Method for identifying starting process of phase modifier in grid connection process of power system

Abstract

The invention belongs to the field of power system grid connection, and particularly relates to a method for identifying the starting process of a phase modifier in the power system grid connection process, which mainly comprises the following steps of: collecting operation data of a power system, repairing the collected data, and comprehensively analyzing the processed operation data by using an improved feedforward neural network; and optimizing the processed operation data of the power system by using a genetic algorithm, and transmitting the optimized operation data to a motor driver of the phase modifier through a driving circuit to realize real-time control and dynamic adjustment of the phase modifier. And through comprehensive judgment of various node information, the accuracy and reliability of startup identification of the phase modifier are improved.

Description

Technical Field

[0001] This invention belongs to the field of power system grid connection, specifically relating to a method for identifying the start-up process of a synchronous condenser during power system grid connection. Background Technology

[0002] In existing power system grid connection processes, the identification methods for synchronous condenser (SCDC) startup mainly rely on traditional circuit breaker position contact information and electrical quantity monitoring. That is, when the circuit breaker position is normal, the circuit breaker position is used as the primary criterion for grid connection. However, when the circuit breaker position is abnormal, the identification method often relies on single electrical quantity information, such as voltage, frequency, or current. This method is easily affected by electrical noise, resulting in low startup identification accuracy and the risk of protection malfunctions and fault failures. Especially when complex faults occur in the power system, the rapid and accurate startup of SCDCs is crucial for maintaining system stability and preventing fault escalation. Therefore, providing a novel method for identifying the startup process of SCDCs during power system grid connection to improve the accuracy and reliability of startup identification is an urgent problem to be solved in the current power system technology field. Summary of the Invention

[0003] This invention provides a method for identifying the start-up process of a synchronous condenser during power system grid connection, which solves the technical problem of the single identification method when the position status of a circuit breaker is abnormal.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for identifying the start-up process of a synchronous condenser during power system grid connection, comprising the following steps:

[0005] Step 1: Collect and monitor the operation data of the power system, including electrical quantity information such as circuit breaker position status S, voltage u, frequency f, and current I, as well as other auxiliary contact information, including electrical equipment fault alarm e, start-up protection m, and motor load protection L;

[0006] Step 2: Repair and process the collected power system operation data;

[0007] Step 3: Perform anomaly detection on the operating data of the repaired power system. Anomalies include values ​​that exceed the normal range, sudden changes, and discontinuous data points.

[0008] Step 4: Based on the normal or abnormal state of the processed power system operation data, different identification strategies are adopted. Under normal conditions, the circuit breaker position is used as the main criterion; under abnormal conditions, an improved feedforward neural network is used to comprehensively analyze the processed power system operation data.

[0009] Step 5: Optimize the processed power system operation data using an improved genetic algorithm to find the optimal parameter combination. At the same time, based on the results of different identification strategies, output corresponding synchronous condenser start-up signals or control commands to control the stable grid-connected operation of the power system.

[0010] Furthermore, the process of repairing power system operation data includes using linear interpolation to repair missing data with small time intervals and using mean smoothing to repair distorted data.

[0011] Furthermore, linear interpolation is used to repair missing data with small time intervals:

[0012]

[0013] In the formula: x t and x t-j The data parameters collected at time t and time tj are respectively; x t-n ω represents the missing data parameter at time ti. i Let i be the weight value at time i.

[0014] Use mean smoothing to repair distorted data:

[0015] or

[0016] In the formula, and These are the thresholds for the error between adjacent data collections, where...

[0017]

[0018] Furthermore, anomaly detection is performed on the operational data of the repaired power system:

[0019] make like Then x t ,t∈{1,23,...} represents an abnormal state, if Then x t ,t∈{1,23,...} represents the normal state.

[0020] Furthermore, the specific implementation process of the improved feedforward neural network is as follows:

[0021] Let the network input I be M-dimensional: x = [x1, x2, ..., x M ] T The hidden layer has K nodes, the output y is L-dimensional, and the length of the input-output sample pairs is N. The interaction function of the hidden layer nodes in the radial basis function network is the Gaussian function.

