A method for site selection and capacity optimization configuration of a distributed phase modifier

Abstract

The present application relates to the technical field of reactive power compensation device configuration of new energy station, and discloses a site selection and capacity optimization configuration method of distributed phase modifier, which obtains a joint data set based on operation data and simulation data of AHP hierarchical analysis method, constructs a wind power abnormal data elimination method combining GMM clustering algorithm and quartile method to analyze and screen wind power data, then constructs a comprehensive weight model based on improved sensitivity index and short-circuit ratio of new energy station to generate a distributed phase modifier site selection hierarchical configuration scheme, simultaneously constructs an adaptive simulated annealing genetic algorithm optimization correction model to dynamically correct the capacity of distributed phase modifier, and finally dynamically adjusts the number of distributed phase modifier integration nodes to solve the problem of frequent integration and removal of distributed phase modifier from the system, which has great significance for realizing stable operation of high-proportion new energy sending-end power grid system.

Description

Technical Field

[0001] This invention relates to the field of reactive power compensation device configuration technology for new energy power plants, and in particular to a method for site selection and capacity optimization configuration of distributed synchronous condensers. Background Technology

[0002] With the growth of wind power installed capacity, the impact of wind power transmission on the power system is becoming increasingly prominent. On the one hand, the large-scale replacement of thermal power units by new energy sources with centralized grid connection leads to a decrease in the short-circuit ratio of new energy power plants and a weakening of the voltage support capacity of sending-end nodes. On the other hand, regions with rapid new energy development are often located at the end of the interconnected grid, with weak grid structures, small local loads, and insufficient reactive power support, making them unable to withstand grid disturbances and resulting in severe transient voltage fluctuations at the sending end after a fault. Therefore, how to improve the short-circuit ratio of high-proportion wind power systems, reduce transient voltage fluctuations, increase installed capacity, and enhance the transmission capacity of sending-end systems has become one of the urgent problems to be solved. Summary of the Invention

[0003] The purpose of this invention is to solve the above problems by designing a method for the location selection and capacity optimization configuration of distributed synchronous condensers.

[0004] The technical solution of the present invention to achieve the above objectives is a method for site selection and capacity optimization configuration of distributed synchronous condensers, which includes the following steps:

[0005] Step 1: Construct a simulation model of a high-voltage direct current (HVDC) transmission network with multiple new energy wind farms at the sending end, and obtain the operation data and simulation data of the HVDC transmission network simulation model;

[0006] Step 2: Process the operational and simulation data in the high-voltage direct current transmission network simulation model based on the AHP (Analytic Hierarchy Process) method, and obtain the joint wind power dataset;

[0007] Step 3: Based on the wind power anomaly data removal model combining GMM clustering algorithm and quartile method, the abnormal wind power data is removed and a complete comprehensive wind power dataset is obtained.

[0008] The abnormal wind power data processing procedure includes: removing outliers from the dataset using the GMM clustering algorithm; and performing secondary processing on the clustered sample points using the quartile method.

[0009] Step 4: Embed the processed data into the PSCAD simulation model to generate target data, and at the same time obtain the simulation results to determine the type of wind turbine disconnection fault.

[0010] It should be noted that the processing in step four is based on a specific fault scenario of wind turbine trawling, to resolve existing electrical faults and ensure the normal and stable operation of the high-voltage direct current transmission system model.

[0011] Step 5: Calculate the short-circuit ratio and improved sensitivity index of each node in the new energy power station, and determine the key quantity proportions of the short-circuit ratio and improved sensitivity index in each node, thereby obtaining the comprehensive score of each node, and determining the distributed synchronous condenser location nodes based on the comprehensive score of the nodes.

[0012] It should be noted that the distributed synchronous condenser location determination method includes: for nodes with a short-circuit ratio greater than 2.5, only the improved sensitivity index is considered; for nodes with an improved sensitivity index less than 0.005, only the short-circuit ratio of the station is considered; and for the remaining nodes, both weights are considered. By calculating the comprehensive score of each node, a distributed synchronous condenser location hierarchical configuration scheme is generated based on the node score.

[0013] The method for determining the weight of the short-circuit ratio and the improved sensitivity index includes: based on the improved sensitivity index of each node and the short-circuit ratio of the new energy power station, using the entropy weight method, proposing a weight design method for the improved sensitivity index and the short-circuit ratio of the new energy power station, constructing a weight model for the improved sensitivity index and the short-circuit ratio of the new energy power station, and calculating the weights of the two typical parameters.

[0014] Step 6: Establish a distributed synchronous condenser capacity optimization and correction model, use dynamic adaptive simulated annealing genetic algorithm to optimize and correct the model, take the transient voltage fluctuation of the rectifier bus as the voltage safety constraint, generate a distributed synchronous condenser capacity optimization configuration scheme and optimize the distributed synchronous condenser capacity.

