A virtual reactive power plant user single-phase asynchronous motor reactive power regulation method

Abstract

The application discloses a kind of virtual reactive power plant user single-phase asynchronous motor reactive power regulation methods, it is related to electrical equipment control technical field, comprising: the real-time state data of key node of distribution network is collected, short-term prediction model is constructed based on state data;According to the prediction result of model, setting determination condition, generating preliminary reactive power regulation instruction, and based on the dynamic update of prediction trend, the priority and regulation amplitude of preliminary reactive power regulation instruction are adjusted in real time;The preliminary reactive power regulation instruction after adjustment is issued to user terminal, and based on distributed coordination control, collaborative processing is carried out, and the final collaborative regulation instruction is generated;According to collaborative regulation instruction, the action of cutting off single-phase asynchronous motor split-phase capacitor is executed;The application is realized by using virtual reactive power plant regulation and control user side single-phase asynchronous motor split-phase capacitor, low cost, fast response and continuous accurate reactive power regulation, to effectively solve the problem of power grid reactive power surplus and high voltage.

Description

Technical Field

[0001] This invention relates to the field of electrical equipment control technology, and more specifically, to a method for reactive power regulation of a single-phase asynchronous motor in a virtual reactive power plant user. Background Technology

[0002] New power systems place high demands on reactive power compensation, requiring precise compensation in many situations. Traditional capacitors and reactors can be intelligently switched in stages, but frequent switching is limited by inrush current issues. While SVG and SVC, based on power electronics technology, utilize semiconductor devices to form controllable reactors or static var generators, enabling continuous and precise reactive power compensation without slow switching response, they generally suffer from high cost and significant losses. Therefore, there is an urgent need for a low-cost, continuously precise, fast-responding, and low-loss reactive power compensation and regulation method. With a high proportion of new energy distributed photovoltaic power integrated into the grid, the problems of excess reactive power and high voltage are particularly challenging. Therefore, it is crucial to effectively utilize existing equipment resources to absorb excess reactive power and control voltage.

[0003] To address the above problems, this invention proposes a solution. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a method for reactive power regulation of a user-side single-phase asynchronous motor using a virtual reactive power plant. By utilizing a virtual reactive power plant to regulate the phase-splitting capacitor of the user-side single-phase asynchronous motor, low-cost, fast-response, and continuous and accurate reactive power regulation is achieved, thereby effectively solving the problems of excess reactive power and high voltage in the power grid.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for reactive power regulation of a single-phase asynchronous motor in a virtual reactive power plant, comprising: collecting real-time status data of key nodes in the distribution network; constructing a short-term prediction model based on the status data to predict the reactive power and voltage change trends over a future period; setting judgment conditions based on the prediction results of the model to generate preliminary reactive power regulation commands; and adjusting the priority and adjustment range of the preliminary reactive power regulation commands in real time based on the dynamic updates of the predicted trends; sending the adjusted reactive power regulation commands to the user terminal and performing collaborative processing based on distributed coordinated control to generate final collaborative regulation commands; and executing the action of cutting off the phase-splitting capacitor of the single-phase asynchronous motor according to the collaborative regulation commands.

[0006] In a preferred embodiment, real-time status data of key nodes in the distribution network is collected, and a short-term prediction model is constructed based on the status data. Specifically, the input feature weights are assigned to the status data based on an adaptive neural network, and weighted fusion is performed to obtain comprehensive reactive power status features. The status data includes instantaneous reactive power disturbance data, phase voltage nonlinearity data, and load response delay data. A supervised learning dataset is constructed based on the comprehensive reactive power status features and future short-term reactive power and voltage. A bidirectional LSTM network is constructed, and the supervised learning dataset is input into the bidirectional LSTM network for training to obtain the trained short-term prediction model. The real-time comprehensive reactive power status features are input into the short-term prediction model to output the predicted values ​​of reactive power and voltage of the power grid in the short term.

[0007] In a preferred embodiment, an adaptive neural network is used to assign input feature weights to the state data and perform weighted fusion to obtain comprehensive reactive power state features. Specifically, a standardized feature matrix is ​​generated based on the state data, and the reactive power and voltage of the power grid in the short term are used as historical prediction targets. A neural network is constructed, and the standardized feature matrix and the corresponding historical prediction targets are input into the network. The mean square error loss function is used to obtain the error between the prediction result and the actual target. The network weights are iteratively updated according to the prediction error through the backpropagation algorithm until the loss function converges. Feature weights are extracted to form an adaptive weight matrix, and weighted fusion is performed with the standardized feature matrix to generate comprehensive reactive power state features.

[0008] In a preferred embodiment, a preliminary reactive power regulation command is generated by setting judgment conditions based on the model's prediction results. Specifically, this involves: acquiring the real-time voltage and real-time reactive power of the grid node; comparing the real-time voltage with a preset voltage upper limit and the real-time reactive power with a preset power threshold to obtain a first judgment result; acquiring the predicted voltage and reactive power values ​​of the grid node; comparing the predicted voltage with the voltage upper limit and the predicted reactive power with the power threshold to obtain a second judgment result; performing a logical judgment based on the first and second judgment results; and generating a preliminary reactive power regulation command when the first judgment result indicates a potential overvoltage and reactive power excess state, or when the second judgment result indicates an overvoltage and reactive power excess state.

