A flexible ring closing phase-shifting voltage regulator selection configuration optimization method and system

Abstract

This invention provides a method and system for optimizing the selection and configuration of flexible closed-loop phase-shifting voltage regulators. The optimization method includes the following steps: collecting distribution network information and historical load data; establishing a load forecasting model based on the CNN-ARIMA-LSTM algorithm; establishing multiple calculation models for voltage regulators and transmission lines; given the capacity and model range of the voltage regulators, obtaining basic configuration schemes; performing calculations on each configuration scheme based on the above calculation models, and performing load forecasting, and then performing safety judgment, outputting configuration schemes that meet preset values; calculating a weighted score for the output values ​​of each model of the configuration schemes that meet the preset values, and selecting the configuration scheme with the highest weighted score as the optimal configuration scheme output. This scheme comprehensively considers the thermal stability of the phase-shifting voltage regulator itself, the thermal stability of the line, and the safety of load growth, obtaining the optimal configuration scheme, achieving a balance between the long-term economic efficiency and safety of distribution network planning, and has high interpretability and feasibility.

Description

Technical Field

[0001] This invention relates to the field of power system optimization technology, specifically to a method and system for optimizing the selection and configuration of a flexible closed-loop phase-shifting voltage regulator. Background Technology

[0002] With the acceleration of urbanization and rapid economic growth, my country's requirements for power system management and intelligence are increasing, and users are demanding more stable and higher-quality power supply. Currently, the problems of unreasonable distribution network structure and lagging development are becoming increasingly prominent. During peak load periods, the receiving-end grid load remains high for extended periods, and uneven load distribution leads to frequent overloads of distribution network equipment and transmission congestion.

[0003] Currently, the planning of flexible loop distribution networks generally adopts rough planning methods, lacking effective planning methods and basic planning basis. At the same time, traditional planning schemes lack consideration for the power flow distribution of the distribution network, lack consideration for the impact of future load growth on the safety of the distribution network, and lack effective means to coordinate and balance the economy and safety of the grid side in the distribution network planning process. Summary of the Invention

[0004] Purpose of the invention: The purpose of this invention is to provide a method and system for optimizing the selection and configuration of a flexible closed-loop phase-shifting voltage regulator that comprehensively considers both the stability of the phase-shifting voltage regulator and the safety of load growth.

[0005] Technical solution: A method for optimizing the selection and configuration of a flexible closed-loop phase-shifting voltage regulator, comprising the following steps:

[0006] (1) Collect information on distribution network equipment, distribution network lines, distribution network electrical quantities, and historical load data;

[0007] (2) Establish a load prediction model based on the CNN-ARIMA-LSTM algorithm;

[0008] (3) Establish calculation models for the closing-loop impact current of distribution network lines, thermal stability of voltage regulators, thermal stability of transmission lines, loss of voltage regulators, and operating efficiency of voltage regulators;

[0009] (4) Based on the information of distribution network equipment and distribution network lines, the capacity and model range of the flexible closed-loop phase-shifting voltage regulator are given to obtain the corresponding configuration scheme;

[0010] (5) Based on the information of distribution network equipment, distribution network line, distribution network electrical quantity and historical load data, calculate each configuration scheme based on the above models, make a safety judgment, and output the configuration scheme that meets the preset value.

[0011] (6) Calculate the weighted score of the output values ​​of each model for the configuration scheme that meets the preset value, and select the configuration scheme with the highest weighted score as the optimal configuration scheme output.

[0012] Specifically, the distribution network equipment information includes transformer type, transformer capacity, no-load loss, load loss, and short-circuit impedance; the distribution network line information includes line type, line unit impedance, and line length; the distribution network electrical quantity information includes three-phase voltage, three-phase current, and power factor; and the historical load data includes the load data of each node of the distribution network on the day of the highest historical load.

