A method and system for harmonic optimization of a photovoltaic power distribution network

Abstract

The present application belongs to the technical field of photovoltaic power distribution network, and provides a harmonic optimization method and system for photovoltaic power distribution network, comprising: collecting harmonic related information of the photovoltaic power distribution network in real time, integrating and generating a harmonic information sequence, and judging whether harmonic optimization is needed; if needed, marking an optimization period and analyzing harmonic generating factors, processing and intercepting the harmonic information sequence to obtain a training data set including a precursor window matrix, constructing a harmonic precursor identification model and training the model using the training data set; if the current time is not in the optimization period, using the trained harmonic precursor identification model to identify, judging whether the optimization period is triggered, if triggered, determining the harmonic generating factors according to the identification result, in the optimization period, selecting a harmonic optimization compensation mechanism based on different harmonic generating factors, optimizing and compensating the photovoltaic power distribution network, and judging whether the optimization and compensation can be stopped in the optimization period, if yes, ending the current optimization period.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of photovoltaic power distribution networks, in particular to a harmonic optimization method and system for a photovoltaic power distribution network. BACKGROUND

[0002] As an important form of renewable energy access to the power grid, the scale and penetration rate of photovoltaic power distribution networks are constantly improving. However, the nonlinear characteristics of a large number of power electronic devices in photovoltaic power stations and the dynamic nature of power grid operation scenarios have led to increasingly prominent harmonic problems. Traditional power distribution network harmonic control methods have exposed multiple technical bottlenecks when applied to photovoltaic scenarios.

[0003] The prior art only relies on voltage total distortion rate threshold to determine whether optimization is needed, does not integrate scene data such as illumination intensity, short-circuit capacity ratio, and load power factor, lacks a multi-parameter collaborative analysis mechanism, leading to inaccurate identification of harmonic generating factors and insufficient targetedness of control strategies. Traditional harmonic compensation devices rely on passive triggering after the voltage distortion rate exceeds the threshold, and cannot capture the precursor characteristics before the occurrence of harmonics. In photovoltaic power distribution networks, factors such as illumination intensity fluctuations and short-circuit capacity changes have obvious time sequence correlations. For example, high-frequency fluctuations in illumination intensity under cloudy weather may indicate an impending harmonic standard violation. However, the prior art does not establish a correlation model between time sequence data and harmonic risk, leading to delayed compensation actions and a high risk of device overload or power grid transient impact.

[0004] On the other hand, different harmonic generating factors require differentiated treatment. Illumination fluctuations require real-time tracking of compensation current, short-circuit capacity deficiencies require dynamic adjustment of reactive power, and low load power factors require filtering of specific frequency harmonics. However, the prior art generally uses a single compensation device or a fixed parameter control strategy, which cannot switch compensation mechanisms according to real-time causes, leading to low treatment efficiency and even causing resonance risks between compensation devices and the power grid. Existing methods usually use voltage distortion rate compliance as the only termination condition, ignoring whether the harmonic generating factors have disappeared. For example, when the short-circuit capacity ratio is insufficient, stopping compensation only because the voltage distortion rate temporarily meets the standard may lead to repeated harmonic problems or continuous operation of compensation devices after the cause disappears, resulting in device wear and energy waste.

[0005] To address the above problems, the present application proposes a harmonic optimization method and system for a photovoltaic power distribution network. SUMMARY

[0006] To address the above problems, the present application proposes a harmonic optimization method and system for a photovoltaic power distribution network.

[0007] The technical solution adopted by the application to solve the technical problems is: a harmonic optimization method for a photovoltaic power distribution network, comprising:

[0008] Real-time acquisition of harmonic related information of photovoltaic power distribution network, integration of harmonic information sequence, and judgment of whether harmonic optimization is needed, if needed, mark the optimization period and analyze the harmonic generating factors;

[0009] The harmonic information sequence is processed and intercepted to obtain a training data set including a precursor window matrix, a harmonic precursor identification model is constructed and trained using the training data set, if the current time is not in the optimization period, the trained harmonic precursor identification model is used for identification to determine whether the optimization period is triggered, if triggered, the harmonic generating factors are determined according to the identification result;

[0010] In the optimization period, different harmonic generating factors are selected to select a harmonic optimization compensation mechanism, the photovoltaic power distribution network is optimized and compensated, and it is judged in real time whether the optimization and compensation can be stopped, if so, the current optimization period is ended;

[0011] The acquisition method of the harmonic information sequence is as follows:

[0012] The harmonic related information includes the total voltage distortion rate of the photovoltaic power distribution network and corresponding scene data, and the scene data includes the light intensity, the short circuit capacity ratio and the load power factor.

[0013] A historical acquisition period ending at the current time is set, in each time point in the historical acquisition period, the harmonic related information of the photovoltaic power distribution network obtained by acquisition is integrated into a harmonic related information vector, and the harmonic information sequence is obtained according to the time sequence.