[0022]

[0023] The input data vector is mapped to the hidden layer, and the output of hidden layer node j is:

[0024]

[0025] In the formula: δ is the normalization constant of the hidden layer nodes, c is the Gaussian function center vector of the hidden layer nodes, and c j =[c j1 ,c j2 ,...,c jM ] T

[0026] The hidden layer to the output layer of the RBF network achieves a linear mapping of dimensional transformation, that is, the output of node k in the output layer is:

[0027]

[0028] In the formula: ω jk θ represents the adjustment weights from the hidden layer to the output layer. k This is the bias for output layer node k. k As a response to the corresponding input signal, the data is output to the workspace as an important control variable according to different backgrounds. The input x is the collected data feature, and the output y is the state feature.

[0029] Furthermore, the mathematical definition of the improved genetic algorithm is:

[0030] BP: Circuit breaker position status, BP∈{0,1};

[0031] V: Voltage, a continuous variable, 100V≤V≤240V;

[0032] F: Frequency, a continuous variable, 50Hz≤f≤60Hz;

[0033] I: Current, a continuous variable, 1A≤I≤100A;

[0034] AF: Auxiliary contact information, fault alarm, start-up protection status, motor load protection status.

[0035] Furthermore, the improved genetic algorithm is defined as follows: considering maximizing the stability and efficiency of the power system while ensuring safety, the objective function F... obj as follows:

[0036] F obj =α1·f stablity (BP,V,F,I,AF)+α2·f efficiency(BP,V,F,I,AF)-α3·f safetypenalty (AF):

[0037] f stability System stability calculated based on electrical parameters and auxiliary contact information;

[0038] f efficiency System efficiency calculated based on electrical parameters and auxiliary contact information;

[0039] f safetypenelty Safety penalty items calculated based on auxiliary contact information (such as fault alarms);

[0040] ω1, ω2, and ω3 are the weights for stability, efficiency, and security, respectively.

[0041] Furthermore, the specific implementation process of the improved genetic algorithm is as follows:

[0042] (1) Initialization: Randomly generate a set of individuals containing circuit breaker position status, voltage, frequency, current and auxiliary contact information. These solutions constitute the initial population.

[0043] (2) Fitness assessment: For each individual in the population, the objective function F is used. obj Calculate the fitness value;

[0044] (3) Selection: Based on the fitness value of an individual, a roulette wheel selection strategy is used to select superior individuals from the current population to enter the next generation;

[0045] (4) Crossover: Perform crossover operation on the selected individuals to generate new offspring individuals. The crossover operation selects a single point crossover.

[0046] (5) Mutation: Mutation operations are performed on offspring individuals to introduce new genetic information;

[0047] (6) Generate a new population: Combine the offspring individuals obtained after crossover and mutation with the parent individuals or a portion thereof to form a new population;

[0048] (7) Termination condition: If the preset maximum number of iterations is reached or the population fitness does not improve significantly for several consecutive generations, the algorithm stops; otherwise, return to step 2 to continue iterating.

[0049] (8) Output results: Output the optimal solution (i.e. the optimal parameter combination) found by the improved genetic algorithm, and adjust the operating state of the power system according to these parameter combinations.

[0050] Furthermore, the synchronous condenser also supports both single-position and dual-position grid-connected circuit breaker status access, which can adapt to the position access of various grid-connected circuit breakers.

[0051] Furthermore, in the single-position access mode, the power system monitors the position status of the grid-connected circuit breaker in real time and determines whether the start-up process has started based on the changes in the position status; in the dual-position access mode, in addition to monitoring the position status, the power system also determines whether the start-up process has started based on the relative positional relationship of the two position contacts.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] (1) When the circuit breaker position is normal, this invention uses the circuit breaker position as the primary criterion for grid connection, and combines it with electrical quantity information such as voltage, frequency, and current for auxiliary judgment to ensure the accuracy of start-up identification. When the circuit breaker position is abnormal, this invention uses electrical quantities such as voltage, frequency, and current, as well as other auxiliary contact information, as the primary criterion for grid connection. Through comprehensive judgment of multiple node information, the accuracy and reliability of synchronous condenser start-up identification are improved.

