Configuration optimization method for dynamic reactive power compensation device of new energy sending end power grid

Abstract

The invention relates to the technical field of power grid equipment, in particular to a configuration optimization method for a dynamic reactive power compensation device of a new energy sending-end power grid, and the method comprises the steps: collecting power grid data and historical operation data; extracting relevance and laws among the power grid data, establishing a reactive power demand prediction model to predict reactive power compensation demands in future preset time, and determining key areas and key nodes of reactive power compensation; the optimal installation position and capacity are determined, and the type of the dynamic reactive power compensation device is selected; and generating an optimal control strategy according to the optimal installation position, the capacity and the type of the dynamic reactive power compensation device, automatically switching and adjusting the reactive power compensation device, and completing a dynamic complete machine debugging task and a test task of the reactive power compensation device. Therefore, the problems that the control strategy cannot be adjusted and optimized in real time and accurate control of the sending-end power grid voltage is difficult to realize due to the fact that the configuration mode of the related technology depends on initially collected data to formulate a compensation scheme and lacks a verification continuous learning mechanism are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power grid equipment, and in particular to a configuration optimization method of a dynamic reactive power compensation device of a new energy sending-end power grid. BACKGROUND

[0002] The sending-end power grid of a large-scale new energy sending base usually has no conventional power support in the initial operation stage, and the power grid characteristics are basically dominated by power electronic equipment. The voltage of the sending-end power grid is completely determined by the flexible direct current station. Compared with the traditional synchronous machine power system, the 100% power electronic equipment dominates the power system. The active and reactive imbalances in the alternating current power grid will be reflected as system voltage changes (voltage collapse, voltage fluctuation exceeding the standard, overvoltage, etc.). Voltage distribution and disturbance response become the core problem of system operation characteristics.

[0003] In related technologies, the configuration of the dynamic reactive power compensation device is the main key technology for improving the voltage support capability of the wide-area new energy sending-end power grid and solving the voltage stability problem caused by the active power fluctuation of the new energy. For example, based on the configuration principles of voltage support demand, dynamic response speed, capacity matching, and harmonic suppression, devices such as static reactive power compensators, static reactive power generators, and static synchronous compensators are selected for the configuration of the dynamic reactive power compensation device.

[0004] However, the configuration method in related technologies mainly relies on the initial collected data to develop a compensation scheme, and the intelligent control strategy is insufficient, which leads to poor voltage stability caused by the active power fluctuation accompanied by the strong randomness and volatility of new energy, and it is difficult to achieve effective and accurate control of the voltage of the sending-end power grid. In addition, there is a lack of verification and continuous learning mechanism, and when the control strategy has a problem, it cannot be adjusted and optimized accordingly. The configuration accuracy is poor, and it needs to be improved urgently. SUMMARY

[0005] The present application provides a configuration optimization method of a dynamic reactive power compensation device of a new energy sending-end power grid to solve the problem that the configuration method in related technologies relies on initial collected data to develop a compensation scheme and lacks a verification and continuous learning mechanism, which further leads to the inability to adjust and optimize in real time when the control strategy has a problem, and the difficulty in achieving effective and accurate control of the voltage of the sending-end power grid.

[0006] The first aspect embodiment of the application provides a configuration optimization method of a dynamic reactive power compensation device of a new energy sending-end power grid, comprising the following steps: determining specific targets of dynamic reactive power compensation device optimization based on the operating power of a plurality of reference operating points of the sending-end power grid, collecting power grid data and historical operating data of the new energy sending-end; extracting the correlation and rules between the power grid data according to the power grid data and the historical operating data to establish a reactive power demand prediction model, and predicting the reactive power compensation demand in a future preset time by using the reactive power demand prediction model, and determining at least one key area and at least one key node of the reactive power compensation by combining the distribution of the reactive power in the power grid; determining the best installation position and capacity of the reactive power compensation device based on the characteristics of the power grid, the at least one key area, the at least one key node and the reactive power compensation demand, and selecting a corresponding dynamic reactive power compensation device type; generating an optimized control strategy according to the best installation position, the capacity and the dynamic reactive power compensation device type, automatically switching and adjusting the reactive power compensation device, and completing the dynamic whole-machine debugging task and the test task of the reactive power compensation device.

[0007] Through the above technical solution, the embodiments of the application can ensure the best reactive power compensation effect, improve the stability and power quality of the power grid, and promote the consumption and utilization of new energy by analyzing and mining the correlation and rules between the power grid data, predicting the reactive power demand in a future period of time, determining the best installation position, capacity and type of the reactive power compensation device; the automatic switching and adjustment of the dynamic reactive power compensation device can be realized through the optimized control strategy to adapt to the rapid change of the active power output of new energy; the whole-machine debugging and testing of the reactive power compensation device can ensure the normal operation of the equipment and the best performance. Compared with related technologies, the continuous optimization learning mechanism is added, and effective and accurate control of the voltage of the sending-end power grid can be realized.

[0008] Optionally, in an embodiment of the application, a real-time feedback mechanism of power grid voltage and reactive power key parameters is established; based on the real-time feedback mechanism, the parameters and threshold values of the optimized control strategy are adjusted according to real-time data.

[0009] Through the above technical solution, the embodiments of the application can dynamically close-loop adjust the parameters and threshold values of the control strategy through the real-time feedback mechanism, so that the control system has the intelligent ability of "perception-decision-adaptation" to improve the control precision.

[0010] Optionally, in an embodiment of the present application, further comprising: simulating the optimization control strategy to generate an optimization control strategy meeting preset effectiveness and stability conditions; and / or, iteratively updating the optimization control strategy according to actual power grid operation data and actual operation state of the reactive power compensation device to obtain an effect of periodically evaluating the control strategy, and updating and upgrading the optimization control strategy according to the effect.

[0011] Through the above technical solutions, the embodiments of the present application can cooperate with simulation verification, continuous learning optimization, and adjustment of parameters and thresholds of the control strategy according to real-time data feedback to timely adjust and optimize the control strategy and existing problems, reduce the need for human verification and adjustment intervention, and improve the intelligence, accuracy, effectiveness, and stability of compensation.