[0014] The marking method of the optimization period is as follows:

[0015] The harmonic related information is obtained, the harmonic related information includes the total voltage distortion rate of the photovoltaic power distribution network, a start flag value is set and its initial value is 0, the total voltage distortion rate of the current time is compared with the preset total voltage distortion rate threshold, if it is greater than the total voltage distortion rate threshold, it is judged that the photovoltaic power distribution network needs to be optimized.

[0016] If it is judged that the photovoltaic power distribution network needs to be optimized and the start flag value is 0, the current time is marked as the optimization starting point, and the start flag value is set to 1, the period when the start flag value is continuously 1 is marked as the optimization period.

[0017] The construction method of the harmonic precursor identification model is as follows:

[0018] The constructed harmonic precursor identification model is a CNN-GRU hybrid deep learning network architecture for the harmonic precursor identification model, including an input layer, a convolutional neural network part, a gated recurrent unit part, a full connection layer and an output layer.

[0019] The input layer receives the precursor window matrix. The convolutional neural network part includes a convolution layer and a pooling layer. The unit structure of the gated recurrent unit part consists of two gate structures: an update gate and a reset gate. The fully connected layer uses the ReLU activation function to output a feature vector. The output layer uses the Softmax activation function to convert the feature vector output by the fully connected layer into the probability distribution of each harmonic generation factor and output the probability distribution vector of each harmonic generation factor.

[0020] The training data set is obtained as follows:

[0021] Acquire the harmonic information sequence and perform normalization processing to obtain a normalized harmonic information sequence;

[0022] Extract each optimized moment within the historical acquisition period and set a buffer window. For the buffer window with the optimized moment as the end point, take the starting point of the buffer window as the end point in the normalized harmonic information sequence and intercept a fixed-length data sequence forward, which is defined as the precursor window sequence. Each precursor window sequence is organized into a matrix to obtain a precursor window matrix.

[0023] For each precursor window matrix, the corresponding harmonic generation factors are marked for the precursor window matrix, and a training data set is obtained;

[0024] Wherein, the method of determining the harmonic generation factor is:

[0025] If the current time is not in the optimized period, the harmonic information sequence is obtained and normalized to obtain a normalized harmonic information sequence. The current window sequence is intercepted on the normalized harmonic information sequence, organized into a current window matrix and input into the harmonic precursor identification model for identification to obtain the probability distribution vector of each harmonic generation factor. Harmonic generation factors include light intensity fluctuation, insufficient short-circuit capacity ratio, and low load power factor.

[0026] When the maximum probability value in the probability distribution vector of each harmonic generation factor is greater than a preset probability threshold, the optimization period is triggered, the end point of the buffer window starting from the current moment is marked as the optimization moment, and the start flag value is set to 1. The harmonic generation factor corresponding to the maximum probability value is used as the harmonic generation factor at the optimization moment;

[0027] The harmonic optimization compensation mechanism includes:

[0028] If the harmonic generation factor at the current optimization moment is light intensity fluctuation, the adaptive active power filter is activated to detect the harmonic current in the grid in real time, generate a compensation current of equal magnitude and opposite direction, and inject it into the photovoltaic distribution network to offset the harmonic current;

[0029] If the harmonic generation factor at the current optimization moment is insufficient short-circuit capacity ratio, the static VAR compensator is put into operation to adjust its own reactive output, change the reactive power distribution of the power grid, and enhance the system's ability to resist harmonic disturbances;

[0030] If the harmonic generation factor at the current optimization moment is due to a low load power factor, the hybrid filter bank is activated to extract the harmonic current, generating an equal and opposite offset current, which is injected into the photovoltaic distribution network to achieve optimal compensation for the harmonics.

[0031] The method for determining whether to stop the optimization compensation is as follows:

[0032] Obtain the stability mark value of the current optimization period. If it is greater than the preset stability value, determine that the total voltage distortion rate of the photovoltaic distribution network continues to meet the standard. Obtain the harmonic generation factor corresponding to the current moment and determine whether the harmonic generation factor has disappeared.

[0033] If the total voltage distortion rate continues to meet the standard and the harmonic generation factors disappear, the start flag value will be reset to 0, and it is determined that the optimization compensation can be stopped;

[0034] The stability flag value is obtained as follows:

[0035] At the optimization starting point of the optimization period, the stability flag value of the current optimization period is set and its initial value is assigned to 0. During the optimization period, the total voltage distortion rate collected is the compensated total voltage distortion rate. If the total voltage distortion rate is less than or equal to the total voltage distortion rate threshold, the stability flag value performs a self-increment operation. Otherwise, the stability flag value is reset to 0.

[0036] A harmonic optimization system for a photovoltaic distribution network includes the following modules:

[0037] Acquisition and Analysis Module: This module collects harmonic information from the photovoltaic distribution network in real time, integrates it to generate a harmonic information sequence, and determines whether harmonic optimization is required. If so, it marks the optimization period and analyzes the factors that cause harmonics.