[0009] In a preferred embodiment, the priority and adjustment range of the initial reactive power regulation command are adjusted in real time, specifically as follows: First data after the initial reactive power regulation command is generated is obtained; a regulation priority index is calculated based on the reactive power regulation weight associated with the first data; and the reactive power regulation weight is dynamically adjusted based on a Bayesian optimization algorithm. Multiple initial reactive power regulation commands are sorted according to the regulation priority index, with higher-priority commands executed before lower-priority commands. The current operating status of the controlled motor is obtained, and the initial adjustment range is calculated using a preset initial adjustment range calculation formula, combined with the first data. The initial adjustment range is optimized based on a robust control method to obtain an optimized adjustment range. The corresponding reactive power regulation commands are executed according to the sorted execution order and the optimized adjustment range, and the changes in grid voltage and reactive power are monitored in real time to obtain the deviation difference between the actual and predicted changes. The adjustment priority and adjustment range of subsequent reactive power regulation commands are dynamically adjusted based on the deviation difference to form the final reactive power regulation command.

[0010] In a preferred embodiment, the reactive power regulation weights are dynamically adjusted based on a Bayesian optimization algorithm. Specifically, the reactive power regulation weights of the system are considered as variables to be optimized, and an optimization objective function is constructed. Based on historical operating data, a Gaussian process prior model is constructed for the reactive power regulation weight space. Initial sampling points are selected in the weight space, the regulation priority index is calculated, and regulation operations are performed. The value of the corresponding objective function is calculated based on the regulation result, and a sampling function is constructed based on the predicted mean and variance of the Gaussian process prior model. The weight combination with the largest sampling function value is selected as the next sampling point, the corresponding objective function value is calculated, and the Gaussian process model is updated. The above selection, calculation, and update process is iteratively executed until a preset iteration termination condition is met. The reactive power regulation weight combination that minimizes the optimization objective function value is selected from all iterative samples as the final optimization result.

[0011] In a preferred embodiment, the initial adjustment range is optimized based on a robust control method to obtain an optimized adjustment range. Specifically, the second data is acquired and modeled as system uncertainty. The deviation between the currently measured actual voltage value and the target voltage reference value is used as an index to construct a robust optimization objective function and safety constraints. The robust optimization objective function is solved using an optimization algorithm to ensure that it meets the voltage and reactive power safety constraints while considering system uncertainty, thus obtaining the optimized adjustment range.

[0012] In a preferred embodiment, the adjusted reactive power regulation command is sent to the user terminal, and coordinated processing is performed based on distributed coordination control to generate the final coordinated regulation command. Specifically, after receiving the final reactive power regulation command, the user terminal obtains the current reactive power regulation capability of the motor; each user terminal sends its adjustable capability to neighboring nodes through a preset communication protocol, and simultaneously receives the capability information of neighboring nodes; based on its own adjustable capability and neighboring node information, the user terminal uses a consensus iterative algorithm to correct the execution ratio of the reactive power regulation command; when the iterative calculation converges, each user terminal generates the final coordinated regulation command according to the convergence result.

[0013] In a preferred embodiment, a consensus iterative algorithm is used to correct the execution ratio of the reactive power regulation command. After the iterative calculation converges, each user terminal generates the final coordinated regulation command based on the convergence result. Specifically, each user terminal calculates the initial regulation ratio based on the received initial reactive power regulation command and its own adjustable capability, and exchanges the regulation ratio information with neighboring nodes through the communication network. Each user terminal iteratively updates the regulation ratio according to the consensus iterative rules based on its local regulation ratio and the regulation ratio information of its neighboring nodes. After each iteration, the change in the local regulation ratio is calculated, and the change information is exchanged with neighboring nodes. When the change in the regulation ratio of all neighboring nodes is less than a preset change threshold, the iteration is considered to have converged. The final coordinated regulation command is generated by combining the converged regulation ratio and the user terminal's own adjustable capability.

[0014] In a preferred embodiment, according to the coordinated adjustment command, the action of disconnecting the split-phase capacitor of the single-phase asynchronous motor is executed, specifically: the user terminal obtains the real-time operating status parameters of the motor and performs condition judgment; 1) the rate of change of motor speed within a preset time window is less than a set threshold; 2) whether the motor temperature rise is within a preset allowable range; 3) whether the deviation of motor stator voltage and current relative to the rated value does not exceed a preset threshold; when all the above conditions are met, it is determined that the motor can safely perform the split-phase capacitor disconnection operation; under the condition that the motor operating conditions are met and the coordinated adjustment command is required, the action of disconnecting the split-phase capacitor of the single-phase asynchronous motor is executed; when the real-time state or predicted state does not meet the overvoltage and reactive power excess conditions, the user terminal restores the split-phase capacitor connection state.

[0015] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0016] 1. By combining short-term forecasting models, adaptive neural network feature weighting, bidirectional LSTM time series forecasting, Bayesian optimized weight adjustment, and robust control amplitude optimization, precise and dynamic regulation of the power grid's reactive power and voltage is achieved. Its advantages include: simultaneously considering the current state of the power grid and short-term future trends, achieving proactive and forward-looking reactive power regulation; improving the accuracy and adaptability of the regulation strategy through adaptive weighting and Bayesian optimization to dynamically allocate regulation priorities and amplitudes; optimizing the regulation amplitude under uncertainties such as system measurement errors, prediction deviations, and external disturbances, ensuring stable operation of voltage and power factor within safe constraints; and forming a closed-loop, adaptive, and intelligent reactive power regulation mechanism, which not only improves the power factor and voltage stability of the power grid but also enhances the reliability of regulation and system robustness, providing a solid guarantee for the safe and efficient operation of the power grid.