[0012] Optionally, in an embodiment of the present application, the optimization formula of the optimization control strategy is:

[0013]

[0014] wherein s is a current state, a is a current action, r is a reward obtained after the action is performed, s' is a next state, a is a learning rate, g is a discount factor, and a' is a selectable action of the next state.

[0015] Through the above technical solutions, the embodiments of the present application can optimize the control strategy through the Q-learning optimization method to realize timely adjustment of the control strategy and improve the accuracy of configuration optimization of the dynamic reactive power compensation device of the new energy sending-end power grid.

[0016] Optionally, in an embodiment of the present application, the best installation position and capacity of the reactive power compensation device and the corresponding dynamic reactive power compensation device type are determined based on the power grid characteristics, the at least one key area, the at least one key node, and the reactive power compensation demand, and the best installation position and capacity of the reactive power compensation device are determined by using a particle swarm algorithm to simulate movement of particles in a solution space to find an optimal solution, wherein the dynamic reactive power compensation device type includes a static synchronous compensator and a static reactive power compensator, and velocity update and position update expressions of the particle swarm algorithm are:

[0017] V id t+1=ωV id t+c1r1(p id best-x id t)+c2r2(g d best-x id t) (4)

[0018] x id t+1=x id t+Vid t+1 (5)

[0019] wherein, V id t+1 is the velocity of particle i in the d-th dimension at the t+1 generation, V id t is the velocity of particle i in the d-th dimension at the t generation, x id t+1 is the position of particle i in the d-th dimension at the t+1 generation, x id t is the position of particle i in the d-th dimension at the t generation, ω is an inertia weight used to control the influence of the velocity of the particle in the last generation on the current velocity, c1 and c2 are learning factors, respectively representing the acceleration weight of the particle approaching its historical optimal position and the global optimal position, r1 and r2 are two random numbers uniformly distributed in the range of [0, 1], which introduce randomness into the search process, p id best is the individual historical optimal position of particle i in the d-th dimension, g d best is the global historical optimal position of the entire particle swarm in the d-th dimension.

[0020] By the above technical solution, the embodiments of the present application can determine the optimal installation position, capacity and type of the reactive power compensation device by using the particle swarm algorithm. Through accurate "position-capacity-type" trinity optimization, the return on investment of the reactive power compensation device can be improved by more than 40%, effectively ensuring the best reactive power compensation effect.

[0021] The second aspect of the embodiments of the present application provides a configuration optimization system of a dynamic reactive power compensation device of a new energy sending end power grid, comprising: an acquisition module configured to determine the specific target of dynamic reactive power compensation device optimization based on the operating power corresponding to each of a plurality of reference operating points of the sending end power grid, and acquire power grid data and historical operating data of the new energy sending end; a modeling module configured to extract the correlation and rules between the power grid data according to the power grid data and the historical operating data, to establish a reactive power demand prediction model, and to predict the reactive power compensation demand in a future preset time by using the reactive power demand prediction model, and to determine at least one key area and at least one key node of the reactive power compensation by combining with the analysis of the distribution of the reactive power in the power grid; a selection module configured to determine the optimal installation position and capacity of the reactive power compensation device based on the power grid characteristics, the at least one key area, the at least one key node and the reactive power compensation demand, and to select the corresponding dynamic reactive power compensation device type; and an optimization module configured to generate an optimized control strategy according to the optimal installation position, the capacity and the dynamic reactive power compensation device type, to automatically switch and adjust the reactive power compensation device, and to complete the dynamic whole-machine debugging task and the test task of the reactive power compensation device.

[0022] By the technical solution, the embodiment of the application can analyze the correlation and rules among power grid data, predict reactive power demand in a future period of time, determine the optimal installation position, capacity and type of the reactive power compensation device, ensure the best reactive power compensation effect, improve the stability and power quality of the power grid, and promote the consumption and utilization of new energy. The embodiment of the application can also realize the automatic switching and adjustment of the dynamic reactive power compensation device through the optimization control strategy to adapt to the rapid change of active power output of new energy. The embodiment of the application can also ensure the normal operation of the device and the best performance through the overall debugging and testing of the reactive power compensation device. Compared with related technologies, the embodiment of the application adds a continuous optimization learning mechanism, and can realize effective and accurate control of the sending-end power grid voltage.

[0023] Optionally, in an embodiment of the application, the system further comprises a feedback mechanism establishment module configured to establish a real-time feedback mechanism for power grid voltage and reactive power key parameters; and a parameter threshold adjustment module configured to adjust parameters and thresholds of the optimization control strategy according to real-time data based on the real-time feedback mechanism.

[0024] Through the technical solution, the embodiment of the application can dynamically close-loop adjust the parameters and thresholds of the control strategy through the real-time feedback mechanism, so that the control system has intelligent capabilities of "perception-decision-adaptation" to improve control accuracy.

[0025] Optionally, in an embodiment of the application, the system further comprises a simulation module configured to simulate the optimization control strategy to generate an optimization control strategy meeting preset effectiveness and stability conditions; and / or an update module configured to iteratively update the optimization control strategy according to actual power grid operation data and actual operation state of the reactive power compensation device to obtain the effect of periodically evaluating the control strategy, and update and upgrade the optimization control strategy according to the effect.

[0026] Through the technical solution, the embodiment of the application can cooperate with simulation and verification, continuous learning and optimization, and adjustment of parameters and thresholds of the control strategy according to real-time data feedback to timely adjust and optimize the control strategy and existing problems, reduce the need for human verification and adjustment intervention, and improve the intelligence, accuracy, effectiveness and stability of compensation.

[0027] Optionally, in an embodiment of the application, the optimization formula of the optimization control strategy is as follows:

[0028]

[0029] wherein s is a current state, a is a current action, r is a reward obtained after the action is performed, s' is a next state, a is a learning rate, g is a discount factor, and a' is a selectable action of the next state.

[0030] By the technical solution, the embodiment of the application can optimize the control strategy by a Q-learning optimization method, realize timely adjustment of the control strategy, and improve the accuracy of configuration optimization of the dynamic reactive power compensation device of the new energy sending-end power grid.

