A real-time inspection and intelligent diagnosis method and system for power grid operation mode with large-scale wind power access

Abstract

The present invention provides a method and system for real-time inspection and intelligent diagnosis of the operation mode of a power grid with a large number of wind power accesses, which relates to the field of data processing technology. The method includes: obtaining a power grid operation state sequence, which includes a voltage signal sequence; performing voltage prediction based on an improved GRU‑RNN according to the voltage signal sequence to obtain a voltage prediction signal; extracting a variety of feature graphs from the voltage prediction signal through wavelet transform; calculating disturbance parameters according to the feature graphs; detecting the operation quality of the power grid in real time according to the disturbance parameters; and performing intelligent diagnosis based on the improved GRU‑RNN when the operation quality of the power grid is low. The present invention can more accurately quantify the disturbance in the power grid, and then detect the operation quality of the power grid in real time according to the disturbance parameters, accurately reflecting the operation status of the power grid when a large number of wind power systems are connected, adapting to the intermittent and uncertain nature of the wind power system, and improving the detection accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a real-time inspection and intelligent diagnosis method and system for the operation mode of a power grid with large-scale access of wind power. Background Art

[0002] In order to ensure the safe, stable and efficient operation of the power grid, real-time detection of the operating status of the power grid is a key link. Currently, the SCADA (Supervisory Control And Data Acquisition) system is mainly used to monitor whether the power parameters such as voltage and power reach the alarm threshold to determine the operating status of the power grid. The SCADA system has functions such as data acquisition, monitoring, control, and alarm, and is an important tool for realizing power grid automation and remote monitoring.

[0003] However, with the development of a large number of clean energy sources such as wind power and photovoltaics, and with the connection of a large number of wind power systems to the power grid, due to the obvious intermittent and uncertain nature of wind power systems, it is difficult to adapt to the intermittent and uncertain nature of wind power systems by simply monitoring whether the power parameters have reached the alarm threshold through the SCADA system to determine the power grid operation status. This is prone to false alarms, has low detection accuracy, and is unable to provide status warnings, making it difficult to respond to changes in the power grid operation status in a timely manner. Summary of the invention

[0004] In order to solve the technical problems in the prior art that when a large number of wind power systems are connected to the power grid, due to the obvious intermittency and uncertainty of the wind power systems, only the SCADA system is used to monitor whether the power parameters reach the alarm threshold to determine the power grid operation status, which is difficult to adapt to the intermittency and uncertainty of the wind power systems, prone to false alarms, low detection accuracy, and unable to perform status warnings, making it difficult to respond to changes in the power grid operation status in a timely manner, the present invention provides a real-time inspection and intelligent diagnosis method and system for the power grid operation mode.

[0005] The technical solution provided by the embodiment of the present invention is as follows:

[0006] First aspect:

[0007] An embodiment of the present invention provides a method for real-time inspection and intelligent diagnosis of power grid operation mode, comprising:

[0008] S1: Acquire a power grid operation state sequence, wherein the power grid operation state sequence includes a voltage signal sequence;

[0009] S2: According to the voltage signal sequence, based on the improved GRU-RNN, voltage prediction is performed to obtain a voltage prediction signal;

[0010] S3: extracting a plurality of feature graphs from the voltage prediction signal through wavelet transform;

[0011] S4: Calculate disturbance parameters according to the characteristic graph;

[0012] S5: Detecting the power grid operation quality in real time according to the disturbance parameters;

[0013] S6: When the grid operation quality is poor, intelligent diagnosis is performed based on the improved GRU-RNN.

[0014] Second aspect:

[0015] An embodiment of the present invention provides a real-time inspection and intelligent diagnosis system for power grid operation mode, comprising:

[0016] processor;

[0017] A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the real-time inspection and intelligent diagnosis method of the power grid operation mode as described in the first aspect is implemented.

[0018] The third aspect:

[0019] An embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the real-time inspection and intelligent diagnosis method of the power grid operation mode as described in the first aspect is implemented.