[0017] 2. By issuing adjusted reactive power regulation commands to user terminals and combining distributed coordinated control and consensus iterative algorithms, collaborative optimization regulation is achieved among each terminal under its own safety constraints, effectively controlling the reactive power and voltage levels of the entire network. Simultaneously, by utilizing the dynamic disconnection and restoration of split-phase capacitors in single-phase asynchronous motors, controllable resources within load equipment are converted into virtual reactive power, enabling rapid, continuous, and precise reactive power regulation, avoiding the cost of adding additional capacitors or electronic compensation devices. This scheme balances real-time response, predictive foresight, safety constraints, and network-wide coordination, not only improving the reliability and robustness of regulation but also significantly enhancing the power factor optimization capability and voltage stability of the grid, achieving the efficient operation of a virtual reactive power plant. Attached Figure Description

[0018] Figure 1 This is a schematic flowchart of a method for reactive power regulation of a single-phase asynchronous motor in a virtual reactive power plant, provided in an embodiment of this application.

[0019] Figure 2 A comparison chart of short-term power grid reactive power and voltage prediction provided for embodiments of this application. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0021] Reference Figure 1 As shown in the diagram, this invention provides a flowchart of a method for reactive power regulation of a single-phase asynchronous motor in a virtual reactive power plant, comprising the following steps:

[0022] S1 collects real-time status data of key nodes in the distribution network and constructs short-term prediction models for reactive power and voltage changes in the power grid.

[0023] In this embodiment, real-time status data of key nodes in the distribution network are collected, and a short-term prediction model is constructed based on the status data to predict the reactive power change and voltage change trends over a future period of time. Specifically:

[0024] Real-time status data of key nodes in the distribution network are collected, including instantaneous reactive power disturbance data, phase voltage nonlinearity data, and load response delay data.

[0025] The adaptive neural network assigns input feature weights to the state data and performs weighted fusion to obtain the comprehensive reactive power state features.

[0026] Based on a preset short-term prediction window, the comprehensive reactive power state features are used as the time series input, and the future short-term power grid reactive power and voltage are used as the output sequence to construct a supervised learning dataset.

[0027] A bidirectional LSTM network is constructed, including an input layer, a bidirectional LSTM hidden layer, and an output layer. The number of nodes in the input layer is consistent with the dimension of the comprehensive reactive power state features. The number of nodes in the hidden layer is determined based on cross-validation or experience. The bidirectional LSTM hidden layer captures the dependence of past and future features simultaneously through two time series paths, forward and backward, to improve the short-term fluctuation prediction capability. The output layer uses a linear activation function to make continuous value predictions of future reactive power and voltage.

[0028] like Figure 2 As shown, the time series input and output sequences are input into a bidirectional LSTM network, and the mean squared error loss function is used to perform iterative training through gradient descent until the preset number of convergences is reached, thus obtaining the trained short-term prediction model.

[0029] The real-time comprehensive reactive power state characteristics are input into the short-term prediction model, and the predicted values ​​of reactive power and voltage of the power grid in the short term are output.

[0030] It should be noted that by constructing a short-term prediction model through a bidirectional LSTM network, the past trends and potential future dependencies of the power grid's reactive power characteristics can be captured simultaneously, enabling high-precision prediction of short-term reactive power and voltage fluctuations. Combined with the comprehensive feature input of the adaptive neural network weighted fusion, the model can dynamically identify the importance of each feature to the prediction results, improve the stability and robustness of the prediction, and thus provide timely and reliable decision-making basis for power grid reactive power regulation, significantly enhancing the safety of power grid operation and the power factor optimization capability.

[0031] Furthermore, based on an adaptive neural network, input feature weights are assigned to the power grid state data, and weighted fusion is performed to obtain comprehensive reactive power state features, specifically:

[0032] Standardize the power grid status data to generate a standardized feature matrix;

[0033] Using the reactive power and voltage of the power grid in the near future as historical prediction targets, labels are provided for model training.

[0034] Construct a neural network with an input layer, hidden layers, and an output layer. The nodes in the input layer correspond to the standardized feature matrices, the activation function of the hidden layer is ReLU, the output layer is linearly activated, and the network weights are initialized to have a mean of 0.

[0035] The standardized feature matrix and the corresponding historical predicted target are input into the network, and the error between the predicted result and the actual target is obtained by using the mean squared error loss function.

[0036] The network weights are iteratively updated based on the prediction error using the backpropagation algorithm until the loss function converges.

[0037] Feature weights are extracted from the connection weights from the input layer to the output layer after training to form an adaptive weight matrix;

[0038] The adaptive weight matrix and the standardized feature matrix are weighted and fused to generate a comprehensive reactive power state feature. The comprehensive reactive power state feature is used in the short-term prediction model to achieve high-precision prediction of future power grid reactive power and voltage deviation.

[0039] During the actual operation of the model, the prediction error is fed back to the adaptive neural network in real time, and the weight matrix is ​​dynamically adjusted according to the latest error to ensure that the contribution of each feature to the prediction result is adaptively optimized as the power grid operating status changes.