[0031] Optionally, in an embodiment of the application, the selection module is configured to simulate the movement of particles in the solution space to find the optimal solution by using a particle swarm algorithm to determine the optimal installation position and capacity of the reactive power compensation device, wherein the dynamic reactive power compensation device type includes a static synchronous compensator and a static reactive power compensator, and the velocity update and position update expressions of the particle swarm algorithm are as follows:

[0032] V id t+1 = ωV id t + c1r1(p id best-x id t) + c2r2(g d best-x id t) (4)

[0033] x id t+1 = x id t + V id t+1 (5)

[0034] wherein V id t+1 is the velocity of the particle i in the dth dimension at the t+1 generation, V id t is the velocity of the particle i in the dth dimension at the t generation, x id t+1 is the position of the particle i in the dth dimension at the t+1 generation, x id t is the position of the particle i in the dth dimension at the t generation, ω is an inertia weight for controlling the influence of the velocity of the previous generation of the particle on the current velocity, c1 and c2 are learning factors, which represent the acceleration weights of the particle approaching the individual historical optimal position and the global optimal position, respectively, r1 and r2 are two random numbers uniformly distributed in the range of [0, 1], which introduce randomness into the search process, p id best is the individual historical optimal position of the particle i in the dth dimension, and g d best is the global historical optimal position of the entire particle swarm in the dth dimension.

[0035] By the technical solution, the embodiment of the application can determine the optimal installation position, capacity and type of the reactive power compensation device by using the particle swarm algorithm, and through the accurate "position-capacity-type" trinity optimization, the return on investment of the reactive power compensation device can be improved by more than 40%, and the optimal reactive power compensation effect can be effectively ensured.

[0036] The third aspect of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the configuration optimization method of the dynamic reactive power compensation device of the new energy sending-end power grid as described in the above embodiments.

[0037] The fourth aspect of the present application provides a computer readable storage medium, which stores a computer program, and the program is executed by a processor to implement the configuration optimization method of the dynamic reactive power compensation device of the new energy sending-end power grid as described above.

[0038] The fifth aspect of the present application provides a computer program product, which stores a computer program, and the program is executed by a processor to implement the configuration optimization method of the dynamic reactive power compensation device of the new energy sending-end power grid as described above.

[0039] The embodiments of the present application can analyze the correlation and rules between the power grid data and predict the reactive power demand in the future period of time, and determine the optimal installation position, capacity and type of the reactive power compensation device by using the particle swarm algorithm, to ensure the best reactive power compensation effect, improve the stability and power quality of the power grid, and promote the consumption and utilization of new energy. The deep learning algorithm is used for automatic switching and adjustment of the dynamic reactive power compensation device, and the parameters and thresholds of the control strategy are adjusted according to the real-time data feedback, simulation verification, continuous learning optimization, and the like, so that the control strategy and existing problems can be adjusted and optimized in time, the demand for human verification and adjustment intervention is reduced, and the intelligence, accuracy, effectiveness and stability of the compensation are improved. Thus, the configuration method of the related art relies on the initial collected data to formulate the compensation scheme and lacks a verification and continuous learning mechanism, which further leads to the inability to adjust and optimize in real time when the control strategy has problems, and the effective and accurate control of the sending-end power grid voltage is difficult to achieve.

[0040] Additional aspects and advantages of the present application will be in part apparent and in part pointed out hereinafter. BRIEF DESCRIPTION OF DRAWINGS

[0041] The above and / or additional aspects and advantages of the present application will become apparent and be readily appreciated from the following description, taken in conjunction with the accompanying drawings, in which:

[0042] Figure 1 A flowchart of a configuration optimization method of a dynamic reactive power compensation device of a new energy sending-end power grid according to an embodiment of the present application;

[0043] Figure 2 A flowchart of a configuration optimization method of a dynamic reactive power compensation device of a new energy sending-end power grid according to an embodiment of the present application;

[0044] Figure 3 This is a structural diagram of a system for optimizing the configuration of a dynamic reactive power compensation device for a new energy sending-end power grid according to an embodiment of the present application;

[0045] Figure 4 A schematic diagram of the structure of an electronic device provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0046] The following describes in detail embodiments of the present application, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0047] The following describes a configuration optimization method for a dynamic reactive power compensation device of a new energy sending-end power grid in accordance with an embodiment of the present application with reference to the accompanying drawings. Aiming at the problem that the configuration method of the related technology mentioned in the background technology center relies on the initial collected data to formulate a compensation plan and lacks a verification and continuous learning mechanism, which leads to the inability to adjust and optimize in real time when problems occur in the control strategy, and it is difficult to achieve effective and accurate control of the voltage of the sending-end power grid, the present application provides a configuration optimization method for a dynamic reactive power compensation device of a new energy sending-end power grid. In this method, the correlation and regularity between the power grid data can be analyzed and mined, and the reactive power demand in the future period can be predicted. The particle swarm algorithm is used to determine the optimal installation position, capacity and type of the reactive power compensation device, thereby ensuring the best reactive power compensation effect, improving the stability and power quality of the power grid, and promoting the absorption and utilization of new energy. A deep learning algorithm is used to automatically switch and adjust the dynamic reactive power compensation device, and the parameters and thresholds of the control strategy are adjusted according to simulation verification, continuous learning optimization and real-time data feedback. The control strategy and existing problems can be adjusted and optimized in a timely manner, reducing the need for manual verification and adjustment intervention, and improving the intelligence, accuracy, effectiveness and stability of the compensation. This solves the problem that the configuration method of the relevant technology relies on the initially collected data to formulate the compensation plan and lacks a verification and continuous learning mechanism, which leads to the inability to adjust and optimize in real time when problems arise in the control strategy, making it difficult to achieve effective and accurate control of the sending-end grid voltage.

[0048] Specifically, Figure 1 A flow chart of a method for optimizing the configuration of a dynamic reactive power compensation device of a new energy sending-end power grid provided in an embodiment of the present application.