[0013] Specifically, step (2) includes the following sub-steps:

[0014] (21) Use the CNN algorithm to extract data features from historical load data. The calculation formula is as follows:

[0015] y t =Pool(σ(x*g)*W+b)

[0016] In the formula: x is the input data, g is the convolution kernel, * indicates the convolution operation, σ is the activation function, Pool is the pooling operation, W is the weight matrix, b is the bias vector, and y is the weight matrix. t It is the output result of data features;

[0017] (22) Using the LSTM algorithm to output the data features extracted by the CNN algorithm y t The formula for load forecasting is as follows:

[0018] h t =LSTM(y t ,h t-1 )

[0019] ε t =hh t

[0020] In the formula: LSTM is the computational unit, y t h represents the data feature output of the CNN algorithm. t-1 h represents the hidden state from the previous moment. t The hidden state at the current moment, h is the actual load value, and ε t The residual sequence between the LSTM prediction results and the actual values;

[0021] (23) Using the ARIMA algorithm to process the residual sequence ε obtained by the LSTM algorithm. t The modeling process yields the load forecast results, and the calculation formula is as follows:

[0022] α t =ARIMA(ε t)

[0023] In the formula: α t For load forecasting results, ARIMA is the modeling and calculation unit, ε t It is a residual sequence.

[0024] Preferably, step (2) further includes:

[0025] (24) Given weight values, perform a weighted average calculation on the residual sequence and load forecast results to obtain the weighted load forecast result. The calculation formula is as follows:

[0026] δ t =w1ε t +w2α t

[0027] Where: δ t For the weighted load forecast results, ε t For the residual sequence, α t The load forecast results are given, where w1 is the weight of the residual sequence and w2 is the weight of the load forecast results.

[0028] Specifically, the calculation model for the closing-loop impact current of the distribution network line includes a steady-state closing-loop current analysis model at the closing point and a maximum closing-loop impact current analysis model. The calculation formula is as follows:

[0029] Steady-state closed-loop current analysis model at the closed-loop point:

[0030]

[0031] In the formula: R1 represents the voltage across the loop closing point, L2 represents the resistance of the sending line at the loop closing point, L1 represents the inductance of the sending line at the loop closing point, R2 represents the resistance of the receiving line at the loop closing point, and L3 represents the inductance of the receiving line at the loop closing point. This represents the steady-state amplitude of the closed-loop current.

[0032] Analysis model of maximum inrush current in closed loop:

[0033]

[0034] In the formula: T is the steady-state amplitude of the closed-loop current. a ii is the decay time constant. M This is the maximum inrush current when the loop closes.

[0035] Specifically, the calculation formula for the thermal stability of the voltage regulator is as follows:

[0036]

[0037] In the formula: J is the current density in the short-circuited winding; t1 is the duration of the short circuit; θ0 is the initial temperature of the winding; θ1 is the average temperature of the winding after the short circuit.

[0038] Specifically, the calculation formula for the thermal stability calculation model of the transmission line is as follows:

[0039]

[0040] In the formula: T L The temperature of the transmission line, T ta The temperature of the surrounding environment; T L0 The initial temperature of the transmission line; T a R is the decay time constant, t is the transmission time, I is the transmission line current; a R is the AC resistance value. d The value is the DC resistance; m is the mass; C p For the specific heat capacity of the conductor; β and K j All are constants.

[0041] Specifically, the voltage regulator loss model is calculated using the following formula:

[0042] P 损 =P0+β 2 P K

[0043] In the formula: P 损 P0 is the total loss of the voltage regulator, P0 is the no-load loss of the voltage regulator, β is the load rate of the voltage regulator, and P0 is the total loss of the voltage regulator. K This refers to the short-circuit loss of the voltage regulator.

[0044] Specifically, the calculation formula for the operating efficiency model of the voltage regulator is as follows:

[0045]

[0046] In the formula: η is the operating efficiency of the voltage regulator, S T This refers to the rated capacity of the voltage regulator. The power factor of the load.

[0047] This invention also provides a flexible closed-loop phase-shifting voltage regulator selection and configuration optimization system, comprising:

[0048] Data acquisition module: used to collect information on distribution network equipment, distribution network lines, distribution network electrical quantities, and historical load data;

[0049] Load forecasting module: used to build a load forecasting model based on the CNN-ARIMA-LSTM algorithm;

[0050] Model building module: used to establish calculation models for the closing-loop impact current of distribution network lines, thermal stability of voltage regulators, thermal stability of transmission lines, loss of voltage regulators, and operating efficiency of voltage regulators;

[0051] The scheme configuration module is used to obtain the corresponding configuration scheme based on the information of distribution network equipment and distribution network lines, given the capacity and model range of the flexible closed-loop phase-shifting voltage regulator;

[0052] The initial screening module is used to calculate each configuration scheme based on the above models according to the information of distribution network equipment, distribution network lines, distribution network electrical quantities and historical load data, and to make a safety judgment and output a configuration scheme that meets the preset value.