[0040] Instantaneous reactive power disturbance data is an indicator that quantifies the amplitude of reactive power fluctuations in the power grid over a short period. It reflects the strength of instantaneous reactive power disturbances in the power grid and can capture short-term power fluctuations caused by load abrupt changes, partial equipment switching, or instantaneous reactive power injection. By acquiring instantaneous reactive power disturbance data and using it as input to a short-term prediction model, the short-term fluctuation characteristics of the power grid's reactive power can be quantified in real time, capturing instantaneous disturbance trends. Building a prediction model based on this quantitative indicator helps to identify potential risks of excess reactive power and high voltage in the next few minutes, thereby enabling proactive adjustment and optimized control, reducing power grid voltage fluctuations, improving power factor stability, and simultaneously reducing equipment losses and power quality problems caused by excess reactive power. The specific steps for acquiring the instantaneous reactive power disturbance data are as follows:

[0041] Real-time acquisition of reactive power of each phase ;

[0042] Calculate the rate of change of reactive power within a preset time window;

[0043] The specific formula for calculating the rate of change of reactive power is as follows:

[0044]

[0045] In the formula, For the rate of change, The sampling interval is... To avoid dividing by zero, small constants For the previous sampling interval Reactive power at any given moment.

[0046] The instantaneous reactive power disturbance data is obtained by averaging N consecutive sampling points of reactive power.

[0047]

[0048] In the formula, This represents instantaneous reactive power disturbance data.

[0049] It should be noted that the higher the instantaneous reactive power disturbance data, the more drastic the short-term reactive power fluctuations in the power grid, indicating a risk of excess or sudden changes.

[0050] Phase voltage nonlinearity data is an indicator that measures the degree to which a voltage waveform deviates from an ideal sine wave. It is represented by the ratio of higher-order harmonic components to the fundamental component. By acquiring phase voltage nonlinearity data and using it as input to a short-term prediction model, the higher-order harmonic components and distortion degree of the voltage waveform can be quantified, reflecting potential reactive power anomalies and local voltage rises in the power grid within a short period. Building a prediction model based on this indicator helps to identify potential risks of excess reactive power and high voltage in the power grid in advance, enabling more precise proactive regulation, improving the stability and power quality of power grid operation, and reducing equipment losses and failure probabilities caused by voltage waveform distortion. The specific steps for acquiring the phase voltage nonlinearity data are as follows:

[0051] Real-time acquisition of phase voltage Perform a Fast Fourier Transform on the sampled waveform to obtain the fundamental component. and higher-order harmonic components ;

[0052] Phase voltage nonlinearity data are obtained by calculating the ratio of higher-order harmonic components to the fundamental component.

[0053] The specific calculation formula for the phase voltage nonlinear data is as follows:

[0054]

[0055] In the formula, This is nonlinear data for phase voltage.

[0056] It should be noted that the higher the phase voltage nonlinearity data, the more severe the voltage waveform distortion, accompanied by local reactive power excess and voltage anomalies.

[0057] Load response delay data measures the lag time of user-side load response to voltage or reactive power changes. By acquiring load response delay data and using it as input to short-term prediction models, the lag characteristics of user-side load response to voltage and reactive power changes can be quantified, revealing potential reactive power accumulation or voltage rise trends in the power grid within a short period. Building prediction models based on this indicator helps to identify the risks of reactive power excess and voltage anomalies in the power grid in advance, enabling proactive regulation, improving the dynamic response capability of the power grid, and optimizing power factor control. The specific steps for acquiring the load response delay data are as follows:

[0058] Collect grid voltage and load current Calculate the cross-correlation function between voltage change and load change. ;

[0059] The cross-correlation function between voltage change and load change is calculated using the following formula:

[0060]

[0061] In the formula, The average voltage. This is the average current. This represents the lag time of the current sequence relative to the voltage sequence.

[0062] Traversing different Value, found Maximum lag time This yields load response delay data.

[0063] The specific calculation formula for the load response delay data is as follows:

[0064]

[0065] It should be noted that the larger the load response delay data, the more obvious the load response lag, and the more likely the power grid will experience reactive power surplus or high voltage in the short term.

[0066] S2 sets judgment conditions based on the model's prediction results, generates preliminary reactive power adjustment commands, and adjusts the priority and adjustment range of the preliminary reactive power adjustment commands in real time based on the dynamic update of the prediction trend.

[0067] In this embodiment, judgment conditions are set based on the model's prediction results to generate preliminary reactive power adjustment instructions, specifically as follows:

[0068] The predicted values ​​of reactive power and voltage of the power grid in the short term are obtained based on the short-term forecasting model.

[0069] Obtain real-time voltage and real-time reactive power;

[0070] The real-time voltage is compared with the preset voltage upper limit, and the real-time reactive power is compared with the preset power threshold.

[0071] If the real-time voltage is greater than the preset voltage limit and the real-time reactive power is greater than or equal to the preset power threshold, then the current real-time state is determined to be a potential overvoltage and reactive power excess state, and recorded as the first determination result.

[0072] The predicted voltage value is compared with the preset voltage upper limit, and the predicted reactive power value is compared with the preset power threshold.

[0073] If the predicted voltage value is greater than the preset voltage upper limit and the predicted reactive power value is greater than or equal to the preset power threshold, the predicted state is determined to be overvoltage and reactive power excess state, and recorded as the second determination result.