[0049] like Figure 1 As shown, the configuration optimization method of the dynamic reactive power compensation device of the new energy sending-end power grid includes the following steps:

[0050] In step S101, based on the operation power corresponding to each reference operation point of the sending end power grid, the specific target of the dynamic reactive power compensation device optimization is determined, and the grid data and historical operation data of the new energy sending end are collected.

[0051] The grid data includes, but is not limited to, topological structure, line parameter, load distribution, new energy power generation output characteristics and the like basic data; the historical operation data includes, but is not limited to, voltage data, power data, short circuit capacity, output characteristics, reactive power control mode and the like basic data.

[0052] In actual execution process, the embodiment of the application can collect the grid data at a preset data sampling interval time by deploying remote terminal units at the new energy booster station, the grid connection point and the key bus, and can obtain the historical operation data of the grid through the new energy power station monitoring system.

[0053] The embodiment of the application can collect the grid data and historical operation data of the new energy sending end to provide effective data support for subsequent analysis of the correlation and rules among the grid data.

[0054] In step S102, the correlation and rules among the grid data are extracted according to the grid data and the historical operation data, to establish a reactive power demand prediction model, predict the reactive power compensation demand in a future preset time by using the reactive power demand prediction model, and determine at least one key area and at least one key node of the reactive power compensation by combining the analysis of the distribution of the reactive power in the grid.

[0055] Those skilled in the art should understand that the reactive power compensation demand mainly depends on the grid structure, load characteristics, new energy penetration rate and power quality requirements, and the reactive power compensation demand can include reactive power demand increment and total reactive power demand.

[0056] As an implementable way, the embodiment of the application can use big data analysis technology to mine the correlation and rules among the grid data, establish a reactive power demand prediction model, predict the reactive power demand in a future period of time according to the historical grid data and real-time grid data, deeply analyze the distribution of the reactive power in the grid, determine the key area and key node of the reactive power compensation, and provide a basis for subsequent configuration of the reactive power compensation device.

[0057] The embodiment of the application can provide a basis for accurately calculating the reactive power demand by mining the correlation and rules among the grid data, and provide a basis for subsequent configuration of the reactive power compensation device by determining the key area and key node of the reactive power compensation.

[0058] In step S103, based on the grid characteristics, at least one key area, at least one key node and the reactive power compensation demand, the best installation position and capacity of the reactive power compensation device are determined, and the corresponding dynamic reactive power compensation device type is selected.

[0059] It can be understood that the dynamic reactive power compensation device type includes but is not limited to static reactive power compensator (such as fixed capacitor, thyristor switched capacitor), static synchronous compensator, static reactive power generator, active power filter, etc., which can be used alone or in combination with different types of devices.

[0060] In some embodiments, the optimal installation position and capacity of the reactive power compensation device can be determined by an optimization algorithm according to the grid characteristics and reactive power compensation requirements, and the appropriate dynamic reactive power compensation device type is selected to ensure the best reactive power compensation effect. Among them, the optimization algorithm can select the particle swarm algorithm.

[0061] The embodiments of the present application can analyze the correlation and rules between the grid data and predict the reactive power demand in the future period of time, and determine the optimal installation position, capacity and type of the reactive power compensation device by using the optimization algorithm, which can ensure the best reactive power compensation effect, improve the stability and power quality of the grid, and promote the consumption and utilization of new energy.

[0062] In step S104, an optimized control strategy is generated according to the optimal installation position, capacity and dynamic reactive power compensation device type, and the reactive power compensation device is automatically switched and adjusted to complete the dynamic whole machine debugging task and test task of the reactive power compensation device.

[0063] Among them, the type of optimized control strategy includes but is not limited to control algorithm optimization strategy, neural network control strategy, fuzzy logic control strategy, reinforcement learning control strategy, etc., which can be selected by persons skilled in the art according to actual scenes. The whole machine debugging and testing includes but is not limited to initialization setting, parameter setting, trial operation, load test and performance monitoring, etc.

[0064] As an implementable way, the embodiments of the present application can use deep learning algorithm to formulate intelligent control strategy, realize automatic switching and adjustment of dynamic reactive power compensation device, and then perform whole machine debugging and testing on the installed dynamic reactive power compensation device.

[0065] The specific logical steps of using deep learning algorithm to formulate intelligent control strategy can be:

[0066] S104-1: Data preprocessing: cleaning the historical operation data of the new energy sending end grid, removing outliers and repeated values, and normalizing or standardizing the data to the same order of magnitude for subsequent processing;

[0067] S104-2: Feature extraction processing: extracting useful features for control strategy formulation from original data, creating new features by transforming or combining original features, and reducing feature dimension by PCA (principal component analysis) or LDA (linear discriminant analysis) method to reduce computational complexity;

[0068] S104-3: Model construction: define the model structure using a fully connected neural network deep learning model, and randomly initialize the weights and biases of the model;

[0069] S104-4: Model training: calculate the input data through the model to obtain the predicted value, and use the mean square error loss function to measure the difference between the predicted value and the actual value, use the chain rule to calculate the gradient of the loss function on the model parameters, and use the gradient descent algorithm to update the weights and biases of the model to minimize the loss function;

[0070] S104-5: Model evaluation and optimization: evaluate the prediction accuracy, recall rate and other indicators of the model using the validation set, and adjust the model structure, hyperparameters and other parameters according to the evaluation results to improve the performance of the model.

[0071] The mean square error loss function expression is:

[0072] (MSE) L = (y_pred-y_true)2 / N (1)

[0073] The gradient descent algorithm expression is:

[0074]

[0075] where θ t+1 is the updated parameter, θ t is the current parameter, ∩ is the learning rate, which determines the distance of each step in the gradient direction, is the gradient of the loss function J(θ) with respect to the parameter θ, pointing to the direction of the fastest function growth, and the gradient descent algorithm updates the variable in the opposite direction of the function gradient to gradually approach the minimum point of the function.

[0076] The embodiments of the present application can realize the automatic switching and adjustment of the dynamic reactive power compensation device by optimizing the control strategy to adapt to the rapid change of active power output of new energy; and can ensure normal operation of the equipment and play the best performance by testing and debugging the whole machine of the reactive power compensation device.