[0053] Solution Output Module: This module calculates a weighted score for each model output value of the configuration solutions that meet preset values, and selects the configuration solution with the highest weighted score as the optimal configuration solution output.

[0054] Beneficial effects: Compared with the prior art, the significant effect of this invention is that after collecting distribution network information and historical load data, given the capacity and model range of the flexible loop phase-shifting voltage regulator, a basic configuration scheme is obtained. Load prediction is performed by integrating the CNN-ARIMA-LSTM algorithm. The output value is calculated through multiple calculation models, compared with the set value, and a safety judgment is made to give a configuration scheme that meets the safety requirements. Furthermore, the thermal stability of the phase-shifting voltage regulator body, the thermal stability of the line, and the safety of load growth are considered to comprehensively output the optimal configuration scheme. This scheme takes into account the long-term economic and safety aspects of distribution network planning and has high interpretability and feasibility. Attached Figure Description

[0055] Figure 1 This is a flowchart of the selection and configuration optimization method for the flexible closed-loop phase-shifting voltage regulator in Embodiment 1 of the present invention.

[0056] Figure 2 This is a flowchart of the load prediction model in Embodiment 1 of the present invention.

[0057] Figure 3 This is a graph showing the load growth prediction results of Embodiment 1 of the present invention.

[0058] Figure 4 A flowchart illustrating the selection of a configuration scheme for Embodiment 1 of the present invention.

[0059] Figure 5 This is a topology diagram of the power distribution network structure in Embodiment 2 of the present invention. Detailed Implementation

[0060] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments.

[0061] Example 1

[0062] Please see Figure 1 As shown, this embodiment provides a method for optimizing the selection and configuration of a flexible closed-loop phase-shifting voltage regulator, including the following steps:

[0063] (1) Collect information on distribution network equipment, distribution network lines, distribution network electrical quantities and historical load data.

[0064] The aforementioned distribution network equipment information includes transformer type, transformer capacity, no-load loss, load loss, and short-circuit impedance; the distribution network line information includes line type, line unit impedance, and line length; the distribution network electrical quantity information includes three-phase voltage, three-phase current, and power factor; and the historical load data includes the load data of each node of the distribution network on the day of the highest historical load.

[0065] (2) Establish a load prediction model based on the CNN-ARIMA-LSTM algorithm.

[0066] Please see Figure 2 As shown, firstly, the CNN algorithm is used to extract data features from the historical workload data. The calculation formula is as follows:

[0067] y t =Pool(σ(x*g)*W+b)

[0068] In the formula: x is the input data, g is the convolution kernel, * indicates the convolution operation, σ is the activation function, Pool is the pooling operation, W is the weight matrix, b is the bias vector, and y is the weight matrix. t It is the output result of data features.

[0069] Next, the LSTM algorithm is used to process the data features extracted by the CNN algorithm, and the output result y is obtained. t The formula for load forecasting is as follows:

[0070] h t =LSTM(y t ,h t-1 )

[0071] ε t =hh t

[0072] In the formula: LSTM is the computational unit, y t h represents the data feature output of the CNN algorithm. t-1 h represents the hidden state from the previous moment. t The hidden state at the current moment, h is the actual load value, and ε t This is the residual sequence between the LSTM prediction results and the actual values.

[0073] Specifically, the LSTM computation unit receives the computation result y of the CNN algorithm at the current time step. t and the hidden state h from the previous moment t-1 Output the hidden state h at the current moment. t .

[0074] Then, the ARIMA algorithm is used to process the residual sequence ε obtained by the LSTM algorithm. t Modeling is performed to obtain load forecasting results, and auxiliary forecasting is then conducted. The calculation formula is as follows:

[0075] α t =ARIMA(ε t )

[0076] In the formula: α t For load forecasting results, ARIMA is the modeling and calculation unit, ε t It is a residual sequence.

[0077] Finally, given the weight values, a weighted average is calculated on the residual sequence and the load forecast results to obtain the weighted load forecast result. The calculation formula is as follows:

[0078] δ t =w1ε t +w2α t

[0079] Where: δ t For the weighted load forecast results, ε t For the residual sequence, α t The load forecast results are given, where w1 is the weight of the residual sequence and w2 is the weight of the load forecast results.