[0074] Logical determination is performed based on the first determination result and the second determination result;

[0075] When the first determination result indicates the existence of the potential overvoltage and reactive power excess risk state, or when the second determination result indicates the existence of the overvoltage and reactive power excess state, a preliminary reactive power adjustment command is generated.

[0076] It should be noted that the purpose of judging the real-time grid status and predicting the grid status is to take into account both the current actual operating conditions and the short-term future trends, thereby achieving safer, more proactive, and more precise reactive power regulation: the real-time status ensures that it can respond immediately when overvoltage and reactive power excess are detected, ensuring current safety; the predictive status identifies possible future anomalies in advance, providing forward-looking guidance for regulation actions, making the regulation amplitude smooth and the sequence optimized, reducing the impact on motors and the grid; the combination of the two forms an intelligent regulation strategy that can respond instantly and predictively, improving system stability and regulation accuracy.

[0077] Furthermore, the priority and adjustment range of the initial reactive power adjustment command are adjusted in real time, specifically as follows:

[0078] The first data obtained after generating the initial reactive power adjustment command includes the magnitude of the predicted voltage exceeding the upper limit (the difference between the predicted voltage and the preset upper voltage limit), the degree of the reactive power prediction value being too high (the difference between the reactive power prediction value and the preset power threshold), and the safety margin of the single-phase asynchronous motor.

[0079] The regulation priority index is calculated based on the first data and its corresponding reactive power regulation weight, and the reactive power regulation weight is dynamically adjusted based on the Bayesian optimization algorithm.

[0080] The specific calculation formula for the adjustment priority index is as follows:

[0081]

[0082] In the formula, To adjust priority indicators, To predict the extent to which the voltage exceeds the upper limit, To determine the degree to which the predicted reactive power is too high. For the safety margin of a single-phase asynchronous motor, The reactive power regulation weight is used to predict the extent to which the voltage exceeds the upper limit. The reactive power adjustment weighting for the degree to which the reactive power prediction is too high. The reactive power regulation weight is used to ensure the safety margin of a single-phase asynchronous motor.

[0083] Multiple preliminary reactive power adjustment commands are sorted (from largest to smallest) according to the adjustment priority index, with higher priority commands taking precedence over lower priority commands in the execution order.

[0084] The current operating status of the controlled motor is obtained and combined with the first data. The preliminary adjustment range is calculated using a preset preliminary adjustment range calculation formula.

[0085] The formula for calculating the initial adjustment range is as follows:

[0086]

[0087] In the formula, For the initial adjustment range, , To adjust the weighting coefficients, This is the motor state coefficient. It is set to 1 when the motor is in a steady state, and otherwise a coefficient less than 1 is used.

[0088] The initial adjustment range is optimized based on a robust control method to obtain the optimized adjustment range;

[0089] Execute the corresponding reactive power regulation commands according to the sorted execution order and optimized adjustment range, and monitor the changes in grid voltage and reactive power in real time to obtain the deviation difference between the actual change value and the predicted change value.

[0090] The adjustment priority and adjustment range of subsequent reactive power adjustment instructions are dynamically adjusted based on the deviation difference to form the final reactive power adjustment instruction, which includes the target adjustment object, execution order, adjustment range and execution conditions.

[0091] The specific calculation formulas for the dynamic adjustment priority and adjustment range are as follows:

[0092]

[0093]

[0094] In the formula, For dynamically adjusted adjustment priorities, , To adjust the coefficient, The adjustment range is dynamically adjusted. This represents the deviation difference.

[0095] Furthermore, the reactive power regulation weights are dynamically adjusted based on the Bayesian optimization algorithm, specifically as follows:

[0096] Adjust the reactive power weight ( , , () are considered as variables to be optimized, and an optimization objective function is constructed;

[0097] The optimization objective function The specific calculation formula is as follows:

[0098]

[0099] In the formula, V is the actual value of the voltage. The target voltage reference value is given, and PF is the actual reactive power value. The target reactive power reference value, , To adjust performance weights.

[0100] Based on historical operating data, the reactive power adjustment weight space , , Construct a Gaussian process prior model to describe the optimization objective function. Uncertainty and potential distribution;

[0101] Selecting initial sampling points in the weight space The adjustment is then substituted into the adjustment priority index for calculation, and the adjustment is performed according to the priority to obtain the adjustment effect. The adjustment effect refers to the degree to which the power grid state approaches the ideal target reference value after the command is executed.

[0102] The value of the objective function corresponding to the adjustment effect is calculated and used as a preliminary observation sample to update the Gaussian process model;

[0103] Based on the predicted mean of the Gaussian process model and variance Construct a data acquisition function, which is the Expected Improvement (EI).

[0104] Select the weight combination with the largest sampling function value as the next sampling point, calculate the value of the corresponding objective function, and update the Gaussian process model. Iterate through the above selection, calculation and update process until the preset maximum number of iterations is reached or the objective function converges.

[0105] Select the reactive power adjustment weight combination that minimizes the value of the objective function from all iterative samples.