[0054] (2) This invention supports both single-position and dual-position access of the grid-connected circuit breaker status, which can adapt to the position access of various grid-connected circuit breakers and improve the applicability and flexibility of the method. Attached Figure Description

[0055] Figure 1 This is a flowchart of the method for identifying the startup process of a synchronous condenser during grid connection, as described in this invention. Detailed Implementation

[0056] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0057] like Figure 1 As shown, a method for identifying the start-up process of a synchronous condenser during power system grid connection includes the following steps:

[0058] Step 1: Collect and monitor the operation data of the power system, including electrical quantity information such as circuit breaker position status S, voltage u, frequency f, and current I, as well as other auxiliary contact information, including electrical equipment fault alarm e, start-up protection m, and motor load protection L;

[0059] Step 2: Repair the collected power system operation data to facilitate subsequent analysis;

[0060] The collected data includes: circuit breaker position status S t Voltage ut Frequency f t Current I t Electrical quantity information and other auxiliary contact information (including electrical equipment fault alarms) t Start-up protection m t Motor load protection L t ).

[0061] Data cleaning and repair methods: For missing data, linear interpolation is used to repair missing data with small time intervals.

[0062]

[0063] In the formula: x t and x t-j The data parameters collected at time t and time tj are respectively; x t-n For missing data parameters at time ti. ω i Let i be the weight value at time i.

[0064] For distorted data, since the data is collected continuously and transmitted sequentially, there will be no abrupt changes in data collected between adjacent time periods. If the range of variation in the collected data at a certain moment is greater than or less than 10% of the monitoring values ​​before and after it, the collected data is considered distorted. Mean smoothing can be used to address this.

[0065] or

[0066] In the formula, and These are the threshold values ​​for the error between adjacent data collections. In the formula...

[0067]

[0068] Step 3: Perform anomaly detection on the operating data of the repaired power system. Anomalies include values ​​that exceed the normal range, sudden changes, and discontinuous data points.

[0069] The goal of anomaly detection is to identify data that does not conform to the normal operating mode of the power grid system, as such abnormal data may affect subsequent analysis and judgment.

[0070] The anomaly detection process is as follows:

[0071]

[0072] make like Then x t ,t∈{1,23,...} represents an abnormal state, if Then xt ,t∈{1,23,...} represents the normal state.

[0073] Step 4: Based on the normal or abnormal state of the processed power system operation data, different identification strategies are adopted. Under normal conditions, the circuit breaker position is used as the main criterion; under abnormal conditions, an improved feedforward neural network is used to comprehensively analyze the processed power system operation data.

[0074] Let the network input I be M-dimensional: x = [x1, x2, ..., x M ] T The hidden layer has K nodes, the output y is L-dimensional, and the length of the input-output sample pairs is N. The interaction function of the hidden layer nodes in the radial basis function network is the Gaussian function.

[0075]

[0076] The input data vector is mapped to the hidden layer, and the output of hidden layer node j is:

[0077]

[0078] In the formula: δ is the normalization constant of the hidden layer nodes, c is the Gaussian function center vector of the hidden layer nodes, and c j =[c j1 ,c j2 ,...,c jM ] T

[0079] The hidden layer to the output layer of the RBF network achieves a linear mapping of dimensional transformation, that is, the output of node k in the output layer is:

[0080]

[0081] In the formula: ω jk θ represents the adjustment weights from the hidden layer to the output layer. k This is the bias for output layer node k. k As a response to the corresponding input signal, it is output to the workspace as an important control variable depending on different backgrounds.

[0082] The input x represents the collected data features, and the output y represents the state features. The specific network architecture is as follows.

[0083]

[0084] Step 5: Optimize the processed power system operation data using a genetic algorithm to find the optimal parameter combination. Simultaneously, based on the identification results, output the corresponding synchronous condenser start-up signal or control command to achieve stable grid-connected operation of the power system.

[0085] An improved genetic algorithm is used to identify parameters such as circuit breaker position status, voltage, frequency, current and other auxiliary contact information, and to provide mathematical models for circuit breaker position status S, voltage u, frequency f, current I and other auxiliary contact information.

[0086] BP: Circuit breaker position status, BP∈{0,1}

[0087] V: Voltage, a continuous variable, within a certain range, 100≤V≤240V.

[0088] F: Frequency, a continuous variable, within a certain range, 50Hz≤f≤60Hz

[0089] I: Current, a continuous variable, within a certain range, 1A≤I≤100A

[0090] AF: Auxiliary contact information, fault alarm (binary variable), start-up protection status (binary variable), motor load protection status (continuous or binary variable, depending on the specific implementation).