[0080] Please see Figure 3 The diagram shows the results of load growth prediction calculation in a specific implementation scenario of this embodiment. The red curve represents the predicted value, and the blue curve represents the actual value. It can be seen that the load prediction is close to the actual value and has good prediction accuracy.

[0081] (3) Establish calculation models for the closing-loop impact current of distribution network lines, thermal stability of voltage regulators, thermal stability of transmission lines, loss of voltage regulators, and operating efficiency of voltage regulators;

[0082] The calculation models described above are explained below:

[0083] The above-mentioned calculation model for the closing-loop inrush current of the distribution network line specifically includes the steady-state closing-loop current analysis model at the closing point and the maximum closing-loop inrush current analysis model. The corresponding calculation formulas are as follows:

[0084] Steady-state closed-loop current analysis model at the closed-loop point:

[0085]

[0086] In the formula: R1 represents the voltage across the loop closing point, L2 represents the resistance of the sending line at the loop closing point, L1 represents the inductance of the sending line at the loop closing point, R2 represents the resistance of the receiving line at the loop closing point, and L3 represents the inductance of the receiving line at the loop closing point. This represents the steady-state amplitude of the closed-loop current.

[0087] Analysis model of maximum inrush current in closed loop:

[0088]

[0089] In the formula: T is the steady-state amplitude of the closed-loop current. a ii is the decay time constant. M This represents the maximum inrush current when the loop closes.

[0090] The calculation formula for the thermal stability of the voltage regulator described above is as follows:

[0091]

[0092] In the formula: J is the current density in the short-circuit winding, calculated according to the root mean square of the symmetrical short-circuit current; t1 is the duration of the short circuit; θ0 is the initial temperature of the winding, which is generally selected as the sum of the highest ambient temperature and the allowable temperature rise of the winding insulation; θ1 is the average temperature of the winding after the short circuit t1.

[0093] The calculation formula for the thermal stability of the above transmission line is as follows:

[0094]

[0095] In the formula: T L The temperature of the transmission line, T ta The temperature of the surrounding environment; T L0 The initial temperature of the transmission line; T a R is the decay time constant, t is the transmission time, I is the transmission line current; a R is the AC resistance value. d The value is the DC resistance; m is the mass; C p For the specific heat capacity of the conductor; β and K j All are constants, β is 0.5, K j Take 1.0123.

[0096] The calculation formula for the above voltage regulator loss model is as follows:

[0097] P 损 =P0+β 2P K

[0098] In the formula: P 损 P0 is the total loss of the voltage regulator, P0 is the no-load loss of the voltage regulator, β is the load rate of the voltage regulator, and P0 is the total loss of the voltage regulator. K This refers to the short-circuit loss of the voltage regulator.

[0099] The calculation formula for the operating efficiency model of the above voltage regulator is as follows:

[0100]

[0101] In the formula: η is the operating efficiency of the voltage regulator, S T This refers to the rated capacity of the voltage regulator. The power factor of the load.

[0102] (4) Based on the information of distribution network equipment and distribution network lines, the capacity and model range of the flexible closed-loop phase-shifting voltage regulator are given to obtain the corresponding configuration scheme.

[0103] Specifically, during implementation, based on the actual situation of the distribution network, the range of selectable phase-shifting voltage regulator capacities is obtained, which in turn leads to the actual range of phase-shifting voltage regulator models to be selected, thus resulting in multiple configuration schemes for the distribution network and the corresponding phase-shifting voltage regulators.

[0104] (5) Based on the information of distribution network equipment, distribution network line, distribution network electrical quantity and historical load data, calculate each configuration scheme based on the above models, make a safety judgment, and output a configuration scheme that meets the preset value.

[0105] Please see Figure 4 As shown, firstly, for each configuration scheme obtained in step (4), the corresponding output value is obtained by using the models in step (3). Then, the load prediction model in step (2) is used to predict the growth of the load value based on historical load data. The load of each node at each time is predicted and the power flow is calculated. The safety judgment is made based on the power flow calculation results at each time. By setting limits, the voltage and line current of each node are ensured not to exceed the limits, and a scheme that meets the preset value is obtained. This process is repeated to obtain all the optional schemes.