[0015] It should be noted that the capacity optimization method for distributed synchronous condensers includes: based on the actual operation and maintenance topology of high-proportion wind power transmission system, constructing simulation models for three operating conditions: three-phase short-circuit fault, two-phase ground fault, and single-phase ground fault; establishing a distributed synchronous condenser capacity optimization correction model; using dynamic adaptive simulated annealing genetic algorithm to optimize the correction model; using transient voltage fluctuations of the rectifier bus as voltage safety constraints; generating a distributed synchronous condenser capacity optimization configuration scheme; and further optimizing the distributed synchronous condenser capacity.

[0016] Step 7: Based on the above distributed synchronous condenser capacity optimization configuration scheme, set the switch matrix according to different operating conditions and dynamically adjust the number of distributed synchronous condensers in the connected nodes.

[0017] It should be noted that, based on the different number of distributed synchronous condensers connected under different operating conditions, this application sets up five sets of switch matrices to dynamically adjust the number of distributed synchronous condensers connected. The process includes: combining different fault combinations, controlling the switch matrices, and simulating the actual number of distributed synchronous condensers connected during operation and maintenance.

[0018] The operational and simulation data in step one include, but are not limited to: wind turbine access location and multiplication model parameter data, conventional thermal power unit installed capacity data, grid load parameters, grid-connected line parameter data, distributed synchronous condensers, SVG and other reactive power compensation device configuration data, 35KV, 500KV, 220KV bus voltage level data, DC transmission system network architecture data, wind farm side lines, and transmission side rectifier and inverter device operation data.

[0019] The process of processing operational and simulation data based on the AHP (Analytic Hierarchy Process) in step two is as follows:

[0020] The analytic hierarchy process (AHP) is used to determine the composite weights of simulated wind power data and actual wind power data, and the formula for these composite weights is as follows:

[0021]

[0022] In the formula: P ij D represents the weight of the i-th factor in layer P to the overall objective. i Represents a hierarchical matrix.

[0023] The process of removing abnormal wind power data using the wind power anomaly data removal model in step three is as follows:

[0024] Let Φ(X|φ) m The power parameter probability density function selected in the GMM algorithm is shown in the following formula:

[0025]

[0026] Where: σ m Let m be the covariance matrix of the m-th Gaussian distribution;

[0027] The abnormal wind power data is initially processed using the power parameter probability density function, and then a secondary processing step is performed using the quartile method. The secondary processing procedure is as follows:

[0028] First, the mixed wind power dataset is horizontally processed based on the quartile method, dividing the wind power in the dataset into four equal parts to obtain the upper fraction Q1, lower fraction Q3, and median fraction Q2, and then calculating the interquartile range I. QR With the data acceptance interval [F d ,F u The calculation formula is as follows:

[0029] I QR =Q3-Q1 (3)

[0030] [F d ,F u ]=[Q1-1.5I QR Q3+1.5IQR (4)

[0031] Then, the mixed data is processed longitudinally based on the quartile method, outliers are removed, and average power data is used to replace outliers to obtain a complete integrated wind power dataset.

[0032] The types of wind turbine grid disconnection faults in step four include:

[0033] It lacks low-voltage ride-through capability, lacks high-voltage ride-through capability, and lacks sufficient reactive power regulation capability.

[0034] The calculation process for the short-circuit ratio and improved sensitivity index of each node in step five is as follows:

[0035] The formula for calculating the node short-circuit ratio is:

[0036]

[0037] In the formula: U Ni For the voltage at the grid connection point of the new energy power station, Let P be the impedance in the i-th row and j-th column of the equivalent impedance matrix at the bus connection point. REj The active power injected into the prime mover;

[0038] A calculation model for the improved sensitivity index is established, and the formula for calculating the value of the improved sensitivity index at each node is as follows:

[0039]

[0040] In the formula: I TSj(SC) To configure distributed synchronous condenser node B j Improved trajectory sensitivity index; V i (t k Q j0 +ΔQ jSC ) is B under a typical fault j The node is configured with a capacity of S. SC Distributed camera condenser after t k Time node voltage V i V i (t k Q j0 ) is B j The node is configured with a capacity of S. SC Distributed condenser front t k Time node voltage V i ;ΔQ jSC For B j Node-distributed camera t k Reactive power at any given moment.

[0041] The calculation process for the comprehensive score of each node in step five is as follows:

[0042] A weighted model for improving sensitivity index and short-circuit ratio of new energy power plants was constructed, and the weights of the two typical parameters were calculated.