[0020] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0021] (1) In the present invention, wavelet transform is used to extract multiple feature graphs from the voltage prediction signal, which can more comprehensively reflect the changes in the power grid state. The disturbance parameters are calculated based on the feature graphs. The disturbance parameters may include frequency disturbance, voltage fluctuation, harmonic distortion, etc., which can more accurately quantify the disturbance in the power grid. Then, the power grid operation quality can be detected in real time based on the disturbance parameters. It can more accurately reflect the power grid operation status when a large number of wind power systems are connected, adapt to the intermittent and uncertain nature of the wind power system, reduce false alarms, and improve detection accuracy.

[0022] (2) In the present invention, the voltage signal is predicted by the improved GRU-RNN model, which can more accurately predict the voltage change trend in the future period, carry out status warning, and respond to the changes in the grid operation status in a timely manner, which helps to make corresponding adjustments and measures in advance before changes occur in the grid.

[0023] (3) In the present invention, when the grid operation quality is poor, the improved GRU-RNN is used for intelligent diagnosis, which can help determine the possible problems or faults in the grid, help to quickly locate the root cause of the problem, and guide the operation and maintenance personnel to perform effective maintenance and processing, which can effectively monitor and manage the grid operation status and improve the reliability, safety and economy of the grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0025] Figure 1 A flowchart of a method for real-time inspection and intelligent diagnosis of power grid operation mode provided by an embodiment of the present invention;

[0026] Figure 2 A schematic diagram of the structure of a real-time inspection and intelligent diagnosis system for power grid operation mode provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0028] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either of the two.

[0029] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same. "of", "corresponding, relevant" and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same.

[0030] In the embodiments of the present invention, sometimes a subscript such as W1 may be mistakenly written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are consistent.

[0031] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0032] Reference Manual Attached Figure 1 , showing a flow chart of a method for real-time inspection and intelligent diagnosis of power grid operation mode provided by an embodiment of the present invention.

[0033] The embodiment of the present invention provides a method for real-time inspection and intelligent diagnosis of power grid operation mode, which can be implemented by a real-time inspection and intelligent diagnosis device for power grid operation mode, and the real-time inspection and intelligent diagnosis device for power grid operation mode can be a terminal or a server. The processing flow of the method for real-time inspection and intelligent diagnosis of power grid operation mode can include the following steps:

[0034] S1: Obtain the grid operation status sequence.

[0035] The grid operation state sequence includes a voltage signal sequence, and further, may include a current signal sequence, a power signal sequence, and the like.

[0036] S2: According to the voltage signal sequence, voltage prediction is performed based on the improved GRU-RNN to obtain a voltage prediction signal.

[0037] Among them, GRU-RNN (Gated Recurrent Unit-Recurrent Neural Network) is a variant of recurrent neural network (RNN) designed to solve the long-term dependency problem in traditional RNN. It controls the flow of information by introducing a gating mechanism, so that it performs better in capturing long-term dependencies. Traditional RNNs have problems with gradient vanishing and gradient exploding, especially when processing long sequence data. This is because during backpropagation, the gradient needs to be propagated through multiple time steps, causing the gradient to gradually decay or explode. To solve this problem, GRU introduces an update gate and a reset gate to better control the transmission and preservation of information.

[0038] In a possible implementation, S2 specifically includes sub-steps S201 to S206:

[0039] S201: In the input layer of the GRU-RNN, a voltage signal sequence is obtained.

[0040] It should be noted that in the traditional GRU-RNN update gate, part of the original signal is retained when updating the state, which leads to an abnormally large number of model parameters and a complicated calculation process, affecting the model calculation efficiency and being unfavorable for rapid response when a large number of wind power systems are connected. The specific calculation formula of the prior art is:

[0041] z t =σ(W uz u t +W hz h t-1 +b z )

[0042] Among them, z t represents the output vector of the update gate at time t, σ() represents the activation function, and W uz represents the weight matrix between the input layer and the update gate, u t Represents the voltage signal at time t, W hz represents the self-connection weight matrix of the update gate between time t and time t-1, h t-1 represents the hidden state vector at time t-1, b z Represents the bias term for the update gate.