[0106] It should be noted that by dynamically adjusting reactive power regulation weights through Bayesian optimization, the weight combination can adaptively match the grid operating characteristics and regulation objectives, thereby maximizing the regulation effect while ensuring that voltage and power factor are close to the target reference values. Specific benefits include: firstly, improved adaptability and accuracy of the reactive power regulation strategy, making the regulation priority and magnitude more reasonable under different operating conditions; secondly, efficient exploration and optimization of the weight space through Gaussian process prediction and expectation-based improvement of the acquisition function, reducing reliance on manual parameter adjustment, and improving the reliability and stability of the strategy under complex grid conditions, providing a reliable foundation for subsequent real-time reactive power regulation and intelligent control.

[0107] Furthermore, the initial adjustment range is optimized based on a robust control method to obtain the optimized adjustment range, specifically:

[0108] Obtain the second set of data and model it as system uncertainty. The second data includes power grid measurement errors, prediction errors, and external disturbances;

[0109] Using the deviation between the currently measured actual voltage value and the target voltage reference value as an indicator, a robust optimization objective function and voltage and reactive power safety constraints are constructed.

[0110] Right now , ;

[0111] The specific calculation formula for the robust optimization objective function is as follows:

[0112]

[0113] In the formula, The robust adjustment amplitude is the optimal amplitude for reactive power regulation considering system uncertainties (measurement errors, prediction errors, and external disturbances). Let Δ be the value of the system uncertainty, and u be the set of uncertainties. for Under the influence of system uncertainty The voltage change that is affected For norm operations, To account for the uncertainty of the worst-case scenario Under the premise of minimizing voltage deviation, select the adjustment range. This is the core idea of ​​robust optimization.

[0114] The robust optimization objective function is solved using an optimization algorithm to ensure that it meets the safety constraints of voltage and reactive power while considering system uncertainties, thus obtaining the optimized regulation range.

[0115] It should be noted that by obtaining the initial adjustment amplitude and priority, and constructing a robust adjustment objective function under a system uncertainty model, the optimal solution and output of the reactive power adjustment amplitude can be achieved. The advantage is that regardless of changes in grid measurement errors, prediction errors, or external disturbances, the adjustment amplitude remains consistent. All of these technologies can adaptively optimize while meeting voltage and power factor safety constraints, thereby ensuring grid voltage stability and power factor close to the target value, improving the reliability and execution accuracy of reactive power regulation strategies, providing a solid foundation for subsequent dynamic regulation and intelligent control, and reducing the risk of grid overvoltage or power factor deviation caused by uncertainties.

[0116] S3 sends the adjusted reactive power regulation command to the user terminal and performs collaborative processing based on distributed coordination control to generate the final collaborative regulation command.

[0117] In this embodiment, the adjusted reactive power regulation command is sent to the user terminal, and coordinated processing is performed based on distributed coordination control to generate the final coordinated regulation command, specifically as follows:

[0118] After receiving the final reactive power adjustment command, the user terminal obtains the current reactive power adjustment capability of the motor, which includes the maximum reactive power that can be absorbed or released and the corresponding adjustment range.

[0119] Each user terminal sends its local adjustable capabilities to neighboring nodes through a preset communication protocol, and at the same time receives capability information from neighboring nodes, thus realizing information exchange between the local and neighboring nodes;

[0120] Based on its own adjustability and neighbor node information, the user terminal uses a consensus iterative algorithm to correct the execution ratio of the reactive power adjustment command. When the iterative calculation converges, each user terminal generates the final coordinated adjustment command based on the convergence result.

[0121] Furthermore, a consensus iterative algorithm is used to correct the execution ratio of the reactive power regulation command. After the iterative calculation converges, each user terminal generates the final coordinated regulation command based on the convergence result, specifically:

[0122] Each user terminal calculates the initial adjustment ratio based on the received reactive power adjustment command and its own adjustment capability, and exchanges the adjustment ratio information with neighboring nodes through the communication network.

[0123]

[0124] In the formula, This is the initial adjustment ratio. The maximum adjustable reactive power of terminal i. This represents the minimum adjustable reactive power of terminal i. The reactive power regulation target is the reactive power regulation command after adjustment, and K is the total number of user terminals participating in the regulation. It represents the total adjustable capabilities of all participating terminals across the entire network.

[0125] The user terminal iteratively updates the adjustment ratio according to the neighbor information and the local initial adjustment ratio, following a consistency iteration rule:

[0126]

[0127] In the formula, Let i be the adjustment ratio for user terminal i in the (g+1)th iteration. Let i be the adjustment ratio of user terminal i in the g-th iteration. This is the iteration step size coefficient, used to control the adjustment range in each iteration. Let i be the set of neighboring nodes of user terminal i. Let be the weight coefficient of neighbor node j with respect to node i. Let be the adjustment ratio of node j, a neighbor of node i, in the g-th iteration.

[0128] After each iteration, calculate the change in the local adjustment ratio, when the changes in all neighboring nodes... When all values ​​are less than a preset threshold for change, the iteration is considered to have converged.

[0129] The final coordinated adjustment command is generated by combining the adjustment ratio after iterative convergence with its own adjustability.

[0130] It should be noted that by adopting a distributed coordinated control approach, single-point failures and communication bottlenecks caused by centralized control can be effectively avoided. This allows each user terminal to meet its own operational safety constraints while collaboratively optimizing the reactive power and voltage levels of the entire network. At the same time, the real-time performance and robustness of regulation are improved through local fast calculation and neighbor iterative convergence mechanisms, avoiding regulation conflicts. Furthermore, the dynamic feedback update process enhances the adaptive capability of the power grid in fluctuating environments, thereby achieving the safety, coordination, and stability of reactive power regulation across the entire network.