[0043] Let the short-circuit ratio model of the new energy power station after the distributed synchronous condenser and its system power flow constraints be:

[0044] MRSCR i >MRSCR min (7)

[0045] 0≤k i ≤k max (8)

[0046]

[0047] i = 1, 2, 3, ..., n (11)

[0048] Where: MRSCR i For the short-circuit ratio of new energy wind farm i, MRSCR min k is the lower limit threshold for the short-circuit ratio of the depot. max Configure an upper limit on the number of distributed synchronous condensers for the site nodes, P i Q i These are the active power and reactive power of node i, respectively, V i V j G represents the voltage at the i-th node and the j-th node, respectively. ij With B ij Let θ represent the real and imaginary parts of the admittance matrix at node (i, j), respectively. ij This represents the phase difference between nodes i and j;

[0049] The two parameters are standardized. The larger the value of the improved sensitivity index, the better. Equation (12) is used for standardization. The nodes with low short-circuit ratios in the station are nodes with weak voltage support capabilities. Equation (13) is used for standardization.

[0050]

[0051] Where x is the value of the j-th indicator, x max x is the maximum value of the j-th indicator. min r is the minimum value of the j-th indicator. ij Standardized values;

[0052] Calculate the weight p of the index value of the i-th scheme under the j-th index. ij :

[0053]

[0054] Calculate the entropy weight e of the j-th index j :

[0055]

[0056] in,

[0057]

[0058] Information entropy redundancy is:

[0059] g j =1-e j (17)

[0060] Calculate the weights of the two influencing factors separately:

[0061]

[0062] The comprehensive score of each station node is calculated based on the improved sensitivity index and the station sensitivity weight:

[0063]

[0064] The distributed synchronous condenser capacity optimization and correction model established in step six is ​​as follows:

[0065]

[0066] In the formula: n1 is the number of distributed synchronous condenser nodes configured to meet the short-circuit ratio requirements of new energy power stations in the first stage, and K i k represents the number of distributed synchronous condensers configured after adjusting for wind farm i capacity. i To configure the number of distributed synchronous condensers S for wind farm i that meets the short-circuit ratio requirements of new energy power stations in the first phase, i γ represents the capacity of the distributed synchronous condenser, and γ is the penalty coefficient.

[0067] In step six, the process of using a dynamic adaptive simulated annealing genetic algorithm to optimize and correct the model, and taking the transient voltage fluctuation of the rectifier bus as a voltage safety constraint, to generate a distributed synchronous condenser capacity optimization configuration scheme is as follows:

[0068] 1) Chromosome encoding is performed using a fixed-length symbol encoding method. Each gene in the chromosome corresponds to the number of distributed synchronous condensers in the wind farm. The number of distributed synchronous condensers initially configured for each wind farm is represented by two states, "1" and "0" and their superposition.

[0069] 2): Define the initial population as:

[0070] W = [p1, p2, ... p n ] (twenty one)

[0071] In the formula: p1, p2, ... p n Chromosome encoding of the number of distributed synchronous condensers initially configured for n wind farm nodes in the first phase;

[0072] Initialize and generate a population W of size N. i The number of distributed synchronous condensers represented by the chromosome genes of the population cannot be less than the number of synchronous condensers configured in the first stage.

[0073] 3): Initial population W i Substitute into equation (21) to calculate, based on the fitness function and constraints, detect the performance of individuals in the population, reproduce offspring for individuals with good performance, and eliminate individuals with poor performance;

[0074] 4): The initial population is subjected to genetic algorithm selection, crossover, and mutation processes to obtain the subpopulation Q. i The parent population is merged with the offspring population, and the objective function value of the new population is calculated.

[0075] 5): Introduce the simulated annealing algorithm and use the Metropolis criterion to compare the fitness function values ​​J2(x2) and J2(x1) of individuals in two generations;

[0076] If J2(x2) < J2(x1), then accept the new generation of offspring; otherwise, according to the... Probability of accepting a new generation of offspring;

[0077] 6): Iterate step by step until the number of iterations is met, and output the optimal configuration capacity of the fixed node camera.

[0078] The voltage safety constraint in step six is ​​as follows:

[0079]

[0080] i = 1, 2, 3, ..., n (24)

[0081] 0≤k i ≤k max (25)

[0082] V min (t)≤V(t)≤V max (t) (26)

[0083] In the formula: V min (t) represents the lower limit of transient voltage fluctuation, V max (t) represents the upper limit of transient voltage fluctuation.