[0077] Optionally, in an embodiment of the present application, it further comprises: establishing a real-time feedback mechanism for key parameters of grid voltage and reactive power; based on the real-time feedback mechanism, adjusting the parameters and threshold values of the optimization control strategy according to real-time data.

[0078] It can be understood that the real-time feedback mechanism is a dynamic adjustment process in the control system that continuously monitors the output and adjusts the input in real time, and its core is the rapid closed loop of "perception-computation-execution".

[0079] In actual implementation, the embodiment of the application can establish a real-time feedback mechanism of key parameters of grid voltage and reactive power, take real-time data as input of the control strategy, and adjust parameters and thresholds of the control strategy according to the real-time data, to ensure accuracy and effectiveness of the control strategy.

[0080] Based on the above technical solution, the embodiment of the application can dynamically close-loop adjust parameters and thresholds of the control strategy through the real-time feedback mechanism, so that the control system has intelligent capabilities of "perception-decision-adaptation", to improve control precision.

[0081] Optionally, in an embodiment of the application, the method further comprises: simulating the optimized control strategy to generate the optimized control strategy meeting preset effectiveness and stability conditions; and / or iteratively optimizing the optimized control strategy according to actual grid operation data and actual operation state of the reactive power compensation device, to obtain effects of periodically evaluating the control strategy, and updating and upgrading the optimized control strategy according to the effects.

[0082] In specific implementation, the embodiment of the application can simulate the control strategy by using power system simulation software; and verify effectiveness and stability of the control strategy in a simulation environment, to timely adjust and optimize problems. The power system simulation software includes but is not limited to MATLAB / Simulink (MATrix LABoratory / Simulation Link, graphical simulation environment based on MATLAB), PSCAD (Power System Computer Aided Design), etc.

[0083] The embodiment of the application can also establish a learning mechanism of the control strategy, to constantly optimize the control strategy according to grid operation data and operation state of the reactive power compensation device; and periodically evaluate effects of the control strategy, to update and upgrade the control strategy according to evaluation results. The specific logic steps of establishing the learning mechanism of the control strategy can be:

[0084] Step S1: data collection and processing: collect historical operation data, load data, and new energy power generation output data of a new energy sending-end grid, and perform cleaning, transformation, and normalization processing on the data;

[0085] Step S2: model training: adopt a convolutional neural network learning model, and train the convolutional neural network learning model by using the processed data, to calculate a predicted value through forward propagation, measure a difference between the predicted value and an actual value by using a loss function, and update model parameters through back propagation;

[0086] Step S3: policy generation and optimization: generate an initial control policy according to the trained convolutional neural network learning model, collect feedback data by interacting with the environment, and further optimize the control policy using the data through a Q-learning optimization method.

[0087] The embodiment of the present application can be used for automatic switching and adjustment of dynamic reactive power compensation devices based on deep learning algorithm, facilitate adaptation to rapid changes in active power output of new energy, improve the accuracy and response speed of compensation, and cooperate with simulation verification, continuous learning optimization, and adjustment of parameters and thresholds of control strategies according to real-time data feedback to timely adjust and optimize the control strategies and existing problems, reduce the need for human verification and adjustment intervention, and improve the intelligence, accuracy, effectiveness, and stability of compensation.

[0088] Optionally, in an embodiment of the present application, the optimization formula for optimizing the control strategy is:

[0089]

[0090] where s is the current state, a is the current action, r is the reward obtained after executing the action, s' is the next state, a is the learning rate, g is the discount factor, and a' is the optional action in the next state.

[0091] In actual execution, the embodiment of the present application can update the Q value through the Q-learning optimization method to further optimize the control strategy, and determine the update of the Q value based on the reward obtained after executing the action, the next state, and the optional action in the next state to determine the optimized control strategy.

[0092] The embodiment of the present application can optimize the control strategy through the Q-learning optimization method to realize timely adjustment of the control strategy and improve the accuracy of configuration optimization of dynamic reactive power compensation devices of new energy sending end power grids.

[0093] Optionally, in an embodiment of the present application, based on the characteristics of the power grid, at least one key area, at least one key node, and the demand for reactive power compensation, the best installation location and capacity of the reactive power compensation device are determined, and the corresponding type of dynamic reactive power compensation device is selected, including: using a particle swarm algorithm to simulate the movement of particles in the solution space to find the optimal solution to determine the best installation location and capacity of the reactive power compensation device, wherein the type of dynamic reactive power compensation device includes a static synchronous compensator and a static reactive power compensator, and the velocity update and position update expressions of the particle swarm algorithm are:

[0094] V id t+1=ωV id t+c1r1(p id best-x id t)+c2r2(g d best-xid t) (4)

[0095] x id t+1=x id t+V id t+1 (5)

[0096] where V id t+1is the velocity of particle i in the dth dimension at the t+1th generation, V id tis the velocity of particle i in the dth dimension at the tth generation, x id t+1is the position of particle i in the dth dimension at the t+1th generation, x id tis the position of particle i in the dth dimension at the tth generation, ω is the inertia weight used to control the influence of the velocity of the last generation on the current velocity, c1and c2are learning factors representing the acceleration weight of the particle to the historical optimal position of itself and the global optimal position, respectively, r1and r2are two random numbers uniformly distributed in the range of [0, 1] to introduce randomness into the search process, p id bestis the individual historical optimal position of particle i in the dth dimension, g d bestis the global historical optimal position of the entire particle swarm in the dth dimension.

[0097] Specifically, the embodiments of the present application can model the velocity update and position update expressions of the particle swarm algorithm based on the velocity, position, inertia weight, and historical optimal position of the particle swarm of the last generation in the same dimension.

[0098] Based on the above scheme, the embodiments of the present application can determine the optimal installation position, capacity, and type of the reactive power compensation device by using the particle swarm algorithm. Through accurate "position-capacity-type" trinity optimization, the return on investment of the reactive power compensation device can be improved by more than 40%, and the optimal reactive power compensation effect can be effectively ensured.

[0099] To enable those skilled in the art to more clearly understand the configuration optimization method of the present application, the optimization method is described in detail below with one specific embodiment.