[0043] Similarly, in the reset gate of the traditional GRU-RNN, part of the original signal will be retained, which leads to an abnormally large number of model parameters and a complicated calculation process, affecting the calculation efficiency of the model and being unfavorable for rapid response when a large number of wind power systems are connected. The specific calculation formula of the prior art is:

[0044] r t =σ(W ur u t +W hr h t-1 +b r )

[0045] Among them, r t represents the output vector of the reset gate at time t, W ur represents the weight matrix between the input layer and the reset gate, W hr represents the self-connection weight matrix between the reset gate at time t and time t-1, b r Represents the bias term for the reset gate.

[0046] To this end, the present invention proposes an improved GRU-RNN, which only considers the previous hidden state and bias when updating the gate and resetting the gate to update the state, and discards part of the original signal. This is because after a large number of wind power equipment are connected, due to the intermittent nature and uncertainty of the wind power equipment, the original signal is not very meaningful for the state update. However, retaining the original signal will greatly increase the difficulty of model training and calculation, which is not conducive to rapid response when a large number of wind power systems are connected. When discarding part of the original signal for state update, on the one hand, the difficulty of model training can be reduced, and on the other hand, the amount of calculation can be reduced, and the power grid operation status when a large number of wind power systems are connected can be reflected in a timely and accurate manner. It can adapt to the intermittent nature and uncertainty of the wind power system, reduce false alarms, and improve detection accuracy.

[0047] S202: In the update gate of the GRU-RNN, the output vector of the update gate is calculated according to the following formula:

[0048] z t =σ(W hz h t-1 +b z )

[0049] Among them, z t represents the output vector of the update gate at time t, σ() represents the activation function, and W hz represents the self-connection weight matrix of the update gate between time t and time t-1, h t-1 represents the hidden state vector at time t-1, b z Represents the bias term for the update gate.

[0050] S203: In the reset gate of the GRU-RNN, the output vector of the reset gate is calculated according to the following formula:

[0051] r t =σ(W hr h t-1 +b r )

[0052] Among them, r t represents the output vector of the reset gate at time t, W hr represents the self-connection weight matrix between the reset gate at time t and time t-1, b r Represents the bias term for the reset gate.

[0053] S204: In the candidate layer of the GRU-RNN, the output vector of the candidate layer is calculated according to the following formula:

[0054] c t =σ(W uc u t +W hc (r t ⊙h t-1 )+b c )

[0055] Among them, c t represents the output vector of the candidate layer at time t, W uc represents the weight matrix between the input layer and the candidate layer, u t Represents the voltage signal at time t, W hc represents the self-connection weight matrix between the candidate layer at time t and time t-1, b c represents the bias term of the candidate layer, and ⊙ represents the element-wise product operation.

[0056] S205: In the hidden layer of GRU-RNN, the hidden state vector at the moment is calculated according to the following formula:

[0057] h t =(1-z t )⊙h t-1 +z t ⊙c t

[0058] Among them, h t Represents the hidden state vector at time t.

[0059] S206: In the output layer of the GRU-RNN, the voltage prediction signal is calculated according to the following formula:

[0060]

[0061] in, Represents the voltage prediction signal at time t, W yh represents the weight matrix between the hidden layer and the output layer, b y Represents the bias term of the output layer.

[0062] In the present invention, the GRU-RNN model can process time series data and memorize information within a certain time range during the training process. In this way, when a new voltage signal sequence is input, the model can predict the voltage condition at the next moment in real time, which is conducive to timely monitoring and adjustment of the power grid operation status. In addition, the GRU-RNN model can capture the complex time series characteristics in the voltage signal sequence, including long-term dependencies and nonlinear relationships. Compared with traditional linear models or simple neural network models, GRU-RNN can better fit the changing law of voltage signals and improve the accuracy of prediction.

[0063] In a possible implementation, the training method of the GRU-RNN specifically includes:

[0064] Construct the loss function:

[0065]

[0066] Among them, J represents the loss function, represents the voltage prediction signal at time t, u t Represents the voltage signal at time t, t=1, 2, ..., T, T represents the total duration.