[0131] S4, execute the action of cutting off the split-phase capacitor of the single-phase asynchronous motor according to the coordinated adjustment command.

[0132] In this embodiment, the action of cutting off the phase-splitting capacitor of the single-phase asynchronous motor according to the coordinated adjustment command is specifically as follows:

[0133] The user terminal obtains the real-time operating status parameters of the motor and makes conditional judgments. The real-time operating status parameters include motor speed, motor temperature rise, and motor stator voltage and current.

[0134] The determination is made based on the operating status parameters of the motor, specifically:

[0135] 1) The rate of change of motor speed within a preset time window is less than a set threshold;

[0136] 2) Whether the motor temperature rise is within the preset allowable range;

[0137] 3) Whether the deviation of the motor stator voltage and current from their rated values ​​does not exceed the preset threshold;

[0138] When the operating status judgment result meets all the above conditions, it is determined that the motor can safely perform the phase-splitting capacitor cut-off operation.

[0139] When the motor operating conditions are met and the coordinated adjustment command is required, the action of cutting off the split-phase capacitor of the single-phase asynchronous motor is executed. The operation is achieved by disconnecting the capacitor from the line, which reduces the power factor of the motor and thus actively absorbs the excess reactive power of the power grid.

[0140] When the real-time or predicted status does not meet the overvoltage and reactive power excess conditions, the user terminal restores the split-phase capacitor connection status.

[0141] It should be noted that single-phase asynchronous motors are widely used in users' low-voltage electrical equipment. These motors rely on split-phase capacitors to generate starting torque. After starting, the capacitors compensate for reactive power and improve the motor's power factor. Disconnecting the split-phase capacitors after starting, without considering the power factor, has minimal impact on motor performance. This invention utilizes this characteristic to propose the concept of a virtual reactive power plant. Through a centralized reactive power control system, commands are wirelessly issued. When a local motor meets certain operating conditions, such as being in a non-starting state and having normal current, voltage, and temperature rise, it has the ability to disconnect the split-phase capacitors to absorb excess reactive power for the grid. In this case, the invention issues commands via a wireless network to disconnect the split-phase capacitors, reducing the motor's power factor and achieving grid-wide reactive power optimization and regulation, thus realizing the effect of a virtual reactive power plant. Most reactive power regulation improvements require adding capacitors, reactors, or electronic compensation equipment. This invention cleverly utilizes the controllable resources of the capacitors within the load equipment. By deeply adjusting the switching of the split-phase capacitors of the user's motor, significant and continuous precise changes in reactive power can be achieved within seconds.

[0142] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0143] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0144] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0145] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0146] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0147] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for reactive power regulation of a single-phase asynchronous motor in a virtual reactive power plant user, characterized in that, include: Real-time status data of key nodes in the distribution network is collected, and a short-term prediction model is built based on the status data to predict the trends of reactive power and voltage changes over a future period. Specifically: The state data is assigned input feature weights based on an adaptive neural network and then weighted and fused to obtain comprehensive reactive power state characteristics; the state data includes instantaneous reactive power disturbance data, phase voltage nonlinear data, and load response delay data. A supervised learning dataset is constructed based on the comprehensive reactive power state characteristics and the future short-term reactive power and voltage. Construct a bidirectional LSTM network and input the supervised learning dataset into the bidirectional LSTM network for training to obtain a trained short-term prediction model. The real-time comprehensive reactive power state characteristics are input into the short-term prediction model, and the predicted values ​​of reactive power and voltage of the power grid in the short term are output. Based on the model's prediction results, decision criteria are set, preliminary reactive power adjustment commands are generated, and the priority and adjustment range of the preliminary reactive power adjustment commands are adjusted in real time based on dynamic updates of the prediction trend. Specifically: The first data after the initial reactive power regulation instruction is generated is obtained, and the regulation priority index is calculated by combining the reactive power regulation weight associated with the first data. The reactive power regulation weight is then dynamically adjusted based on the Bayesian optimization algorithm. Multiple preliminary reactive power adjustment commands are sorted according to the adjustment priority index, and commands with higher priority are executed before commands with lower priority. The current operating status of the controlled motor is obtained, and the initial adjustment range is calculated by combining the first data with the preset initial adjustment range calculation formula. The initial adjustment range is optimized based on a robust control method to obtain the optimized adjustment range. Execute the corresponding reactive power regulation commands according to the sorted execution order and optimized adjustment range, and monitor the changes in grid voltage and reactive power in real time to obtain the deviation difference between the actual change value and the predicted change value. The adjustment priority and adjustment range of subsequent reactive power adjustment commands are dynamically adjusted based on the deviation difference to form the final reactive power adjustment command. The adjusted reactive power regulation command is sent to the user terminal and processed collaboratively based on distributed coordination control to generate the final coordinated regulation command, specifically: After receiving the final reactive power adjustment command, the user terminal obtains the current reactive power adjustment capability of the motor. Each user terminal sends its adjustable capabilities to neighboring nodes via a preset communication protocol, and simultaneously receives capability information from neighboring nodes. Based on its own adjustability and neighbor node information, the user terminal uses a consensus iterative algorithm to correct the execution ratio of the reactive power adjustment command; when the iterative calculation converges, each user terminal generates the final coordinated adjustment command according to the convergence result. According to the coordinated adjustment command, the action of cutting off the split-phase capacitor of the single-phase asynchronous motor is executed.