[0038] Model training module: Process and intercept the harmonic information sequence to obtain a training data set including a precursor window matrix, and construct and use the training data set to train a harmonic precursor recognition model;

[0039] Precursor recognition module: If the current time is not in the optimization period, the trained harmonic precursor recognition model is used to identify whether the optimization period is triggered. If triggered, the harmonic generation factor is determined based on the recognition result;

[0040] Optimization compensation module: During the optimization period, the harmonic optimization compensation mechanism is selected based on different harmonic generation factors to optimize the compensation of the photovoltaic distribution network. It is determined in real time whether the optimization compensation can be stopped. If so, the current optimization period is ended.

[0041] The beneficial effects of the present application are as follows:

[0042] 1、The present application monitors power quality in multiple dimensions, takes the total voltage distortion rate as the core, accurately locates harmonic problems combined with scene data, comprehensively traces the harmonic causes of the photovoltaic distribution network, extracts harmonic precursor characteristics using buffer window and precursor window sequences, combines a deep learning model to predict risks in advance, improves identification accuracy, uses special compensation equipment for different causes, dynamically adjusts parameters, realizes accurate governance, effectively improves compensation efficiency and accuracy, and ensures stable operation of the photovoltaic distribution network.

[0043] 2、The present application determines whether to stop optimization based on the dual conditions of voltage distortion rate continuously meeting the standard and harmonic causes disappearing, avoids the lag or false stop of the traditional single time threshold termination method, quantizes the voltage quality stability degree by adding and zeroing the flag value, stops compensation operation in time, reduces equipment loss and energy consumption, improves system operation economy, and balances the prediction lead time and accuracy, providing an efficient, reliable and economical solution for harmonic optimization of the photovoltaic distribution network. BRIEF DESCRIPTION OF DRAWINGS

[0044] The present application will be further described below with reference to the accompanying drawings.

[0045] Figure 1 is a step flow chart of a harmonic optimization method of a photovoltaic distribution network according to an embodiment of the present application;

[0046] Figure 2 is a module architecture diagram of a harmonic optimization system of a photovoltaic distribution network according to an embodiment of the present application. DETAILED DESCRIPTION

[0047] In order to make the technical means, creative features, purposes and effects achieved by the present application easy to understand, the present application will be further described below with reference to the specific embodiments.

[0048] Embodiment 1:

[0049] Please refer to Figure 1 The harmonic optimization method of a photovoltaic distribution network according to an embodiment of the present application includes the following steps:

[0050] S1: Real-time acquisition of harmonic related information of the photovoltaic distribution network, integration to generate a harmonic information sequence, and judgment of whether harmonic optimization is needed, if needed, marking the optimization period and analyzing the harmonic generating factors;

[0051] Real-time acquisition of harmonic related information of the photovoltaic distribution network, the harmonic related information including the total voltage distortion rate of the photovoltaic distribution network and the corresponding scene data, the scene data including the light intensity, the short circuit capacity ratio and the load power factor;

[0052] wherein the total voltage distortion rate is measured and calculated by the power quality monitor installed in the photovoltaic power distribution network, and the total voltage distortion rate The calculation formula is as follows:

[0053] ;

[0054] In the calculation formula, t represents the current time, represents the effective value of the fundamental voltage, represents the effective value of the h-th harmonic voltage, the collected voltage signal is decomposed into the fundamental and harmonic signals by the fast Fourier transform (FFT) algorithm in the power quality monitor, and then the effective values of the fundamental and harmonic voltages are calculated, h = 2, 3, …, H, and H represents the highest monitorable harmonic order;

[0055] It should be noted that the total voltage distortion rate is one of the important standards for measuring power quality. If the total voltage distortion rate is too high, it may have an adverse effect on the normal operation of the electrical equipment, and therefore the total voltage distortion rate is the core basis for determining whether harmonic optimization is needed;

[0056] wherein the light intensity S is measured by installing a photosynthetic radiation sensor at the photovoltaic module array of the photovoltaic power station, and the short-circuit capacity ratio is obtained and calculated through the technical data of the photovoltaic power distribution network, and the load power factor is measured and calculated in real time by installing a power measuring device on the load side of the photovoltaic power distribution network;

[0057] A historical collection period ending at the current time is set. At each time point in the historical collection period, the collected harmonic-related information of the photovoltaic power distribution network is integrated into a harmonic-related information vector, and the harmonic information sequence is obtained according to the time sequence;

[0058] Based on the real-time collected harmonic-related information of the photovoltaic power distribution network, it is determined whether harmonic optimization is needed at the current time. If so, the harmonic generating factor is determined;

[0059] Specifically, a start flag value is set and its initial value is set to 0. According to the relevant national standards and the specific operation requirements of the photovoltaic power distribution network, the total voltage distortion rate threshold is set. The total voltage distortion rate at the current time is compared with the total voltage distortion rate threshold. If the total voltage distortion rate at the current time is greater than the total voltage distortion rate threshold, it is determined that the photovoltaic power distribution network needs to be optimized;