[0091] Objective function:

[0092] Consider maximizing system stability and efficiency while ensuring security. The objective function is F. obj as follows:

[0093] F obj =α1·f stablity (BP,V,F,I,AF)+α2·f efficiency (BP,V,F,I,AF)-α3·f safetypenalty (AF)

[0094] f stability System stability calculated based on electrical parameters and auxiliary contact information;

[0095] f efficiency System efficiency calculated based on electrical parameters and auxiliary contact information;

[0096] f safetypenelty Safety penalty items calculated based on auxiliary contact information (such as fault alarms);

[0097] ω1, ω2, and ω3 are the weights for stability, efficiency, and security, respectively.

[0098] Improved genetic algorithm process:

[0099] (1) Initialization: Randomly generate a set of individuals containing circuit breaker position status, voltage, frequency, current and auxiliary contact information. These solutions constitute the initial population.

[0100] (2) Fitness assessment: For each individual in the population, the objective function F is used. obj Calculate the fitness value;

[0101] (3) Selection: Based on the fitness value of an individual, a roulette wheel selection strategy is used to select superior individuals from the current population to enter the next generation;

[0102] (4) Crossover: Perform crossover operation on the selected individuals to generate new offspring individuals. The crossover operation selects a single point crossover.

[0103] (5) Mutation: Mutation operations are performed on offspring individuals to introduce new genetic information;

[0104] (6) Generate a new population: Combine the offspring individuals obtained after crossover and mutation with the parent individuals or a portion thereof to form a new population;

[0105] (7) Termination condition: If the preset maximum number of iterations is reached or the population fitness does not improve significantly for several consecutive generations, the algorithm stops; otherwise, return to step 2 to continue iterating.

[0106] (8) Output results: Output the optimal solution (i.e. the optimal parameter combination) found by the improved genetic algorithm, and adjust the operating state of the power system according to these parameter combinations.

[0107] To accommodate different types and specifications of grid-connected circuit breakers, this invention also supports both single-position and dual-position connection modes. Grid-connected circuit breakers with only one position contact use the single-position connection mode; while grid-connected circuit breakers with two position contacts use the dual-position connection mode.

[0108] In single-position access mode, the system monitors the position status of the grid-connected circuit breaker in real time and determines whether the start-up process has begun based on changes in the position status. In dual-position access mode, in addition to monitoring the position status, the system also determines whether the start-up process has begun based on the relative positional relationship of the two position contacts. This dual-judgment method further improves the accuracy and reliability of start-up identification.

[0109] By supporting both single-position and dual-position access for grid-connected circuit breakers, this invention can adapt to the position access requirements of various grid-connected circuit breakers, improving the applicability and flexibility of the method. Simultaneously, it also provides a more flexible and reliable access method for other equipment and systems within the power system.

[0110] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A method for identifying the start-up process of a synchronous condenser during power system grid connection, characterized in that, Includes the following steps: Step 1: Collect the operating data of the power system, including electrical quantity information such as circuit breaker position status S, voltage u, frequency f, and current I, as well as other auxiliary contact information, including electrical equipment fault alarm e, start-up protection m, and motor load protection L; Step 2: Repair and process the collected power system operation data; Step 3: Perform anomaly detection on the operating data of the repaired power system. Anomalies include values ​​that exceed the normal range, sudden changes, and discontinuous data points. Step 4: Based on the normal or abnormal state of the processed power system operation data, different identification strategies are adopted. Under normal conditions, the circuit breaker position is used as the main criterion. In abnormal conditions, an improved feedforward neural network is used to perform a comprehensive analysis of the processed power system operation data; Step 5: Optimize the processed power system operation data using an improved genetic algorithm to find the optimal parameter combination. At the same time, based on the results of different identification strategies, output corresponding synchronous condenser start-up signals or control commands to control the stable grid-connected operation of the power system.

2. The method for identifying the start-up process of a synchronous condenser during power system grid connection as described in claim 1, characterized in that, The process of repairing power system operation data includes using linear interpolation to repair missing data with small time intervals and using mean smoothing to repair distorted data.

3. The method for identifying the start-up process of a synchronous condenser during power system grid connection as described in claim 2, characterized in that, Linear interpolation is used to repair missing data with small time intervals. In the formula: x t and x t-j The data parameters collected at time t and time tj are respectively; x t-n ω represents the missing data parameter at time ti. i Let i be the weight value at time i. Use mean smoothing to repair distorted data: or In the formula, and These are the thresholds for the error between adjacent data collections, where...