2. The reactive power regulation method for single-phase asynchronous motors in virtual reactive power plants according to claim 1, characterized in that, The method of assigning input feature weights to the state data based on an adaptive neural network and then performing weighted fusion to obtain comprehensive reactive power state features is as follows: A standardized feature matrix is ​​generated based on state data, and the power grid reactive power and voltage in the near future are used as historical prediction targets. A neural network is constructed by inputting a standardized feature matrix and the corresponding historical prediction target into the network, and the error between the prediction result and the actual target is obtained by using the mean squared error loss function. The network weights are iteratively updated based on the prediction error using the backpropagation algorithm until the loss function converges. Feature weights are extracted to form an adaptive weight matrix, which is then weighted and fused with the standardized feature matrix to generate comprehensive reactive power state features.

3. The reactive power regulation method for single-phase asynchronous motors in virtual reactive power plants according to claim 1, characterized in that, The step of setting judgment conditions based on the model's prediction results and generating preliminary reactive power adjustment instructions specifically involves: Obtain the real-time voltage and real-time reactive power of the power grid nodes; The real-time voltage is compared with the preset voltage upper limit, and the real-time reactive power is compared with the preset power threshold to obtain the first judgment result; Obtain the predicted voltage and reactive power values ​​of the power grid nodes; The voltage prediction value is compared with the voltage upper limit, and the reactive power prediction value is compared with the power threshold to obtain a second determination result; Logical determination is performed based on the first determination result and the second determination result; When the first determination result indicates the existence of potential overvoltage and excess reactive power, or when the second determination result indicates the existence of overvoltage and excess reactive power, a preliminary reactive power adjustment command is generated.

4. The reactive power regulation method for single-phase asynchronous motors in virtual reactive power plants according to claim 1, characterized in that, The dynamic adjustment of reactive power regulation weights based on the Bayesian optimization algorithm is specifically as follows: The reactive power regulation weight of the system is regarded as the variable to be optimized, and an optimization objective function is constructed. Based on historical operating data, a Gaussian process prior model is constructed for the reactive power regulation weight space; Select initial sampling points in the weight space, calculate the adjustment priority index, and perform adjustment operations; The value of the corresponding objective function is calculated based on the adjustment results, and the acquisition function is constructed based on the predicted mean and variance of the Gaussian process prior model. Select the weight combination with the largest sampling function value as the next sampling point, calculate the corresponding objective function value, and update the Gaussian process model; The above selection, calculation and update process is executed iteratively until the preset iteration termination condition is met; Select the reactive power adjustment weight combination that minimizes the objective function value from all iterative samples as the final optimization result.

5. The reactive power regulation method for single-phase asynchronous motors in virtual reactive power plants according to claim 1, characterized in that, The initial adjustment range is optimized using a robust control method to obtain the optimized adjustment range, specifically as follows: The second data is acquired and modeled as system uncertainty. The deviation between the current measured voltage value and the target voltage reference value is used as the index to construct a robust optimization objective function and safety constraints. The robust optimization objective function is solved using an optimization algorithm to ensure that it meets the safety constraints of voltage and reactive power while considering system uncertainties, thus obtaining the optimized regulation range.

6. The reactive power regulation method for single-phase asynchronous motors in virtual reactive power plants according to claim 1, characterized in that, The consistency iterative algorithm is used to correct the execution ratio of the reactive power regulation command; after the iterative calculation converges, each user terminal generates the final coordinated regulation command based on the convergence result, specifically: Each user terminal calculates the initial adjustment ratio based on the received initial reactive power adjustment command and its own adjustment capability, and exchanges the adjustment ratio information with neighboring nodes through the communication network. Each user terminal updates its adjustment ratio iteratively according to the consistency iteration rules, based on the local adjustment ratio and the adjustment ratio information of neighboring nodes. After each iteration, the change in the local adjustment ratio is calculated and the change information is exchanged with neighboring nodes. When the change in the adjustment ratio of all neighboring nodes is less than the preset change threshold, the iteration is considered to have converged. The final coordinated adjustment command is generated by combining the adjustment ratio after iterative convergence with its own adjustability.

7. The reactive power regulation method for a single-phase asynchronous motor in a virtual reactive power plant according to claim 6, characterized in that, The action of cutting off the split-phase capacitor of the single-phase asynchronous motor according to the coordinated adjustment command is specifically as follows: Obtain the real-time operating status parameters of the user terminal based on the motor it belongs to and make conditional judgments; 1) The rate of change of motor speed within a preset time window is less than a set threshold; 2) Whether the motor temperature rise is within the preset allowable range; 3) Whether the deviation of the motor stator voltage and current from their rated values ​​does not exceed the preset threshold; When all of the above conditions are met, it is determined that the motor can safely perform the split capacitor cut-off operation; Under the condition that the motor operating conditions are met and the requirements of the coordinated adjustment command are met, the action of cutting off the split-phase capacitor of the single-phase asynchronous motor is executed. When the real-time or predicted status does not meet the overvoltage and reactive power excess conditions, the user terminal restores the split-phase capacitor connection status.

Images