[0060] If it is determined that the photovoltaic power distribution network needs to be optimized and the start flag value is 0, the current time is marked as the optimization starting point, and the start flag value is set to 1. The period during which the start flag value is continuously 1 is marked as the optimization period;

[0061] If the current time is the optimization starting point, the specific factors of harmonic generation are determined based on the scene data analysis corresponding to the current time, an analysis window is intercepted with the current time as the end point in the historical collection period, the difference value of the light intensity between the end point and the starting point of the analysis window is obtained, which is marked as the light intensity fluctuation amplitude at the current time, if the light intensity fluctuation amplitude at the current time is greater than the preset fluctuation amplitude threshold, it is judged that the harmonic generation factor that needs to be optimized is the light intensity fluctuation;

[0062] According to the actual situation of the photovoltaic power distribution network, the short-circuit capacity ratio threshold is set, if the short-circuit capacity ratio of the photovoltaic power distribution network at the current time is less than the set short-circuit capacity ratio threshold, it indicates that the support ability of the photovoltaic power distribution network to the photovoltaic power station is weak, it is judged that the harmonic generation factor that needs to be optimized is the insufficient short-circuit capacity ratio, similarly, the load power factor standard value is set, if the load power factor of the photovoltaic power distribution network at the current time is less than the set load power factor standard value, it indicates that the proportion of nonlinear load in the load is high, it is judged that the harmonic generation factor that needs to be optimized is the too low load power factor;

[0063] It should be noted that the role of this step is to monitor the power quality in multiple dimensions, taking the total voltage distortion rate as the core index, combining the scene data to realize accurate positioning of the harmonic problem, analyzing the light fluctuation amplitude, short-circuit capacity ratio threshold and other parameters based on the historical data window, improving the accuracy of harmonic generation factor judgment, avoiding misjudgment, comprehensively covering the unique harmonic inducement of the photovoltaic power distribution network, and realizing multi-dimensional tracing of the harmonic factor;

[0064] S2: process and intercept the harmonic information sequence to obtain a training data set including a precursor window matrix, construct a harmonic precursor identification model and train it using the training data set, if the current time is not in the optimization period, use the trained harmonic precursor identification model for identification, determine whether the optimization period is triggered, if it is triggered, determine the harmonic generation factor according to the identification result;

[0065] Obtain the harmonic information sequence in the historical collection period, normalize each component in the harmonic related information vector for the harmonic information sequence, and arrange to obtain the normalized harmonic information sequence;

[0066] Extract each optimization time in the historical collection period, set a buffer window, for the buffer window with the optimization time as the end point, intercept a data sequence of fixed length in the normalized harmonic information sequence with the starting point of the buffer window as the end point, define as a precursor window sequence, arrange each precursor window sequence into a matrix with dimension Lx4, marked as the precursor window matrix, wherein L represents the length of the precursor window sequence, i.e. the number of times in the precursor window sequence;

[0067] It should be noted that the role of the buffer window is to timely optimize the response to the harmonic to be generated after identifying the precursor characteristics generated by the harmonic, to prevent the photovoltaic power distribution network from being damaged due to the delayed response, and to prevent the preventive optimization from deviating too much and affecting the photovoltaic power distribution network. The buffer window length setting cannot lose the precursor characteristics too much. The buffer window length of all optimization moments is the same;

[0068] For each precursor window matrix, the harmonic generation factor corresponding to the optimization moment is taken as the harmonic generation factor of the precursor window matrix, and the corresponding harmonic generation factor is labeled for the precursor window matrix. All precursor window matrices in the historical collection period are labeled, and a training data set is obtained by sorting;

[0069] The harmonic precursor identification model constructed based on the CNN-GRU hybrid deep learning network is trained using the training data set;

[0070] Specifically, a CNN-GRU hybrid deep learning network architecture is built for the harmonic precursor identification model, including an input layer, a convolutional neural network (CNN) part, a gated recurrent unit (GRU) part, a fully connected layer, and an output layer.

[0071] The input layer receives the precursor window matrix, the convolutional neural network part includes a convolutional layer and a pooling layer, the convolutional layer extracts features from the input precursor window matrix to generate a feature map, and the pooling layer down-samples the feature map output by the convolutional layer to reduce the data dimension and reduce the model calculation amount. The unit structure of the GRU part consists of two gate structures, update gate and reset gate, which capture the change trend and dependency relationship of the precursor window matrix in time sequence. The fully connected layer uses a ReLU activation function to further extract and integrate key features related to harmonic generation factors, output a feature vector, and the output layer uses a Softmax activation function to convert the feature vector output by the fully connected layer into a probability distribution of each harmonic generation factor. The output is a probability distribution vector of each harmonic generation factor, and the sum of each component of the probability distribution vector is 1. The harmonic generation factors include illumination intensity fluctuation, short-circuit capacity ratio deficiency, and low load power factor;