[0100] As Figure 2 shown, the optimization method of the embodiment can include the following steps:

[0101] Step S201: preliminary preparation and data collection: clearly define the specific target of dynamic reactive power compensation optimization, and collect the grid data and historical operation data of the new energy sending end;

[0102] Step S202: deep analysis and mining of reactive power compensation: analyze and mine the correlation and rules between the grid data, establish a reactive power demand prediction model to predict the reactive power demand in the future period of time, and analyze and determine the key areas and key nodes of reactive power compensation;

[0103] Step S203: Dynamic reactive power compensation device configuration scheme making: determine the optimal installation location and capacity of the reactive power compensation device using an optimization algorithm, and select the appropriate dynamic reactive power compensation device type;

[0104] Step S204: Optimization strategy making: use a deep learning algorithm to develop an intelligent control strategy to achieve automatic switching and adjustment of the dynamic reactive power compensation device;

[0105] Step S205: Real-time feedback: establish a real-time feedback mechanism and adjust the parameters and thresholds of the control strategy according to real-time data;

[0106] Step S206: Model simulation and verification: simulate the control strategy using power system simulation software and verify the effectiveness and stability of the control strategy in the simulation environment, and make timely adjustments and optimizations to existing problems;

[0107] Step S207: Continuous learning and optimization: establish a learning mechanism for the control strategy, continuously optimize the control strategy, and periodically evaluate the effectiveness of the control strategy, and update and upgrade the control strategy according to the results.

[0108] The configuration optimization method of the dynamic reactive power compensation device of the new energy sending-end power grid according to the embodiments of the present application can analyze and mine the correlation and rules between grid data and predict the reactive power demand in the future period of time, and determine the optimal installation location, capacity and type of the reactive power compensation device using the particle swarm algorithm, to ensure the best reactive power compensation effect, improve the stability and power quality of the power grid, and promote the consumption and utilization of new energy. The deep learning algorithm is used for automatic switching and adjustment of the dynamic reactive power compensation device, and the parameters and thresholds of the control strategy are adjusted according to the real-time data feedback, simulation verification, continuous learning and optimization, which can timely adjust and optimize the control strategy and existing problems, reduce the need for human verification and adjustment intervention, and improve the intelligence, accuracy, effectiveness and stability of the compensation. Thus, the configuration method of the related art relies on the initial collected data to develop a compensation scheme and lacks a verification and continuous learning mechanism, which further leads to the inability to adjust and optimize in real time when the control strategy has problems, making it difficult to achieve effective and accurate control of the sending-end power grid voltage.

[0109] Secondly, refer to the attached Figure 3 The configuration optimization system of the dynamic reactive power compensation device of the new energy sending-end power grid according to the embodiments of the present application is described.

[0110] Figure 3 is a block diagram of the configuration optimization system of the dynamic reactive power compensation device of the new energy sending-end power grid according to the embodiments of the present application.

[0111] As Figure 3As shown, the configuration optimization system 10 of the dynamic reactive power compensation device of the new energy sending end power grid comprises: an acquisition module 100, a modeling module 200, a selection module 300, and an optimization module 400.

[0112] The acquisition module 100 is configured to determine the specific target of the dynamic reactive power compensation device optimization based on the operating power of the multiple reference operating points of the sending end power grid, and to acquire the grid data and historical operating data of the new energy sending end.

[0113] The modeling module 200 is configured to extract the correlation and rules between the grid data based on the grid data and the historical operating data, to establish a reactive power demand prediction model, to predict the reactive power compensation demand in a future preset time period using the reactive power demand prediction model, and to determine at least one key area and at least one key node of the reactive power compensation by analyzing the distribution of the reactive power in the power grid.

[0114] The selection module 300 is configured to determine the optimal installation location and capacity of the reactive power compensation device based on the grid characteristics, the at least one key area, the at least one key node, and the reactive power compensation demand, and to select the corresponding dynamic reactive power compensation device type.

[0115] The optimization module 400 is configured to generate an optimized control strategy based on the optimal installation location, the capacity, and the dynamic reactive power compensation device type, to automatically switch and adjust the reactive power compensation device, and to complete the dynamic whole-machine debugging task and the testing task of the reactive power compensation device.

[0116] Optionally, in an embodiment of the present application, the optimization system 10 further comprises a feedback mechanism establishment module and a parameter threshold adjustment module. The feedback mechanism establishment module is configured to establish a real-time feedback mechanism for the key parameters of the grid voltage and the reactive power. The parameter threshold adjustment module is configured to adjust the parameters and thresholds of the optimized control strategy based on the real-time data according to the real-time feedback mechanism.

[0117] Optionally, in an embodiment of the present application, the optimization system 10 further comprises a simulation module and an updating module. The simulation module is configured to simulate the optimized control strategy to generate an optimized control strategy that satisfies the preset effectiveness and stability conditions. The updating module is configured to iteratively optimize the optimized control strategy according to the actual grid operating data and the actual operating state of the reactive power compensation device, to evaluate the effect of the optimized control strategy at regular intervals, and to update and upgrade the optimized control strategy according to the effect.

[0118] Optionally, in an embodiment of the present application, the optimization formula of the optimized control strategy is as follows:

[0119]

[0120] Wherein, s is the current state, a is the current action, r is the reward obtained after performing the action, s' is the next state, a is the learning rate, g is the discount factor, and a' is the next state selectable action.

[0121] Optionally, in an embodiment of the present application, the selection module is configured to use a particle swarm algorithm to simulate the movement of particles in the solution space to find the optimal solution to determine the optimal installation location and capacity of the reactive power compensation device, wherein the dynamic reactive power compensation device type includes a static synchronous compensator and a static reactive power compensator, and the velocity update and position update expressions of the particle swarm algorithm are as follows:

[0122] V id t+1=ωV id t+c1r1(p id best-x id t)+c2r2(g d best-x id t) (4)

[0123] x id t+1=x id t+V id t+1 (5)

[0124] Wherein, V id t+1 is the velocity of particle i in the dth dimension at the t+1 generation, V id t is the velocity of particle i in the dth dimension at the t generation, x id t+1 is the position of particle i in the dth dimension at the t+1 generation, x id t is the position of particle i in the dth dimension at the t generation, w is the inertia weight for controlling the influence of the previous generation velocity of the particle on the current velocity, c1 and c2 are learning factors, representing the acceleration weight of the particle approaching its own historical optimal position and the global optimal position, respectively, r1 and r2 are two random numbers uniformly distributed in the range of [0, 1], introducing randomness into the search process, p id best is the individual historical optimal position of particle i in the dth dimension, g d best is the global historical optimal position of the entire particle swarm in the dth dimension.