[0067] With the goal of minimizing the function value of the loss function, the model parameters of the GRU-RNN are iterated:

[0068]

[0069] Among them, θi+1 represents the model parameters at the i+1th iteration, θ i represents the model parameters at the i-th iteration, η represents the learning rate, Represents the loss function J on the model parameters θ i The gradient of , β represents the inertia weight factor.

[0070] Among them, those skilled in the art can set the learning rate η and the inertia weight factor β according to actual conditions, and the present invention does not limit this.

[0071] In the present invention, the model parameters are updated by gradient descent, so that the loss function is gradually reduced and the model gradually converges to the optimal solution. This method can accelerate the training process of the model and enable the model to achieve ideal performance more quickly. By minimizing the loss function, the prediction results of the model can be made closer to the actual observed values, thereby optimizing the parameters of the model and improving the accuracy and reliability of the prediction.

[0072] Optionally, the learning rate is innovatively set to an adaptive learning rate, and the inertia weight factor is innovatively set to an adaptive inertia weight factor.

[0073] Then the transformation is to iterate the model parameters of GRU-RNN with the goal of minimizing the function value of the loss function:

[0074]

[0075] Among them, θ i+1 represents the model parameters at the i+1th iteration, θ i represents the model parameters at the i-th iteration, η i represents the adaptive learning rate at the i-th iteration, Represents the loss function J on the model parameters θ i The gradient of i Represents the adaptive inertia weight factor at the i-th iteration.

[0076] Optionally, the adaptive learning rate is calculated as:

[0077]

[0078] Among them, η i represents the adaptive learning rate at the i-th iteration, η0 represents the initial learning rate, g k represents the gradient at the kth iteration, k = 1, 2, …, i, and μ represents the hyperparameter.

[0079] In the present invention, the adaptive learning rate can adjust the learning rate according to the current gradient. If the gradient is large, the learning rate is large, which is conducive to rapid convergence; conversely, if the gradient is small, the learning rate is small to avoid oscillation or instability caused by learning too fast. In this way, the optimal solution of the loss function can be approached faster.

[0080] in, θ i represents the model parameters at the i-th iteration.

[0081] Optionally, the adaptive inertia weight factor is calculated as:

[0082]

[0083] Among them, β i represents the adaptive inertia weight factor at the i-th iteration, e represents the natural constant, γ represents the initial coefficient, ||g i || represents the gradient g i Perform the two-norm calculation.

[0084] In the present invention, the adaptive inertia weight factor can dynamically adjust the weight according to the size of the current gradient. If the gradient is large, the inertia weight is small, which helps to make faster adjustments in the direction of steep gradients; if the gradient is small, the inertia weight is large, which can maintain a certain stability in the direction of gentle gradients.

[0085] In the present invention, by adaptively adjusting the learning rate and the inertia weight factor, it is possible to better adapt to different data distributions and model complexities, thereby improving the generalization ability and adaptability of the model.

[0086] S3: Through wavelet transform, various feature maps are extracted from the voltage prediction signal.

[0087] Among them, wavelet transform is a time-frequency analysis method used to analyze signals in two dimensions: time and frequency. Compared with Fourier transform, wavelet transform can provide better localization characteristics in time domain and frequency domain. Through wavelet transform, a variety of feature graphs can be extracted from the signal, including amplitude curve graph, absolute value and curve graph, power frequency curve graph, etc. The basic idea of ​​wavelet transform is to decompose the signal using wavelet basis functions of different scales and positions. These wavelet basis functions can be compressed and expanded in the time domain, so they can adapt well to the local characteristics of the signal.

[0088] Among them, the characteristic graphs include amplitude curve graphs, absolute value and curve graphs, power frequency curve graphs, etc.

[0089] Among them, the amplitude curve graph shows how the amplitude of the signal changes over time.

[0090] Among them, the absolute value and curve graph are curves obtained after performing absolute value processing on the signal.

[0091] Among them, the power frequency curve shows how the frequency of the signal changes over time.