[0072] The training data set is divided into a training set and a validation set in a ratio of 7:3 by using stratified sampling, the hyperparameters of the CNN-GRU hybrid deep learning network are set, the training set is input into the CNN-GRU hybrid deep learning network in batches, the loss value is calculated using the cross-entropy loss function, the calculated loss value is propagated in the reverse direction along the network structure from the output layer to the input layer by using the back propagation algorithm, the parameter gradient is calculated, the Adam optimizer is used, the parameters are updated according to the calculated parameter gradient and the set learning rate, the performance of the harmonic precursor identification model is gradually optimized, the harmonic precursor identification model is evaluated using the validation set data, the identification accuracy is calculated, the early stopping mechanism is introduced to prevent overfitting, and when the identification accuracy of the harmonic precursor identification model on the validation set reaches a preset threshold or the identification accuracy no longer improves, the trained harmonic precursor identification model is obtained;

[0073] If the current time is not in the optimization period, a data sequence with a length of L is intercepted from the normalized harmonic information sequence with the current time as the end point, and is defined as a current window sequence, which slides with the change of the current time;

[0074] At the current time, the current window sequence is arranged into a matrix with a dimension of Lx4, which is marked as a current window matrix, and the current window matrix is input into the trained harmonic precursor identification model for identification to obtain a probability distribution vector of each harmonic generating factor, including illumination intensity fluctuation, short-circuit capacity ratio deficiency, and low load power factor;

[0075] The maximum probability value in the probability distribution vector of each harmonic generating factor is compared with a preset probability threshold value, if the maximum probability value is less than or equal to the preset probability threshold value, it is judged that the precursor feature of harmonic generation cannot be identified in the current window sequence, the probability of harmonic generation in the photovoltaic power distribution network within the buffer window with the current time as the starting point is low, and preventive optimization is not needed, and the optimization period is not triggered;

[0076] If the maximum probability value is greater than the preset probability threshold value, the optimization period is triggered, the end point of the buffer window with the current time as the starting point is marked as the optimization time, and the start flag value is set to 1, and the harmonic generating factor corresponding to the maximum probability value is taken as the harmonic generating factor at the optimization time;

[0077] It should be noted that the purpose of this step is to extract the time sequence feature before the harmonic occurs through the buffer window and the precursor window sequence, to predict the harmonic risk in advance, to shorten the response time, to combine the spatial feature extraction capability of the convolutional neural network and the time sequence dependence capture capability of the gated recurrent unit, to balance the prediction advance and accuracy without losing features, and to avoid excessive response or lag of preventive optimization;

[0078] S3: During the optimization period, a harmonic optimization compensation mechanism is selected based on different harmonic generation factors to optimize the compensation of the photovoltaic distribution network;

[0079] If the current moment is the optimization moment, the harmonic generation factors at the optimization moment are obtained, a harmonic optimization compensation mechanism is selected for the photovoltaic distribution network based on different harmonic generation factors, and the selected harmonic optimization compensation mechanism is used to optimize and compensate for the harmonics during the optimization period;

[0080] Specifically, if the harmonic generation factor at the current optimization moment is light intensity fluctuation, the adaptive active power filter (APF) is activated for harmonic compensation. During the optimization period, the adaptive active power filter detects the harmonic current in the power grid in real time, generates a compensation current of equal magnitude and opposite direction, and injects it into the photovoltaic distribution network, thereby offsetting the harmonic current.

[0081] Compensation current command of adaptive active power filter The calculation formula is:

[0082] ;

[0083] Among them, the current moment is the tth moment in the optimization period, and the optimization starting point is the first moment in the optimization period. Indicates the difference between the total voltage distortion rate at the current moment and the total voltage distortion rate threshold. 、 Represent the proportional coefficient and integral coefficient respectively, Indicates the time between the start of the optimization period and the current time The points, represents the time variable within the optimization period;

[0084] If the harmonic generation factor at the current optimization time is insufficient short-circuit capacity ratio, a static VAR compensator (SVC) is used to optimize harmonic compensation. The SVC adjusts its own reactive power output to change the reactive power distribution of the power grid, thereby increasing the short-circuit capacity ratio of the power grid and enhancing the system's ability to resist harmonic disturbances.