[0106] (6) Calculate the weighted score of each model output value of the configuration scheme that meets the preset value, and select the configuration scheme with the highest weighted score as the optimal configuration scheme output.

[0107] Please see Figure 4 As shown, a weighted score is calculated for the output values ​​of each model in the optional schemes in step (5). In this embodiment, the weighted calculation formula is as follows:

[0108] Y = α1T L +α2P损 +α3η

[0109] In the formula: Y is the weighted score of the configuration scheme, T L P represents the temperature of the transmission line. 损 Let η be the total loss of the voltage regulator, η be the operating efficiency of the voltage regulator, and α1 be the voltage regulator operating efficiency. L The weighting coefficient, α2 is P 损 The weighting coefficients are α3 and α3 are the weighting coefficients of η.

[0110] The weighting coefficients take into account both economic factors and the impact of future load growth on the safe operation of the distribution network, and are determined by expert opinions.

[0111] Example 2

[0112] This embodiment applies the flexible closed-loop phase-shifting voltage regulator selection and configuration optimization method proposed in Embodiment 1, using a generalized model of a 10kV radial distribution network as a simulation example. Please refer to [link / reference]. Figure 4 The diagram shown is a structural topology diagram of a generalized 10kV radial distribution network model in this embodiment.

[0113] The data information of the main transformer equipment and lines in the distribution area are shown in Tables 1 and 2 below.

[0114] Table 1. Relevant data of the main transformers in both distribution transformer areas.

[0115]

[0116]

[0117] Table 2. Relevant data for Line A and Line B

[0118] Route Information Line model Line current Active power reactive power Line A YJV22-8.7 / 15-3*400 65A-262A 3.141MVA 1.501MVA Line B YJV22-8.7 / 15-3*400 66A-280A 4.142MVA 1.979MVA

[0119] In this embodiment, the loop-closing objects are two 10kV distribution lines. The specifications of the main transformers corresponding to the two lines are: rated capacity 50MVA, rated voltage 110 / 10kV, connection type YNd11; the two line models are: YJV22-8.7 / 15-3*400 (copper core cross-linked polyethylene insulated double-layer steel tape armored PVC sheathed power cable, rated working voltage 8.7 / 15kV, cable is three-core, single conductor cross-sectional area 400mm²). 2 The lengths are 27.5516km and 18.1207km respectively.

[0120] Based on the optimization method for selecting and configuring flexible closed-loop phase-shifting voltage regulators proposed in Example 1, the optimal planning schemes for operating losses and operating efficiency of phase-shifting voltage regulators with different capacities are calculated and shown in the table below.

[0121]

[0122] When selecting a scheme, the construction and maintenance costs of distribution transformers should be fully considered, and the planning timeframe should be as long as possible, taking into account the economic efficiency of the plan within 10 years. In the current scenario, the S13-M-6300 / 10-6.3 phase-shifting voltage regulator with a capacity of 6300kVA is selected. Although its initial construction cost is higher than that of the S13-M-5000 / 10-6.3, from the perspective of long-term operation and load growth forecast results, its economic efficiency is better than that of the S13-M-5000 / 10-6.3 phase-shifting voltage regulator with a capacity of 5000kVA. The overall planning scheme takes into account the long-term economic efficiency and safety of the power grid. This embodiment fully verifies the feasibility of the flexible closed-loop phase-shifting voltage regulator selection and configuration optimization method of the present invention.

[0123] Example 3

[0124] This embodiment provides a flexible loop phase-shifting voltage regulator selection and configuration optimization system that is compatible with the flexible loop phase-shifting voltage regulator selection and configuration optimization method described in Embodiment 1, including:

[0125] Data acquisition module: used to collect information on power distribution network equipment, power distribution network lines, power distribution network electrical quantities, and historical load data.

[0126] Load forecasting module: Used to build a load forecasting model based on the CNN-ARIMA-LSTM algorithm and perform load forecasting based on historical load data output by the data acquisition module.

[0127] Model building module: used to establish calculation models for the closing-loop impact current of distribution network lines, thermal stability of voltage regulators, thermal stability of transmission lines, loss of voltage regulators, and operating efficiency of voltage regulators.

[0128] Scheme configuration module: Based on the information of distribution network equipment and distribution network lines, it is used to obtain the corresponding configuration scheme by specifying the capacity and model range of the flexible closed-loop phase-shifting voltage regulator.