[0084] Compared with the prior art, this application has the following advantages:

[0085] 1. This application ensures the validity and representativeness of wind power data imported into the model by processing mixed wind power data datasets, which helps to enhance the accuracy of distributed synchronous condenser site selection and capacity optimization methods;

[0086] 2. This application considers the short-circuit ratio and improved sensitivity index of new energy power stations, and adopts a hierarchical location strategy for distributed synchronous condensers based on the entropy weight method to improve the efficiency of reactive power resource allocation and extend the service life of distributed synchronous condensers.

[0087] 3. This application adopts a dynamically adaptive distributed synchronous condenser capacity optimization method to improve the stability of high-voltage power transmission, and at the same time applies a switch matrix to reduce maintenance costs during long-term operation. Attached Figure Description

[0088] Figure 1 This is a flowchart of a method for site selection and capacity optimization configuration of a distributed synchronous condenser according to the present invention;

[0089] Figure 2 This is the high-proportion wind power transmission-end grid topology diagram with multiple voltage levels described in this invention;

[0090] Figure 3 This is a topology diagram of the external transmission structure of the ultra-high voltage direct current transmission system described in this invention;

[0091] Figure 4 This is a scatter plot of the hybrid wind power data processed according to the present invention;

[0092] Figure 5 This is a comparison diagram of the voltage support effect of the configured nodes before and after the location selection of the distributed synchronous condenser based on multiple fault conditions, as described in this invention.

[0093] Figure 6 This is a comparison chart of the improvement effect of the rectifier bus voltage before and after the capacity optimization of the distributed synchronous condenser based on multiple fault conditions as described in this invention;

[0094] Figure 7 This is the topology diagram of the distributed synchronous condenser control using a switch matrix as described in this invention, which involves the insertion and removal of synchronous condensers. Detailed Implementation

[0095] The present invention will now be described in detail with reference to the accompanying drawings, such as... Figure 1-7 As shown;

[0096] First, such as Figure 2 and Figure 3As shown: Based on the PSCAD simulation model, a high-voltage direct current (HVDC) transmission network architecture with a voltage level of ±800kV is constructed. The receiving-end converter station adopts constant DC voltage and constant trigger angle control, while the sending end adopts constant power control. Three voltage levels of the sending-end grid nodes are set: 35kV, 220kV, and 500kV. Based on the integration of traditional thermal power units into 6 new energy wind turbine units, the wind turbine units all adopt the x50 aggregation multiplication model, and both the turbine side and the grid side adopt dual-loop control. The converter valve internally adopts a single-pole 24-pulse rectifier.

[0097] Secondly, a hierarchical model based on a combination of actual wind power data and simulated wind power data was constructed. Economic factors, risk factors, and data reference value factors were decomposed into wind turbine operation and maintenance costs, site temporary engineering costs, and simulation platform costs; natural risk factors and platform operation risk factors; actual wind farm operation data; and simulated operation data under various operating scenarios. By comparing the relative importance of each pair of factors, a comprehensive decision-making result was obtained. The proportions of economic factors, risk factors, and data reference value factors were 0.17%, 0.13%, and 0.7%, respectively.

[0098] The analytic hierarchy process (AHP) determines the weighting formula for combining simulated wind power data and actual wind power data as follows:

[0099]

[0100] In the formula: P ij D represents the weight of the i-th factor in layer P to the overall objective. i Represents a hierarchical matrix.

[0101] Substituting the numerical values ​​into the calculation, the ratio of actual wind power data to simulated wind power data is 0.6512 and 0.348, respectively.

[0102] Then, the obtained dataset is processed using the GMM clustering algorithm. Power is selected as the variable, as it is influenced by multiple factors such as wind speed, turbine rotor current, and doubly-fed induction generator (DFIG) parameters. Clustering is performed based on the power probability density function and its characteristics, eliminating discrete points in the wind power dataset. Φ(X∣φ m The power parameter probability density function selected in the GMM algorithm is shown in the following formula:

[0103]

[0104] Where: σ m Let represent the covariance matrix of the m-th Gaussian distribution.

[0105] For the data after initial processing, the quartile method is used to handle outlier data;

[0106] The mixed wind power dataset is processed horizontally using the quartile method. Power parameters are divided into units of 60MW, and the dataset is sorted by power value. The wind power is then divided into four equal parts to obtain the upper fraction Q1, lower fraction Q3, and median fraction Q2. The interquartile range I is then calculated. QR With the data acceptance interval [F d ,F u Based on this standard, wind power data is retained, and the calculation formula is as follows:

[0107] I QR =Q3-Q1 (3)

[0108] [F d ,F u ]=[Q1-1.5I QR Q3+1.5I QR (4)

[0109] The mixed data is processed longitudinally using the quartile method. Wind speed data is divided into units of 1 m / s, and the corresponding wind power data is retained within the wind speed data receiving interval. Anomalous data is further identified.