[0085] Admittance compensation of static VAR compensator The calculation formula is:

[0086] ;

[0087] ;

[0088] in, Indicates the reactive power to be compensated, which is the active power output by the photovoltaic inverter and target power factor The active power output by the photovoltaic inverter is calculated The target power factor is obtained in real time by a monitoring device of the photovoltaic inverter The parameters are set in advance according to the operation requirements of the photovoltaic distribution network and the characteristics of the equipment, The voltage at the point of common coupling is represented, and is measured by a power quality monitor installed at the point of common coupling;

[0089] If the harmonic generation factor at the current optimization moment is the low load power factor, the hybrid filter group (HBF) is activated for harmonic compensation, and the harmonic extraction operator of the hybrid filter group extracts the harmonic current The calculation formula is as follows:

[0090] ;

[0091] ;

[0092] Among them, The load current waveform is represented, and is obtained in real time by a power measurement device installed on the load side, and FFT represents fast Fourier transform, and f represents The frequency is converted from the time domain to the frequency domain by fast Fourier transform, The frequency domain filter function is represented, and when the frequency is the fundamental frequency, that is, The function value is 0, and the fundamental component is not extracted, when the frequency The function value is 1, and the corresponding harmonic component is extracted, and the harmonic current in the load current is separated out ;

[0093] The hybrid filter group generates a cancellation current with equal size and opposite direction according to the calculated harmonic current , and injects it into the photovoltaic distribution network to realize the optimal compensation of the harmonic;

[0094] The compensation effect is verified in real time, and the total voltage distortion rate after compensation is compared with the total voltage distortion rate before compensation, and the compensation efficiency is calculated to evaluate the optimal compensation effect, if the compensation efficiency reaches the efficiency threshold, and the total voltage distortion rate after compensation Is less than or equal to the total voltage distortion rate threshold, it is judged that the current optimization compensation is effective, otherwise the optimization compensation parameter readjustment mechanism is triggered, and the optimization compensation parameters are adjusted, including the proportional coefficient, the integral coefficient of the adaptive active power filter, the admittance compensation amount of the static var compensator, and the harmonic extraction operator parameters of the hybrid filter group, and the optimization compensation operation is performed again according to the adjusted optimization compensation parameters, until the current optimization compensation is effective;

[0095] It should be noted that the role of this step is to use special compensation equipment for different harmonic causes, improve compensation efficiency, dynamically adjust compensation parameters based on the difference between real-time detection voltage distortion rate and threshold value, realize adaptive optimization of compensation strategy, realize accurate management of harmonic problems, and improve compensation accuracy in dynamic scenarios;

[0096] S4: determining whether the optimization compensation can be stopped in the optimization period, and ending the current optimization period if so;

[0097] In the optimization compensation process in the optimization period, it is determined whether the optimization compensation can be stopped;

[0098] Specifically, in the optimization period, the real-time collected harmonic related information of the photovoltaic power distribution network is obtained, including the voltage total distortion rate of the photovoltaic power distribution network and corresponding scene data, and the scene data includes the light intensity, the short-circuit capacity ratio and the load power factor;

[0099] A stable flag value is set at the optimization starting point and its initial value is 0. In the optimization period, the collected voltage total distortion rate is the compensated voltage total distortion rate. The voltage total distortion rate is compared with the voltage total distortion rate threshold. If the voltage total distortion rate at the current time is less than or equal to the voltage total distortion rate threshold, the stable flag value is increased by one. Otherwise, the stable flag value is reset to 0.

[0100] If the stable flag value is greater than a preset stable value, it is determined that the voltage total distortion rate of the photovoltaic power distribution network is continuously up to standard. Based on the harmonic generating factor corresponding to the current optimization time, it is determined whether the harmonic generating factor disappears.

[0101] Specifically, if the harmonic generating factor is light intensity fluctuation, the light intensity fluctuation amplitude at the current time is calculated. If the light intensity fluctuation amplitude is less than or equal to a preset fluctuation amplitude threshold, it is determined that the light intensity fluctuation stops and the harmonic generating factor disappears.

[0102] If the harmonic generating factor is insufficient short-circuit capacity ratio, and the short-circuit capacity ratio of the photovoltaic power distribution network at the current time is greater than or equal to a set short-circuit capacity ratio critical value, it is determined that the short-circuit capacity ratio returns to normal and the harmonic generating factor disappears.

[0103] If the harmonic generating factor is too low load power factor, and the load power factor of the photovoltaic power distribution network at the current time is greater than or equal to a load power factor standard value, it is determined that the operation state of the load is improved and the harmonic generating factor disappears.

[0104] If the voltage total distortion rate is continuously up to standard and the harmonic generating factor disappears, the start flag value is reset to 0, indicating that the current harmonic optimization operation is completed, the current optimization period is ended, all ongoing optimization compensation operations are stopped, and the system returns to the normal monitoring state, avoiding unnecessary resource consumption and equipment operation loss.

[0105] It should be noted that this step ensures the reliability of optimization termination based on the dual conditions of continuous compliance with voltage distortion standards and the disappearance of harmonic inducements. It promptly stops compensation operations, reduces equipment loss and energy consumption, and improves the economic efficiency of system operation. A dynamic termination mechanism is designed to combine power quality indicators with the elimination status of harmonic inducements, avoiding the lag or false shutdown of the traditional single time threshold termination method. Through the logic design of self-increment and zeroing of the flag value, the stability of voltage quality is quantified, realizing intelligent judgment of optimization termination.