[0092] In a possible implementation, S3 specifically includes sub-steps S301 and S302:

[0093] S301: Perform wavelet transform on the voltage prediction signal, decompose the voltage prediction signal into different frequency components, and give the temporal distribution of these components.

[0094] S302: Determine an amplitude curve graph, an absolute value and curve graph, and a power frequency curve graph according to the coefficients after wavelet decomposition.

[0095] Specifically, for the coefficients after wavelet decomposition, the modulus (or absolute value) of each layer of coefficients can be calculated, which represents the amplitude of the corresponding frequency component. The changes of these amplitudes over time are plotted into a graph, that is, an amplitude graph is obtained. For the original voltage prediction signal or a certain layer of coefficients after wavelet transformation, its absolute value is directly calculated. Then, the sum of the absolute value over time is plotted into a graph to obtain an absolute value sum graph. For the coefficients after wavelet decomposition, the energy or power of each layer of coefficients can be calculated (usually the sum of the squares of the coefficients). These power values ​​are plotted into a graph according to the corresponding frequency, that is, a power frequency graph is obtained.

[0096] As for how to utilize wavelet transform to extract a variety of feature graphs from the voltage prediction signal, reference may be made to many other prior arts. This aspect is already a very mature prior art, and the present invention will not elaborate on it here.

[0097] In the present invention, different characteristic graphs obtained by wavelet transform can more comprehensively analyze the frequency, amplitude and other characteristics of the signal, provide richer information, and can more comprehensively and accurately analyze the characteristics of the signal in the power system, which is helpful for real-time monitoring of the operating status and quality of the power grid.

[0098] S4: Calculate the perturbation parameters according to the feature map.

[0099] Optionally, the disturbance parameters specifically include: voltage harmonic distortion, amplitude disturbance, amplitude disturbance time, absolute value and disturbance degree, absolute value and disturbance time, power frequency disturbance degree and power frequency disturbance time.

[0100] In a possible implementation, S4 specifically includes sub-steps S401 to S404:

[0101] S401: Determine the voltage harmonic distortion degree according to the fundamental wave component and the harmonic component obtained by wavelet transformation.

[0102] Optionally, the voltage harmonic distortion degree is specifically:

[0103]

[0104] Where THD represents voltage harmonic distortion, V1 represents fundamental voltage value, V2, V3, ..., V n Indicates the value of each harmonic voltage.

[0105] In the present invention, by calculating the voltage harmonic distortion, the influence of the fundamental wave and each harmonic component on the total voltage can be comprehensively considered, and the voltage distortion can be evaluated more comprehensively, rather than just focusing on the fundamental wave voltage.

[0106] S402: Determine the amplitude disturbance degree and the amplitude disturbance time according to the amplitude curve diagram.

[0107] Optionally, the amplitude disturbance degree is determined in the following manner:

[0108] δ A =A max -A min

[0109] Among them, δ A A represents the amplitude disturbance, A max Indicates the maximum amplitude in the amplitude curve, A min Indicates the minimum amplitude value in the amplitude curve graph.

[0110] It should be noted that by calculating the amplitude disturbance degree, the fluctuation in the amplitude curve can be quantified into a value, which helps to more intuitively understand the range of amplitude variation. Large fluctuations in amplitude may indicate stability problems in the power grid or equipment. Quantitative calculation of the amplitude disturbance degree helps to detect these problems in a timely manner and conduct further investigation and treatment.

[0111] The amplitude disturbance time is determined as follows:

[0112] T A =∑t A

[0113]

[0114] Among them, T A Indicates the amplitude disturbance time, specifically the sum of the times when the absolute value of the slope in the amplitude curve is greater than the preset amplitude slope, t A Indicates the moment when the absolute value of the slope in the amplitude curve is greater than the preset amplitude slope. A represents the amplitude. Indicates the amplitude curve t A The slope at time k A Indicates the preset amplitude slope.