[0129] The initial screening module is used to calculate each configuration scheme based on the load prediction module and the model building module, according to the distribution network equipment information, distribution network line information, distribution network electrical quantity information and historical load data. It also performs safety judgment and outputs the configuration scheme that meets the preset value.

[0130] Solution Output Module: This module calculates a weighted score for each model output value of the configuration solutions that meet the preset values ​​output by the solution screening module, and selects the configuration solution with the highest weighted score as the optimal configuration solution.

Claims

1. A method for optimizing the selection and configuration of a flexible closed-loop phase-shifting voltage regulator, characterized in that, Includes the following steps: (1) Collect information on distribution network equipment, distribution network lines, distribution network electrical quantities, and historical load data; (2) Establish a load prediction model based on the CNN-ARIMA-LSTM algorithm; Includes the following sub-steps: (21) Use the CNN algorithm to extract data features from historical load data. The calculation formula is as follows: y t =Pool(σ(x*g)*W+b) In the formula: x is the input data, g is the convolution kernel, * indicates the convolution operation, σ is the activation function, Pool is the pooling operation, W is the weight matrix, b is the bias vector, and y is the weight matrix. t It is the output result of data features; (22) Using the LSTM algorithm to output the data features extracted by the CNN algorithm y t The formula for load forecasting is as follows: h t =LSTM(y t ,h t-1 ) ε t =h-h t In the formula: LSTM is the computational unit, y t h represents the data feature output of the CNN algorithm. t-1 h represents the hidden state from the previous moment. t The hidden state at the current moment, h is the actual load value, and ε t The residual sequence between the LSTM prediction results and the actual values; (23) Using the ARIMA algorithm to process the residual sequence ε obtained by the LSTM algorithm. t The modeling process yields the load forecast results, and the calculation formula is as follows: a t =ARIMA(e t ) In the formula: α t For load forecasting results, ARIMA is the modeling and calculation unit, ε t It is a residual sequence; (24) Given weight values, perform a weighted average calculation on the residual sequence and load forecast results to obtain the weighted load forecast result. The calculation formula is as follows: d t =w1e t +w2a t Where: δ t For the weighted load forecast results, ε t For the residual sequence, α t For the load forecast results, w1 is the weight of the residual sequence, and w2 is the weight of the load forecast results; (3) Establish calculation models for the closing-loop impact current of distribution network lines, thermal stability of voltage regulators, thermal stability of transmission lines, loss of voltage regulators, and operating efficiency of voltage regulators; (4) Based on the information of distribution network equipment and distribution network lines, the capacity and model range of the flexible closed-loop phase-shifting voltage regulator are given to obtain the corresponding configuration scheme; (5) Based on the information of distribution network equipment, distribution network line, distribution network electrical quantity and historical load data, calculate each configuration scheme based on the above models, make a safety judgment, and output the configuration scheme that meets the preset value. (6) Calculate the weighted score of the output values ​​of each model for the configuration scheme that meets the preset value, and select the configuration scheme with the highest weighted score as the optimal configuration scheme output.

2. The method for selecting and configuring a flexible closed-loop phase-shifting voltage regulator according to claim 1, characterized in that: The distribution network equipment information includes transformer type, transformer capacity, no-load loss, load loss, and short-circuit impedance; the distribution network line information includes line type, line unit impedance, and line length; the distribution network electrical quantity information includes three-phase voltage, three-phase current, and power factor; the historical load data includes the load data of each node of the distribution network on the day of the highest historical load.

3. The method for selecting and optimizing the flexible closed-loop phase-shifting voltage regulator according to claim 1, characterized in that: The calculation model for the closing-loop impact current of the distribution network line includes a steady-state closing-loop current analysis model at the closing point and a maximum closing-loop impact current analysis model. The calculation formula is as follows: Steady-state closed-loop current analysis model at the closed-loop point: In the formula: R1 represents the voltage across the loop closing point, L2 represents the resistance of the sending line at the loop closing point, L1 represents the inductance of the sending line at the loop closing point, R2 represents the resistance of the receiving line at the loop closing point, and L3 represents the inductance of the receiving line at the loop closing point. This represents the steady-state amplitude of the closed-loop current. Analysis model of maximum inrush current in closed loop: In the formula: T is the steady-state amplitude of the closed-loop current. a The decay time constant, This is the maximum inrush current when the loop closes.