[0106] The technical solution of an embodiment of the present invention is: real-time collection of harmonic-related information of a photovoltaic distribution network, integration to generate a harmonic information sequence, and judgment on whether harmonic optimization is required; if necessary, marking the optimization period and analyzing harmonic generation factors; processing and intercepting the harmonic information sequence to obtain a training data set including a precursor window matrix; constructing a harmonic precursor recognition model and using the training data set for training; if the current moment is not in the optimization period, using the trained harmonic precursor recognition model for recognition, and judging whether the optimization period is triggered; if triggered, determining the harmonic generation factors according to the recognition results; within the optimization period, selecting a harmonic optimization compensation mechanism based on different harmonic generation factors, and optimizing and compensating the photovoltaic distribution network; judging whether the optimization compensation can be stopped within the optimization period; and if so, ending the current optimization period.

[0107] Example 2:

[0108] like Figure 2 As shown, a harmonic optimization system for a photovoltaic distribution network according to an embodiment of the present invention includes the following modules:

[0109] Acquisition and Analysis Module: This module collects harmonic information from the photovoltaic distribution network in real time, integrates it to generate a harmonic information sequence, and determines whether harmonic optimization is required. If so, it marks the optimization period and analyzes the factors that cause harmonics.

[0110] Model training module: Process and intercept the harmonic information sequence to obtain a training data set including a precursor window matrix, and construct and use the training data set to train a harmonic precursor recognition model;

[0111] Precursor recognition module: If the current time is not in the optimization period, the trained harmonic precursor recognition model is used to identify whether the optimization period is triggered. If triggered, the harmonic generation factor is determined based on the recognition result;

[0112] Optimization compensation module: During the optimization period, the harmonic optimization compensation mechanism is selected based on different harmonic generation factors to optimize the compensation of the photovoltaic distribution network. It is determined in real time whether the optimization compensation can be stopped. If so, the current optimization period is ended.

[0113] The above shows and describes the basic principles, main features and advantages of the present application. Those skilled in the art should understand that the present application is not limited to the above-mentioned embodiments, and the above-mentioned embodiments and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A harmonic optimization method for a photovoltaic distribution network, characterized by: The application relates to a method for optimizing and compensating harmonics of a photovoltaic power distribution network. Real-time acquisition of harmonic-related information of the photovoltaic power distribution network, integration to generate a harmonic information sequence, and judgment of whether harmonic optimization is needed, if needed, marking an optimization period and analyzing harmonic generation factors; The acquisition mode of the harmonic information sequence is as follows: The harmonic-related information includes a total voltage distortion rate of the photovoltaic power distribution network and corresponding scene data, and the scene data further includes an illumination intensity, a short-circuit capacity ratio and a load power factor; A historical acquisition period with the current time as the terminal point is set, at each time point in the historical acquisition period, the acquired harmonic-related information of the photovoltaic power distribution network is integrated into a harmonic-related information vector, and a harmonic information sequence is obtained according to the time sequence; The harmonic information sequence is processed and intercepted to obtain a training data set including a precursor window matrix, a harmonic precursor identification model is constructed and trained by using the training data set, if the current time is not in the optimization period, the trained harmonic precursor identification model is used for identification, whether the optimization period is triggered is judged, if triggered, the harmonic generation factors are determined according to the identification result; In the optimization period, different harmonic generation factors are selected to select a harmonic optimization compensation mechanism, the photovoltaic power distribution network is optimized and compensated, and whether the optimization and compensation can be stopped is judged in real time, if yes, the current optimization period is ended.

2. The method of claim 1, wherein: The marking mode of the optimization period is as follows: Harmonic-related information is acquired, the harmonic-related information includes a total voltage distortion rate of the photovoltaic power distribution network, a starting flag value is set and the initial value is 0, the total voltage distortion rate at the current time is compared with a preset total voltage distortion rate threshold value, if greater than the total voltage distortion rate threshold value, it is judged that the photovoltaic power distribution network needs harmonic optimization; If it is judged that the photovoltaic power distribution network needs optimization and the starting flag value is 0, the current time is marked as an optimization starting point, the starting flag value is set to 1, and the period with the starting flag value being 1 continuously is marked as an optimization period.

3. The method of claim 1, wherein: The construction mode of the harmonic precursor identification model is as follows: The constructed harmonic precursor identification model is a CNN-GRU hybrid deep learning network architecture for the harmonic precursor identification model, including an input layer, a convolutional neural network part, a gated recurrent unit part, a full connection layer and an output layer; The input layer receives the precursor window matrix, the convolutional neural network part includes a convolutional layer and a pooling layer, the unit structure of the gated recurrent unit part is composed of two gate structures of an update gate and a reset gate, the full connection layer adopts a ReLU activation function, an output feature vector is output, and the output layer adopts a Softmax activation function, the feature vector output by the full connection layer is converted into a probability distribution of each harmonic generation factor, and a probability distribution vector of each harmonic generation factor is output.