[0125] It should be noted that the aforementioned explanation of the configuration optimization method embodiment of the dynamic reactive power compensation device of the new energy sending end power grid also applies to the configuration optimization system of the dynamic reactive power compensation device of the new energy sending end power grid of the embodiment, which will not be described here.

[0126] The configuration optimization system of the dynamic reactive power compensation device of the new energy sending-end power grid provided by the embodiment of the application can analyze the correlation and rules among power grid data and predict reactive power demand in a future period of time, and determine the optimal installation position, capacity and type of the reactive power compensation device by using a particle swarm algorithm, so as to ensure that the reactive power compensation effect is optimal, improve the stability and power quality of the power grid, and promote the consumption and utilization of new energy. The automatic switching and adjustment of the dynamic reactive power compensation device are performed by using a deep learning algorithm, and the parameters and thresholds of the control strategy are adjusted according to real-time data feedback, simulation verification, continuous learning optimization and control strategy. The control strategy and existing problems can be adjusted and optimized in a timely manner, the demand for human verification and adjustment intervention is reduced, and the intelligence, accuracy, effectiveness and stability of compensation are improved. Thus, the problems of the related art, such as the configuration mode relying on initial collected data to formulate a compensation scheme and lacking a verification and continuous learning mechanism, and thus being unable to adjust and optimize in a timely manner when the control strategy has a problem and being difficult to realize effective and accurate control of the voltage of the sending-end power grid, are solved.

[0127] Figure 4 A structural schematic diagram of an electronic device is provided for the embodiments of the application. The electronic device can include:

[0128] The memory 401, the processor 402 and the computer program stored in the memory 401 and executable on the processor 402.

[0129] The processor 402 implements the configuration optimization method of the dynamic reactive power compensation device of the new energy sending-end power grid provided in the above embodiments when executing the program.

[0130] Further, the electronic device further includes:

[0131] The communication interface 403 is used for communication between the memory 401 and the processor 402.

[0132] The memory 401 is used to store the computer program executable on the processor 402.

[0133] The memory 401 can include a high-speed RAM memory, and can also include a non-volatile memory, such as at least one disk memory.

[0134] If the memory 401, the processor 402 and the communication interface 403 are implemented independently, the communication interface 403, the memory 401 and the processor 402 can be connected with each other through a bus and complete communication between each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For convenience of representation, Figure 4 Only one thick line is used to represent the bus in the figure, but it does not mean that there is only one bus or only one type of bus.

[0135] Optionally, in a specific implementation, if the memory 401, the processor 402 and the communication interface 403 are integrated on a chip, the memory 401, the processor 402 and the communication interface 403 can complete communication between each other through an internal interface.

[0136] The processor 402 can be a Central Processing Unit (CPU), or an Application Specific Integrated Circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0137] The embodiments of the present application also provide a computer readable storage medium, having stored thereon a computer program, which, when executed by a processor, implements the configuration optimization method of the dynamic reactive power compensation device of the new energy sending-end power grid as above.

[0138] The embodiments of the present application also provide a computer program product, having stored thereon a computer program, which, when executed by a processor, implements the configuration optimization method of the dynamic reactive power compensation device of the new energy sending-end power grid as above.

[0139] In the description of the application, reference to "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" means that a particular feature, structure, material, or characteristic being described is included in at least one embodiment or example of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment or example. Furthermore, the described specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples. In addition, the usage of "N" means at least two, for example, two, three or the like, unless explicitly stated otherwise.

[0140] Furthermore, the terms "first", "second", or the like, are used merely as a designation of certain elements or features, and do not imply or connote relative importance or a specific order of categorization thereof. Accordingly, features described as "first" or "second" can be explicitly or implicitly included in at least one of the features. In the description of the application, the meaning of "N" is at least two, for example, two, three, etc., unless explicitly specified otherwise.

[0141] Any process or method descriptions or blocks in flow charts or otherwise described herein represent embodiments which can be managed as one or more modules, segments, or portions of code which include one or more executable instructions for implementing specific logic functions or steps, and alternate implementations are possible. In some embodiments, the processes and methods described can be executed by one or more apparatuses or devices, either directly or after conversion to another language. Alternate implementations are possible.

[0142] The logic and / or steps represented in the flowcharts and / or described herein, for example, can be considered as a sequence of executable instructions stored in a computer readable medium, which can be executed by an instruction execution system, apparatus or device, such as a computer-based system, a processor-based system, or other system that can fetch the instructions from the instruction execution system, apparatus or device and execute the instructions, or a combination of the above. For the purposes of this specification, a "computer readable medium" can be any apparatus that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus or device. The computer readable medium can be a computer readable storage medium or a computer readable signal medium. The computer readable storage medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a propagation medium. The computer readable signal medium can include, but is not limited to, a computer readable medium that facilitates transfer of the program from one place to another. A specific example of a computer readable medium is a non-transitory computer-readable storage medium. A specific example of a computer readable signal medium is a source or destination of the computer readable medium. Another specific example of a computer readable signal medium is a computer readable signal travelling through space. Thus, a computer readable medium can take many forms of hardware to carry out the program for use by or in connection with the instruction execution system, apparatus or device.

[0143] It should be understood that aspects of the application can be implemented in hardware, software, firmware or combinations thereof. In the above embodiments, the N steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. If implemented in hardware and in another embodiment, the hardware can be implemented using any or a combination of the following technologies, which are each well known in the art: a discrete logic circuit(s) having logic gates for implementing logic functions upon an application of data signals, an application specific integrated circuit having appropriate combinational logic gates, a programmable gate array(s) (PGA), a field programmable gate array (FPGA), etc.