[0110] To maintain data integrity, average power data was used to replace outlier power data, thus improving the comprehensive wind power dataset. After removing 19.08% of outliers, the power scatter plot is as follows: Figure 4 As shown.

[0111] The processed data is embedded into the PSCAD simulation model to generate target data. Simultaneously, stability characteristic simulation is performed to obtain simulation results and determine the wind turbine grid disconnection fault modes under various operating scenarios.

[0112] Subsequently, the improved sensitivity value and the short-circuit ratio value of each new energy power station are calculated. For nodes with a short-circuit ratio greater than 2.5, only the improved sensitivity index is considered, i.e., the improved sensitivity index is weighted at 1 and the short-circuit ratio index is weighted at 0. For nodes with an improved sensitivity index less than 0.005, only the short-circuit ratio index is considered, i.e., the short-circuit ratio index is weighted at 1 and the improved sensitivity index is weighted at 0. Combining the two weights, the comprehensive score of each node is calculated. Based on the node scores, a distributed synchronous condenser location hierarchical configuration scheme is generated.

[0113] The formula for calculating the short-circuit ratio of the renewable energy power plant at the wind turbine access node in the power transmission system simulation model is as follows:

[0114]

[0115] In the formula: U Ni For the voltage at the grid connection point of the new energy power station, Let P be the impedance in the i-th row and j-th column of the equivalent impedance matrix at the bus connection point.REj The active power injected into the prime mover.

[0116] Obtain the voltage and reactive power changes of the target node before and after configuring the distributed synchronous condenser. Establish a model for calculating the improved sensitivity index, and calculate the improved sensitivity index value of the wind turbine access node within the selected study range of the transmission system simulation model. The formula is as follows:

[0117]

[0118] In the formula: I TSj(SC) To configure distributed synchronous condenser node B j Improved trajectory sensitivity index; V i (t k Q j0 +ΔQ jSC ) is B under a typical fault j The node is configured with a capacity of S. SC Distributed camera condenser after t k Time node voltage V i V i (t k Q j0 ) is B j The node is configured with a capacity of S. SC Distributed condenser front t k Time node voltage V i ;ΔQ jSC For B j The reactive power of the distributed synchronous condenser at time tk.

[0119] For the remaining nodes, the proportions of the key quantities of the short-circuit ratio of the new energy power station and the improved sensitivity index are determined. Based on the improved sensitivity index and the short-circuit ratio of the new energy power station at each node, and based on the entropy weight method model, a weight design method for the improved sensitivity index and the short-circuit ratio of the new energy power station is proposed. A weight model for the improved sensitivity index and the short-circuit ratio of the new energy power station is constructed, and the weights of the two typical parameters are calculated.

[0120] The short-circuit ratio model of new energy power plants after distributed synchronous condenser and its system power flow constraints are as follows:

[0121] MRSCR i >MRSCR min (7)

[0122] 0≤k i ≤k max (8)

[0123]

[0124] i = 1, 2, 3, ..., n (11)

[0125] Where: MRSCR i For the short-circuit ratio of new energy wind farm i, MRSCR min This is the lower limit threshold for the short-circuit ratio of the power station (the threshold for weak systems is 2.5). max Configure an upper limit on the number of distributed synchronous condensers for the site nodes. i Q i These are the active power and reactive power of node i, respectively, V i V j G represents the voltage at the i-th node and the j-th node, respectively. ij With B ij Let θ represent the real and imaginary parts of the admittance matrix at node (i, j), respectively. ij This represents the phase difference between nodes i and j.

[0126] The two parameters are standardized. The larger the value of the improved sensitivity index, the better. Equation (12) is used for standardization. The nodes with low short-circuit ratios in the station are nodes with weak voltage support capabilities. Equation (13) is used for standardization.

[0127]

[0128] Where x is the value of the j-th indicator, x max x is the maximum value of the j-th indicator. min r is the minimum value of the j-th indicator. ij These are standardized values.

[0129] Calculate the weight p of the index value of the i-th scheme under the j-th index. ij :

[0130]

[0131] Calculate the entropy weight e of the j-th index j :

[0132]

[0133] Information entropy redundancy is:

[0134] g j =1-e j (17)

[0135] Calculate the weights of the two influencing factors separately:

[0136]

[0137] The comprehensive score of each station node is calculated based on the improved sensitivity index and the station sensitivity weight:

[0138]

[0139] Arrange the node comprehensive scores from largest to smallest, while taking into account the node's weak voltage support capability and the compensation effect after configuration, to determine the access location of distributed synchronous condensers, avoid unnecessary waste of reactive power resources, and significantly improve the efficiency of reactive power resource allocation. Figure 4 A comparison chart of the voltage support capabilities of the configured nodes after the initial site selection and configuration of the distributed synchronous condensers.