4. The method for identifying the start-up process of a synchronous condenser during power system grid connection as described in claim 1, characterized in that, Perform anomaly detection on the operating data of the repaired power system: make like Then x t ,t∈{1,23,...} represents an abnormal state, if Then x t ,t∈{1,23,...} represents the normal state.

5. The method for identifying the start-up process of a synchronous condenser during power system grid connection as described in claim 1, characterized in that, The specific implementation process of the improved feedforward neural network is as follows: Let the network input I be M-dimensional: x = [x1, x2, ..., x M ] T The hidden layer has K nodes, the output y is L-dimensional, and the length of the input-output sample pairs is N. The interaction function of the hidden layer nodes in the radial basis function network is the Gaussian function. The input data vector is mapped to the hidden layer, and the output of hidden layer node j is: In the formula: δ is the normalization constant of the hidden layer nodes, c is the Gaussian function center vector of the hidden layer nodes, and c j =[c j1 ,c j2 ,...,c jM ] T The hidden layer to the output layer of the RBF network achieves a linear mapping of dimensional transformation, that is, the output of node k in the output layer is: In the formula: ω jk θ represents the adjustment weights from the hidden layer to the output layer. k The bias of output layer node k; y k As a response to the corresponding input signal, it is output to the workspace as an important control variable according to different backgrounds. The input x above is the collected data feature, and the output y is the state feature.

6. The method for identifying the start-up process of a synchronous condenser during power system grid connection as described in claim 1, characterized in that, The improved genetic algorithm includes the following parameters: BP: Circuit breaker position status, BP∈{0,1}; V: Voltage, a continuous variable, 100V≤V≤240V; F: Frequency, a continuous variable, 50Hz≤f≤60Hz; I: Current, a continuous variable, 1A≤I≤100A; AF: Auxiliary contact information, fault alarm, start-up protection status, motor load protection status.

7. The method for identifying the start-up process of a synchronous condenser during power system grid connection as described in claim 6, characterized in that, The improved genetic algorithm includes: an objective function F based on maximizing the stability, efficiency, and security of the power system. obj as follows: F obj =α1·f stablity (BP,V,F,I,AF)+α2·f efficiency (BP,V,F,I,AF)-α3·f safetypenalty (OF): f stability System stability calculated based on electrical parameters and auxiliary contact information; f efficiency System efficiency calculated based on electrical parameters and auxiliary contact information; f safetypenelty Safety penalty items calculated based on auxiliary contact information (such as fault alarms); ω1, ω2, and ω3 are the weights for stability, efficiency, and security, respectively.

8. The method for identifying the start-up process of a synchronous condenser during power system grid connection as described in claim 7, characterized in that, The specific implementation process of the improved genetic algorithm is as follows: (1) Initialization: Randomly generate a set of individuals containing circuit breaker position status, voltage, frequency, current and auxiliary contact information. These solutions constitute the initial population. (2) Fitness assessment: For each individual in the population, the objective function F is used. obj Calculate the fitness value; (3) Selection: Based on the fitness value of an individual, a roulette wheel selection strategy is used to select superior individuals from the current population to enter the next generation; (4) Crossover: Perform crossover on the selected individuals to generate new offspring individuals. The crossover operation is a single-point crossover. (5) Mutation: Mutation operations are performed on offspring individuals to introduce new genetic information; (6) Generate a new population: Combine the offspring individuals obtained after crossover and mutation with the parent individuals or a portion thereof to form a new population; (7) Termination condition: If the preset maximum number of iterations is reached or the population fitness does not improve significantly for several consecutive generations, the algorithm stops; otherwise, return to step 2 to continue iterating. (8) Output results: Output the optimal solution (i.e. the optimal parameter combination) found by the improved genetic algorithm, and adjust the operating state of the power system according to these parameter combinations.

9. The method for identifying the start-up process of a synchronous condenser during power system grid connection as described in claim 1, characterized in that, The synchronous condenser also supports both single-position and dual-position access for grid-connected circuit breakers, enabling it to adapt to various grid-connected circuit breaker position access methods.

10. The method for identifying the start-up process of a synchronous condenser during power system grid connection as described in claim 9, characterized in that, In the single-position access mode, the power system monitors the position status of the grid-connected circuit breaker in real time and determines whether the start-up process has started based on the changes in the position status. In the dual-position access mode, in addition to monitoring the position status, the power system also determines whether the start-up process has started based on the relative positional relationship of the two position contacts.

Images