4. The method of claim 2, wherein: The acquisition mode of the training data set is as follows: The harmonic information sequence is acquired and normalized to obtain a normalized harmonic information sequence; Each optimization time point in the historical acquisition period is extracted, a buffer window is set, for the buffer window with the optimization time point as the terminal point, a data sequence with a fixed length is intercepted from the normalized harmonic information sequence with the starting point of the buffer window as the terminal point, defined as a precursor window sequence, each precursor window sequence is arranged into a matrix to obtain a precursor window matrix. For each precursor window matrix, the corresponding harmonic generation factor is labeled for the precursor window matrix, and a training data set is obtained.

5. The method of claim 3, wherein: The determination of the harmonic generation factor is as follows: If the current time is not in the optimization period, the harmonic information sequence is obtained and normalized to obtain the normalized harmonic information sequence, and the current window sequence is obtained by intercepting the normalized harmonic information sequence, which is arranged into a current window matrix and input into a harmonic precursor identification model for identification to obtain a probability distribution vector of each harmonic generation factor, including illumination intensity fluctuation, short-circuit capacity ratio deficiency and low load power factor; If the maximum probability value in the probability distribution vector of each harmonic generation factor is greater than a preset probability threshold, the optimization period is triggered, the end point of the buffer window starting from the current time is marked as the optimization time, the start flag value is set to 1, and the harmonic generation factor corresponding to the maximum probability value is taken as the harmonic generation factor of the optimization time.

6. The method of claim 5, wherein: The harmonic optimization compensation mechanism includes: If the harmonic generation factor of the current optimization time is illumination intensity fluctuation, an adaptive active power filter is enabled, the harmonic current in the power grid is detected in real time, a compensation current with the same size and opposite direction is generated, and the compensation current is injected into the photovoltaic distribution network to offset and compensate the harmonic current; If the harmonic generation factor of the current optimization time is short-circuit capacity ratio deficiency, a static var compensator is put into operation to change the reactive power distribution of the power grid by adjusting its own reactive power output, thereby enhancing the ability of the system to resist harmonic disturbance; If the harmonic generation factor of the current optimization time is low load power factor, a hybrid filter group is activated to extract the harmonic current, generate a compensation current with the same size and opposite direction, and inject the compensation current into the photovoltaic distribution network to realize the optimization compensation of the harmonic.

7. The method of claim 1, wherein: The judgment of whether the optimization compensation can be stopped is as follows: The stability flag value of the current optimization period is obtained, and if it is greater than a preset stability value, it is judged that the voltage total distortion rate of the photovoltaic distribution network is continuously up to standard, the harmonic generation factor corresponding to the current time is obtained, and it is judged whether the harmonic generation factor disappears. If the voltage total distortion rate is continuously up to standard and the harmonic generation factor disappears, the start flag value is reset to 0, and it is judged that the optimization compensation can be stopped.

8. The method of claim 7, wherein: The stability flag value is obtained as follows: At the optimization starting point of the optimization period, the stability flag value of the current optimization period is set and assigned an initial value of 0, the voltage total distortion rate obtained in the optimization period is the voltage total distortion rate after compensation, if the voltage total distortion rate is less than or equal to the voltage total distortion rate threshold, the stability flag value is increased by one, otherwise, the stability flag value is reset to 0.

9. A system for harmonic optimization of a photovoltaic power distribution network, the system being configured to implement the optimization method according to any one of claims 1 to 8, characterized in that: The following modules are included: The acquisition and analysis module: real-time acquisition of harmonic-related information of the photovoltaic distribution network, integration to generate a harmonic information sequence, and judgment of whether harmonic optimization is needed, if needed, marking the optimization period and analyzing the harmonic generation factor; The harmonic information sequence is obtained as follows: The harmonic-related information includes the voltage total distortion rate of the photovoltaic distribution network and the corresponding scene data, and the scene data includes the illumination intensity, the short-circuit capacity ratio and the load power factor. A historical collection time period ending at the current time is set, at each time in the historical collection time period, the collected harmonic related information of the photovoltaic power distribution network is integrated into a harmonic related information vector, and a harmonic information sequence is obtained according to time sequence integration; The model training module: processes and intercepts the harmonic information sequence to obtain a training data set including a precursor window matrix, constructs and trains the harmonic precursor identification model using the training data set; The precursor identification module: if the current time is not in the optimization period, the trained harmonic precursor identification model is used for identification to determine whether the optimization period is triggered, and if so, the harmonic generation factor is determined according to the identification result; The optimization compensation module: in the optimization period, different harmonic optimization compensation mechanisms are selected based on different harmonic generation factors to optimize and compensate the photovoltaic power distribution network, and it is determined in real time whether the optimization compensation can be stopped, and if so, the current optimization period is ended.

Images