[0140] Based on the above analysis, this application can determine the capacity optimization method for distributed synchronous condensers, including constructing simulation models for three operating conditions—three-phase short-circuit fault, two-phase ground fault, and single-phase ground fault—based on the actual operation and maintenance topology of a high-proportion wind power transmission system, establishing a distributed synchronous condenser capacity optimization correction model, using a dynamic adaptive simulated annealing genetic algorithm to optimize the correction model, and using transient voltage fluctuations on the rectifier bus as constraints to generate a distributed synchronous condenser capacity optimization configuration scheme, thereby further optimizing the distributed synchronous condenser capacity.

[0141] Based on the transient voltage fluctuation characteristics, the optimization objective is to minimize the cost of transient voltage instability control, and a distributed synchronous condenser capacity correction objective function is constructed as follows:

[0142]

[0143] In the formula: n1 is the number of distributed synchronous condenser nodes configured to meet the short-circuit ratio requirements of new energy power stations in the first stage, and K i k represents the number of distributed synchronous condensers configured after adjusting for wind farm i capacity. i To configure the number of distributed synchronous condensers S for wind farm i that meets the short-circuit ratio requirements of new energy power stations in the first phase, i γ represents the capacity of the distributed synchronous condenser, and γ is the penalty coefficient.

[0144] Considering the high and low voltage ride-through capabilities of the wind turbines at the power station, to further limit transient voltage fluctuations, the voltage safety constraint during the transient response process is increased as follows:

[0145]

[0146] i = 1, 2, 3, ..., n (23)

[0147] 0≤k i ≤k max (twenty four)

[0148] V min (t)≤V(t)≤V max (t) (25)

[0149] In the formula: V min (t) represents the lower limit of transient voltage fluctuation, V max(t) represents the upper limit of transient voltage fluctuation.

[0150] Short-circuit fault scanning was performed on the new energy power plants that had completed the first phase of distributed synchronous condenser configuration to obtain the transient time-domain simulation characteristics of the wind farm in the sending-end power grid. Based on the bus voltage fluctuation trajectories of different wind farms under various short-circuit faults, a distributed synchronous condenser capacity optimization model using an improved simulated annealing genetic algorithm was designed to realize the capacity correction of the distributed synchronous condenser and improve the voltage fluctuation of the rectifier side bus under typical faults. The algorithm steps are as follows:

[0151] Algorithm Step 1: Chromosome encoding is performed using a fixed-length symbol encoding method. Each gene in the chromosome corresponds to the number of distributed synchronous condensers in the wind farm. The number of distributed synchronous condensers initially configured for each wind farm is represented by two states, "1" and "0" and their superposition.

[0152] Algorithm step 2: Define the initial population as:

[0153] W = [p1, p2, ... p n ] (twenty one)

[0154] In the formula: p1, p2, ... p n Chromosome encoding of the number of distributed synchronous condensers initially configured for n wind farm nodes in phase one.

[0155] Initialize and generate a population W of size N. i The number of distributed synchronous condensers represented by the chromosome genes of the population cannot be less than the number of synchronous condensers configured in the first stage.

[0156] Algorithm Step 3: Initial Population W i Substituting into equation (21), the performance of individuals in the population is assessed based on the fitness function and constraints, thus evaluating the wind farm's ability to resist transient overvoltages after the initial configuration of distributed synchronous condensers. Individuals with good performance reproduce to produce offspring, while those with poor performance are eliminated.

[0157] Algorithm step 4: Perform genetic selection, crossover, and mutation processes on the initial population to obtain the subpopulation Q. i The parent population is merged with the offspring population, and the objective function value of each individual in the new population is calculated.

[0158] Algorithm Step 5: To improve global search capability, simulated annealing is introduced, using the Metropolis criterion to compare the fitness function values ​​J2(x2) and J2(x1) of two generations of individuals. If J2(x2) < J2(x1), the next generation of offspring is accepted; otherwise, the next generation is selected based on probability. Accept the new generation of offspring.

[0159] Algorithm step 6: Iterate step by step until the number of iterations is met, and output the optimal configuration capacity of the fixed node camera.

[0160] in, Figure 6 The diagram shows the effect of suppressing transient voltage fluctuations on the rectifier bus under different fault conditions after optimizing the capacity configuration of the distributed synchronous condenser.

[0161] Step Seven: As Figure 7 As shown, based on the above distributed synchronous condenser configuration and capacity optimization scheme, in order to meet the needs of different operating conditions, a switch matrix is ​​set up to dynamically adjust the number of distributed synchronous condensers in the connected nodes.