4. The method for selecting and configuring a flexible closed-loop phase-shifting voltage regulator according to claim 1, characterized in that: The calculation model for the thermal stability of the voltage regulator is as follows: In the formula: J is the current density in the short-circuited winding; t1 is the duration of the short circuit; θ0 is the initial temperature of the winding; θ1 is the average temperature of the winding after the short circuit.

5. The method for selecting and optimizing the flexible closed-loop phase-shifting voltage regulator according to claim 1, characterized in that: The calculation model for the thermal stability of the transmission line is as follows: In the formula: T L The temperature of the transmission line, T ta The temperature of the surrounding environment; T L0 The initial temperature of the transmission line; T a R is the decay time constant, t is the transmission time, I is the transmission line current; a R is the AC resistance value. d The value is the DC resistance; m is the mass; C p For the specific heat capacity of the conductor; β and K j All are constants.

6. The method for selecting and configuring a flexible closed-loop phase-shifting voltage regulator according to claim 1, characterized in that: The voltage regulator loss model is calculated using the following formula: P 损 =P0+β 2 P K In the formula: P 损 P0 is the total loss of the voltage regulator, P0 is the no-load loss of the voltage regulator, β is the load rate of the voltage regulator, and P0 is the total loss of the voltage regulator. K This refers to the short-circuit loss of the voltage regulator.

7. The method for selecting and configuring a flexible closed-loop phase-shifting voltage regulator according to claim 6, characterized in that: The operating efficiency model of the voltage regulator is calculated using the following formula: In the formula: η is the operating efficiency of the voltage regulator, S T This refers to the rated capacity of the voltage regulator. The power factor of the load.

8. A flexible closed-loop phase-shifting voltage regulator selection and configuration optimization system, characterized in that, include: Data acquisition module: used to collect information on distribution network equipment, distribution network lines, distribution network electrical quantities, and historical load data; Load forecasting module: used to build a load forecasting model based on the CNN-ARIMA-LSTM algorithm; The specific steps involved in establishing a load prediction model based on the CNN-ARIMA-LSTM algorithm are as follows: The CNN algorithm is used to extract features from historical workload data. The calculation formula is as follows: y t =Pool(σ(x*g)*W+b) In the formula: x is the input data, g is the convolution kernel, * indicates the convolution operation, σ is the activation function, Pool is the pooling operation, W is the weight matrix, b is the bias vector, and y is the weight matrix. t It is the output result of data features; The output result y of the data features extracted by the CNN algorithm using the LSTM algorithm. t The formula for load forecasting is as follows: h t =LSTM(y t ,h t-1 ) ε t =h-h t In the formula: LSTM is the computational unit, y t h represents the data feature output of the CNN algorithm. t-1 h represents the hidden state from the previous moment. t The hidden state at the current moment, h is the actual load value, and ε t The residual sequence between the LSTM prediction results and the actual values; Using the ARIMA algorithm to process the residual sequence ε obtained by the LSTM algorithm t The modeling process yields the load forecast results, and the calculation formula is as follows: a t =ARIMA(e t ) In the formula: α t For load forecasting results, ARIMA is the modeling and calculation unit, ε t It is a residual sequence; Given weight values, a weighted average is calculated on the residual sequence and the load forecast results to obtain the weighted load forecast result. The calculation formula is as follows: d t =w1e t +w2a t Where: δ t For the weighted load forecast results, ε t For the residual sequence, α t For the load forecast results, w1 is the weight of the residual sequence, and w2 is the weight of the load forecast results; Model building module: used to establish calculation models for the closing-loop impact current of distribution network lines, thermal stability of voltage regulators, thermal stability of transmission lines, loss of voltage regulators, and operating efficiency of voltage regulators; The scheme configuration module is used to obtain the corresponding configuration scheme based on the information of distribution network equipment and distribution network lines, given the capacity and model range of the flexible closed-loop phase-shifting voltage regulator; The initial screening module is used to calculate each configuration scheme based on the above models according to the information of distribution network equipment, distribution network lines, distribution network electrical quantities and historical load data, and to make a safety judgment and output a configuration scheme that meets the preset value. Solution Output Module: This module calculates a weighted score for each model output value of the configuration solutions that meet preset values, and selects the configuration solution with the highest weighted score as the optimal configuration solution output.

Images