[0115] It should be noted that by calculating the amplitude disturbance time, the duration of the slope of the amplitude curve greater than the preset amplitude slope can be quantified, which helps to understand the persistence of the amplitude change. The amplitude disturbance time describes the sum of the moments when the slope of the amplitude curve is greater than the preset amplitude slope, which reflects the characteristics of the amplitude change. A disturbance with a longer duration may mean that there are persistent problems or abnormal conditions in the system.

[0116] S403: Determine the absolute value and the disturbance degree and the absolute value and the disturbance time according to the absolute value and the curve graph. Optionally, the absolute value and the disturbance degree are determined in the following manner:

[0117] δ S =S max -S min

[0118] Among them, δ S Indicates the absolute value and disturbance degree, S max Indicates the absolute value and the absolute value and maximum value in the curve graph, S min Indicates the absolute value and the minimum value in the absolute value and curve graph.

[0119] It should be noted that the absolute value and disturbance degree provide a measure of the absolute value and amplitude change in the curve. By calculating the difference between the maximum and minimum values, the overall change of the curve can be understood. Large absolute values ​​and changes may be an indication of system abnormalities or failures. By monitoring the absolute value and disturbance degree, abnormal conditions in the system can be discovered in time and early warning can be given, which helps reduce potential problems.

[0120] The absolute value and disturbance time are determined as follows:

[0121] T S =∑t S

[0122]

[0123] Among them, T S Represents the absolute value and disturbance time, specifically the sum of the absolute value and the time when the absolute value of the slope in the curve is greater than the preset absolute value and slope, t S Indicates the moment when the absolute value of the slope in the absolute value and curve is greater than the preset absolute value and slope, S represents the absolute value and, Indicates the absolute value and t in the curve S The slope at time k S Indicates the preset absolute value and slope.

[0124] It should be noted that the absolute value and disturbance time can detect the rate of change of the absolute value and by considering the slope of the absolute value and disturbance time curve. When the absolute value of the slope is greater than the preset slope threshold, it means that the change in the absolute value and disturbance time is more drastic and needs attention. Large absolute value and disturbance time may indicate that there are long-lasting changes in the absolute value and disturbance time in the system. This fine-grained anomaly detection helps operation and maintenance personnel locate problems more accurately and take timely measures to adjust or handle them.

[0125] S404: Determine the power frequency disturbance degree and the power frequency disturbance time according to the power frequency curve.

[0126] Optionally, the power frequency disturbance degree is determined in the following manner:

[0127] δ f =f max -f min

[0128] Among them, δ f represents the power frequency disturbance, f max Indicates the maximum power frequency in the power frequency curve, f min Indicates the minimum power frequency in the power frequency curve.

[0129] It should be noted that the power frequency disturbance provides a quantitative indicator of power frequency changes by calculating the difference between the maximum and minimum values ​​in the power frequency curve. This helps to understand the magnitude of power frequency changes in the system. A larger power frequency disturbance may indicate an abnormal situation in the system, such as load fluctuations or the influence of other external factors. This provides a simple but effective indicator to detect abnormal conditions in the system.

[0130] The power frequency disturbance time is determined as follows:

[0131] T f =∑t f

[0132]

[0133] Among them, T f Indicates the power frequency disturbance time, specifically the sum of the times when the absolute value of the slope in the power frequency curve is greater than the preset power frequency slope, t f Indicates the moment when the absolute value of the slope in the power-frequency curve is greater than the preset power-frequency slope, f represents the power frequency, Indicates the power frequency curve t f The slope at time k f Indicates the preset power frequency slope.

[0134] It should be noted that the power frequency disturbance time provides a quantitative indicator of the duration of the frequency disturbance by calculating the sum of the moments when the absolute value of the slope in the power frequency curve is greater than the preset power frequency slope. This helps to understand the duration and duration of the frequency disturbance. The continuous power frequency disturbance time may indicate that there are continuous abnormal conditions in the system, such as unstable frequency changes. This provides an indicator to evaluate the duration of the abnormal conditions in the system.