[0162] The above technical solutions only embody the preferred technical solutions of the present invention. Any modifications that may be made by those skilled in the art to certain parts thereof embody the principles of the present invention and fall within the protection scope of the present invention.

Claims

1. A method for site selection and capacity optimization configuration of distributed synchronous condensers, characterized in that, The method includes the following steps: Step 1: Construct a simulation model of a high-voltage direct current (HVDC) transmission network with multiple new energy wind farms at the sending end, and obtain the operation data and simulation data of the HVDC transmission network simulation model; Step 2: Process the operational and simulation data in the high-voltage direct current transmission network simulation model based on the AHP (Analytic Hierarchy Process) method, and obtain the joint wind power dataset; Step 3: Based on the wind power anomaly data removal model combining GMM clustering algorithm and quartile method, the abnormal wind power data is removed and a complete comprehensive wind power dataset is obtained. Step 4: Embed the processed data into the PSCAD simulation model to generate target data, and at the same time obtain the simulation results to determine the type of wind turbine disconnection fault. Step 5: Calculate the short-circuit ratio and improved sensitivity index of each node in the new energy power station, and determine the key quantity proportions of the short-circuit ratio and improved sensitivity index in each node, thereby obtaining the comprehensive score of each node, and determining the distributed synchronous condenser location nodes based on the comprehensive score of the nodes. The calculation process for the short-circuit ratio and improved sensitivity index of each node in step five is as follows: The formula for calculating the node short-circuit ratio is: In the formula: U Ni For the voltage at the grid connection point of the new energy power station, Let P be the impedance in the i-th row and j-th column of the equivalent impedance matrix at the bus connection point. REj The active power injected into the prime mover; A calculation model for the improved sensitivity index is established, and the formula for calculating the value of the improved sensitivity index at each node is as follows: In the formula: I TSj(SC) To configure distributed synchronous condenser node B j Improved trajectory sensitivity index; V i (t k Q j0 +ΔQ jSC ) is B under a typical fault j The node is configured with a capacity of S. SC Distributed camera condenser after t k Time node voltage V i V i (t k Q j0 ) is B j The node is configured with a capacity of S. SC Distributed condenser front t k Time node voltage V i ;ΔQ jSC For B j Node-distributed camera t k Reactive power at any given moment; The calculation process for the comprehensive score of each node in step five is as follows: A weighted model for improving sensitivity index and short-circuit ratio of new energy power plants was constructed, and the weights of the two typical parameters were calculated. Let the short-circuit ratio model of the new energy power station after the distributed synchronous condenser and its system power flow constraints be: MRSCR i >MRSCR min (7) i = 1, 2, 3, ..., n (11) Where: MRSCR i For the short-circuit ratio of new energy wind farm i, MRSCR min k is the lower limit threshold for the short-circuit ratio of the depot. max Configure an upper limit on the number of distributed synchronous condensers for the site nodes, P i Q i These are the active power and reactive power of node i, respectively, V i V j G represents the voltage at the i-th node and the j-th node, respectively. ij With B ij Let θ represent the real and imaginary parts of the admittance matrix at node (i, j), respectively. ij This represents the phase difference between nodes i and j; The two parameters are standardized. The larger the value of the improved sensitivity index, the better. Equation (12) is used for standardization. The nodes with low short-circuit ratios in the station are nodes with weak voltage support capabilities. Equation (13) is used for standardization. Where x is the value of the j-th indicator, x max x is the maximum value of the j-th indicator. min r is the minimum value of the j-th indicator. ij Standardized values; Calculate the weight p of the index value of the i-th scheme under the j-th index. ij : Calculate the entropy weight e of the j-th index j : in, Information entropy redundancy is: gj=1-e j (17) Calculate the weights of the two influencing factors separately: The comprehensive score of each station node is calculated based on the improved sensitivity index and the station sensitivity weight: Step 6: Establish a distributed synchronous condenser capacity optimization and correction model, use dynamic adaptive simulated annealing genetic algorithm to optimize and correct the model, take the transient voltage fluctuation of the rectifier bus as the voltage safety constraint, generate a distributed synchronous condenser capacity optimization configuration scheme and optimize the distributed synchronous condenser capacity. The distributed synchronous condenser capacity optimization and correction model established in step six is ​​as follows: In the formula: n1 is the number of distributed synchronous condenser nodes configured to meet the short-circuit ratio requirements of new energy power stations in the first stage, and K i k represents the number of distributed synchronous condensers configured after adjusting for wind farm i capacity. i To configure the number of distributed synchronous condensers S for wind farm i that meets the short-circuit ratio requirements of new energy power stations in the first phase, i Where γ is the capacity of the distributed synchronous condenser, and γ is the penalty coefficient. Step 7: Based on the above distributed synchronous condenser capacity optimization configuration scheme, set up the switch matrix according to different operating conditions and dynamically adjust the number of distributed synchronous condensers in the connected nodes.

