi-dimensional data of the power system, and build a digital twin model; The construction of the digital twin model involves using the improved Newton-Raphson method to calculate the power flow, combining it with the dynamic load model and training the digital twin model with historical data to achieve convergence. The indicator determination and evaluation module is used to design multi-condition test scenarios for normal operation, single equipment failure and extreme weather, determine the node voltage amplitude, voltage phase angle, voltage stability margin and reactive power distribution test indicators, establish a vulnerability assessment model based on PSO optimized AHP, and evaluate the vulnerability value of power system nodes based on the vulnerability assessment model.

Images