[0135] Furthermore, voltage harmonic distortion, amplitude disturbance, amplitude disturbance time, absolute value and disturbance, absolute value and disturbance time, power frequency disturbance and power frequency disturbance time are important parameters for measuring the quality of power grid operation. The reason why they can guide the determination of power grid operation quality is that: voltage harmonic distortion reflects the harmonic content in the voltage waveform. High voltage harmonic distortion may lead to unstable equipment operation, increased heat loss and other problems, thus affecting the safety and stability of the power grid. Amplitude disturbance and amplitude disturbance time reflect the change of voltage amplitude. Large amplitude changes and long disturbance time may cause unstable equipment performance and affect power quality. Absolute value and disturbance and absolute value and disturbance time are used to detect the fluctuation of current or power signals. Abnormal absolute value and fluctuation may indicate faults or abnormal conditions in the power system. Power frequency disturbance and power frequency disturbance time reflect the fluctuation of power grid frequency. Frequency instability will affect the operation and power quality of power equipment.

[0136] S5: Detect the power grid operation quality in real time according to the disturbance parameters.

[0137] In a possible implementation, S5 is specifically:

[0138] According to the following formula, the power grid operation quality parameters are calculated:

[0139] Q=λ THD THD+λ A δ A T A +λ S δ S T S +λ f δ f T f

[0140] Among them, Q represents the power grid operation quality parameter, THD represents the voltage harmonic distortion, λ THD The weight coefficient of voltage harmonic distortion, δ A represents the amplitude disturbance, T A represents the amplitude perturbation time, λ A Represents the weight coefficient of amplitude disturbance, δ S Indicates the absolute value and disturbance degree, TS represents the absolute value and disturbance time, λ S Represents the weight coefficient of absolute value and perturbation degree, δ f Indicates the power frequency disturbance, T f represents the power frequency disturbance time, λ f The weight coefficient representing the power frequency disturbance degree.

[0141] Among them, those skilled in the art can set the weight coefficient λ of the voltage harmonic distortion according to actual conditions. THD , the weight coefficient λ of the amplitude disturbance A , absolute value and perturbation weight coefficient λ S And the weight coefficient λ of the power frequency disturbance f The present invention does not limit the size.

[0142] It should be noted that the larger the grid operation quality parameter is, the worse the current grid operation quality is. Conversely, the smaller the grid operation quality parameter is, the better the current grid operation quality is.

[0143] In the present invention, by comprehensively considering parameters such as voltage harmonic distortion, amplitude disturbance, absolute value and disturbance, power frequency disturbance and the like, the grid operation quality parameters obtained can more comprehensively reflect the overall situation of the grid quality, taking into account not only the stability of voltage and the accuracy of frequency, but also factors such as harmonic distortion and amplitude change.

[0144] S6: When the grid operation quality is poor, intelligent diagnosis is performed based on the improved GRU-RNN.

[0145] It should be noted that intelligent diagnosis algorithms usually require a lot of computing resources and time. Only performing diagnosis when the grid operation quality is abnormal can avoid unnecessary computing overhead. When the grid is operating normally, there is no need to run the diagnosis algorithm frequently, which saves computing resources and improves the efficiency of the system.

[0146] In a possible implementation, S6 specifically includes: when the power grid operation quality parameter is higher than a preset value, it indicates that the power grid operation quality is poor, and intelligent diagnosis is performed based on the improved GRU-RNN.

[0147] Among them, those skilled in the art can set the size of the preset value according to actual conditions, and the present invention does not limit it.

[0148] Optionally, the voltage signal sequence can be used to train a new GRU-RNN for intelligent diagnosis.

[0149] In a possible implementation, the intelligent diagnosis based on the improved GRU-RNN in S6 specifically includes sub-steps S601 to S606:

[0150] S601: The disturbance parameters in a preset time period before the current moment are combined into a sequence x, and input into the improved GRU-RNN.

[0151] It should be noted that taking the disturbance parameter sequence as input and using the sequence modeling capability of GRU-RNN to learn and capture the time series information in the sequence helps to understand the relationship, pattern and evolution law between the disturbance parameters. Compared with using the voltage signal sequence as input, using the disturbance parameter as input, since the disturbance parameter is a high-level representation of the voltage signal and contains the key information and characteristics of the power grid operation state, using the disturbance parameter as input makes it easier for the model to learn the common characteristics of the power grid operation state, thereby improving the generalization ability of the model, that is, it has better predictive ability for unseen data.