2. The method for site selection and capacity optimization configuration of distributed synchronous condensers according to claim 1, characterized in that, The operational and simulation data in step one include, but are not limited to: wind turbine access location and multiplication model parameter data, conventional thermal power unit installed capacity data, grid load parameters, grid-connected line parameter data, distributed synchronous condensers, SVG and other reactive power compensation device configuration data, 35KV, 500KV, 220KV bus voltage level data, DC transmission system network architecture data, wind farm side lines, and transmission side rectifier and inverter device operation data.

3. The method for site selection and capacity optimization configuration of distributed synchronous condensers according to claim 1, characterized in that, The process of processing operational and simulation data based on the AHP (Analytic Hierarchy Process) in step two is as follows: The analytic hierarchy process (AHP) is used to determine the composite weights of simulated wind power data and actual wind power data, and the formula for these composite weights is as follows: In the formula: P ij D represents the weight of the i-th factor in layer P to the overall objective. i Represents a hierarchical matrix.

4. The method for site selection and capacity optimization configuration of distributed synchronous condensers according to claim 1, characterized in that, The process of removing abnormal wind power data using the wind power anomaly data removal model in step three is as follows: Let Φ(X|φ) m The power parameter probability density function selected in the GMM algorithm is shown in the following formula: Where: σ m Let m be the covariance matrix of the m-th Gaussian distribution; The abnormal wind power data is initially processed using the power parameter probability density function, and then a secondary processing step is performed using the quartile method. The secondary processing procedure is as follows: First, the mixed wind power dataset is horizontally processed based on the quartile method, dividing the wind power in the dataset into four equal parts to obtain the upper fraction Q1, lower fraction Q3, and median fraction Q2, and then calculating the interquartile range I. QR With the data acceptance interval [F d ,F u The calculation formula is as follows: I QR =Q3-Q1(3)[F d ,F u ]=[Q1-1.5I QR ,Q3+1.5I QR ] (4) Then, the mixed data is processed longitudinally based on the quartile method, outliers are removed, and average power data is used to replace outliers to obtain a complete integrated wind power dataset.

5. The method for site selection and capacity optimization configuration of distributed synchronous condensers according to claim 1, characterized in that, The wind turbine grid disconnection fault types in step four include: lack of low voltage ride-through capability, lack of high voltage ride-through capability, and insufficient reactive power regulation capability.

6. The method for site selection and capacity optimization configuration of distributed synchronous condensers according to claim 1, characterized in that, In step six, the process of using a dynamic adaptive simulated annealing genetic algorithm to optimize and correct the model, and taking the transient voltage fluctuation of the rectifier bus as a voltage safety constraint, to generate a distributed synchronous condenser capacity optimization configuration scheme is as follows: 1) Chromosome encoding is performed using a fixed-length symbol encoding method. Each gene in the chromosome corresponds to the number of distributed synchronous condensers in the wind farm. The number of distributed synchronous condensers initially configured for each wind farm is represented by two states, "1" and "0" and their superposition. 2): Define the initial population as: W=[p1,p2,...p n ] (21) In the formula: p1, p2, ... p n Chromosome encoding of the number of distributed synchronous condensers initially configured for n wind farm nodes in the first phase; Initialize and generate a population W of size N. i The number of distributed synchronous condensers represented by the chromosome genes of the population cannot be less than the number of synchronous condensers configured in the first stage. 3): Initial population W i Substitute into equation (21) to calculate, based on the fitness function and constraints, detect the performance of individuals in the population, reproduce offspring for individuals with good performance, and eliminate individuals with poor performance; 4): The initial population is subjected to genetic algorithm selection, crossover, and mutation processes to obtain the subpopulation Q. i The parent population is merged with the offspring population, and the objective function value of the new population is calculated. 5): Introduce the simulated annealing algorithm and use the Metropolis criterion to compare the fitness function values ​​J2(x2) and J2(x1) of individuals in two generations; If J2(x2) < J2(x1), then accept the next generation of offspring; otherwise, proceed according to probability. Accepting the next generation of offspring; 6): Iterate step by step until the number of iterations is met, and output the optimal configuration capacity of the fixed node camera.

7. The method for site selection and capacity optimization configuration of distributed synchronous condensers according to claim 6, characterized in that, The voltage safety constraint in step six is ​​as follows: i=1,2,3....n (24)0≤k i ≤k max (25) V min (t)≤V(t)≤V max (t) (26) In the formula: V min (t) represents the lower limit of transient voltage fluctuation, V max (t) represents the upper limit of transient voltage fluctuation.

Images