[0144] Those of skill in the art would understand that the steps of the methods carried out above can be carried out wholly or partly by a program instructing relevant hardware, and the program can be stored in a computer readable storage medium, and when executed, includes one or a combination of the steps of the method embodiments.

[0145] In addition, each of the functional units in the various embodiments of the present application can be integrated in one processing module, or each of the units can be physically present separately, or two or more units can be integrated in one module. The integrated module can be implemented in the form of hardware or in the form of a software functional module. When the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer readable storage medium.

[0146] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it should be understood that the above embodiments are exemplary and should not be construed as limiting the present application, and those skilled in the art can make changes, modifications, replacements and variations to the above embodiments within the scope of the present application.

Claims

1. A configuration optimization method for a dynamic reactive power compensation device of a new energy sending-end power grid, characterized in that: The following steps are involved: Based on the operating power corresponding to multiple benchmark operating points of the sending-end power grid, the specific goals of dynamic reactive power compensation device optimization are clarified, and the power grid data and historical operating data of the renewable energy sending end are collected; Extracting correlations and patterns between the grid data based on the grid data and the historical operation data to establish a reactive power demand forecasting model, and using the reactive power demand forecasting model to forecast reactive power compensation demand within a preset future time period, and combining analysis of reactive power distribution in the grid to determine at least one key area and at least one key node for reactive power compensation; Determining an optimal installation location and capacity of a reactive power compensation device based on grid characteristics, the at least one key area, the at least one key node, and the reactive power compensation demand, and selecting a corresponding type of dynamic reactive power compensation device; An optimization control strategy is generated according to the optimal installation position, the capacity and the type of the dynamic reactive compensation device, and the reactive compensation device is automatically switched on and off to complete the dynamic whole-machine debugging and testing tasks of the reactive compensation device.

2. The method according to claim 1, characterized in that Also includes: Establish a real-time feedback mechanism for key parameters of grid voltage and reactive power; Based on the real-time feedback mechanism, the parameters and thresholds of the optimization control strategy are adjusted according to real-time data.

3. The method according to claim 1 or 2, characterized in that Also includes: Simulating the optimization control strategy to generate an optimization control strategy that meets preset validity and stability conditions; And / or, iterating the optimization control strategy according to actual grid operation data and the actual operation status of the reactive power compensation device to obtain regular evaluation of the effect of the control strategy, and updating and upgrading the optimization control strategy according to the effect.

4. The method according to claim 3, characterized in that The optimization formula of the optimization control strategy is: Where s is the current state, a is the current action, r is the reward after executing the action, s′ is the next state, α is the learning rate, γ is the discount factor, and a′ is the optional action for the next state.

5. The method according to claim 1, wherein The determining of the optimal installation location and capacity of the reactive power compensation device based on the grid characteristics, the at least one key area, the at least one key node, and the reactive power compensation demand, and selecting the corresponding dynamic reactive power compensation device type includes: The particle swarm algorithm is used to simulate the movement of particles in the solution space to find the optimal solution to determine the optimal installation position and capacity of the reactive power compensation device. The types of dynamic reactive power compensation devices include static synchronous compensators and static VAR compensators. The speed update and position update expressions of the particle swarm algorithm are: V id t+1=ωV id t+c1r1(p id best-x id t)+c2r2(g d best-x id t) x id t+1=x id t+V id t+1 Among them, V id t+1 is the velocity of particle i in the t+1th generation in the dth dimension, V id t is the velocity of particle i in the tth generation in the dth dimension, x id t+1 is the position of particle i in the t+1th generation in the dth dimension, x id t is the position of particle i in the tth generation on the dth dimension, ω is the inertia weight, which is used to control the influence of the particle's previous generation velocity on the current velocity, c1 and c2 are learning factors, which represent the acceleration weights of the particle approaching its own historical optimal position and the global optimal position, respectively, r1 and r2 are random numbers uniformly distributed within a preset range, which introduce randomness into the search process, p id best is the individual historical optimal position of particle i in the dth dimension, g d best is the global historical optimal position of the entire particle swarm in the dth dimension.

6. A configuration optimization system for a dynamic reactive power compensation device of a new energy sending-end power grid, characterized in that: include: The acquisition module is used to determine the specific optimization goals of the dynamic reactive power compensation device based on the operating power corresponding to multiple benchmark operating points of the sending-end power grid, and to collect power grid data and historical operating data of the new energy sending end; a modeling module for extracting correlations and patterns among the grid data based on the grid data and the historical operation data to establish a reactive power demand forecasting model, using the reactive power demand forecasting model to forecast reactive power compensation demand within a preset future time period, and determining at least one key area and at least one key node for reactive power compensation in combination with analyzing the distribution of reactive power in the grid; a selection module for determining an optimal installation location and capacity of a reactive power compensation device based on grid characteristics, the at least one key area, the at least one key node, and the reactive power compensation demand, and selecting a corresponding type of dynamic reactive power compensation device; The optimization module is used to generate an optimization control strategy based on the optimal installation position, the capacity and the type of the dynamic reactive compensation device, automatically switch and adjust the reactive compensation device, and complete the dynamic whole machine debugging and testing tasks of the reactive compensation device.

7. The system according to claim 6, characterized in that Also includes: Feedback mechanism establishment module, used to establish a real-time feedback mechanism for key parameters of grid voltage and reactive power; The parameter threshold adjustment module is used to adjust the parameters and thresholds of the optimization control strategy according to real-time data based on the real-time feedback mechanism.

8. An electronic device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the configuration optimization method for the dynamic reactive power compensation device of the new energy sending-end power grid as described in any one of claims 1 to 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the configuration optimization method of the dynamic reactive power compensation device of the new energy sending-end power grid as described in any one of claims 1 to 5.

10. A computer program product comprising a computer program, characterized in that The computer program is executed to implement the configuration optimization method of the dynamic reactive power compensation device of the new energy sending-end power grid according to any one of claims 1 to 5.

Images