[0152] S602: In the update gate of the GRU-RNN, the output vector of the update gate is calculated according to the following formula:

[0153] z t =σ(W hz h t-1 +b z )

[0154] Among them, z t represents the output vector of the update gate at time t, σ() represents the activation function, and W hz represents the self-connection weight matrix of the update gate between time t and time t-1, h t-1 represents the hidden state vector at time t-1, b z Represents the bias term for the update gate.

[0155] S603: In the reset gate of the GRU-RNN, the output vector of the reset gate is calculated according to the following formula:

[0156] r t =σ(W hr h t-1 +b r )

[0157] Among them, r t represents the output vector of the reset gate at time t, W hr represents the self-connection weight matrix between the reset gate at time t and time t-1, b r Represents the bias term for the reset gate.

[0158] S604: In the candidate layer of the GRU-RNN, the output vector of the candidate layer is calculated according to the following formula:

[0159] c t =σ(W xc x t +Whc (r t ⊙h t-1 )+b c )

[0160] Among them, c t represents the output vector of the candidate layer at time t, W xc represents the weight matrix between the input layer and the candidate layer, x t represents the disturbance parameter combination at time t, W hc represents the self-connection weight matrix between the candidate layer at time t and time t-1, b c represents the bias term of the candidate layer, and ⊙ represents the element-wise product operation.

[0161] S605: In the hidden layer of the GRU-RNN, the hidden state vector at the moment is calculated according to the following formula:

[0162] h t =(1-z t )⊙h t-1 +z t ⊙c t

[0163] Among them, h t Represents the hidden state vector at time t.

[0164] S606: In the output layer of the GRU-RNN, the voltage prediction signal is calculated according to the following formula:

[0165] y t =Softmax(W yh h t +b y )

[0166] y t =[y t (1), y t (2), …, y t (m)]

[0167] Among them, y t represents the intelligent diagnosis result at time t, Softmax() represents the Softmax activation function, W yh represents the weight matrix between the hidden layer and the output layer, b y represents the bias term of the output layer, y t (j) represents the probability of classification into the jth state category, j = 1, 2, ..., m, m represents the total number of state categories.

[0168] Optionally, the improved GRU-RNN used in the intelligent diagnosis process can adopt the same structure as the improved GRU-RNN used to predict the operating status in the previous article, and can achieve the same technical effect. To avoid repetition, the present invention will not go into details.

[0169] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0170] (1) In the present invention, wavelet transform is used to extract multiple feature graphs from the voltage prediction signal, which can more comprehensively reflect the changes in the power grid state. The disturbance parameters are calculated based on the feature graphs. The disturbance parameters may include frequency disturbance, voltage fluctuation, harmonic distortion, etc., which can more accurately quantify the disturbance in the power grid. Then, the power grid operation quality can be detected in real time based on the disturbance parameters. This can more accurately reflect the power grid operation status when a large number of wind power systems are connected, adapt to the intermittent and uncertain nature of the wind power system, reduce false alarms, and improve detection accuracy.

[0171] (2) In the present invention, the voltage signal is predicted by the improved GRU-RNN model, which can more accurately predict the voltage change trend in the future period, carry out status warning, and respond to the changes in the grid operation status in a timely manner, which helps to make corresponding adjustments and measures in advance before changes occur in the grid.

[0172] (3) In the present invention, when the grid operation quality is poor, the improved GRU-RNN is used for intelligent diagnosis, which can help determine the possible problems or faults in the grid, help to quickly locate the root cause of the problem, and guide the operation and maintenance personnel to perform effective maintenance and processing, which can effectively monitor and manage the grid operation status and improve the reliability, safety and economy of the grid.

[0173] Reference Manual Attached Figure 2 , showing a structural schematic diagram of a real-time inspection and intelligent diagnosis system for power grid operation mode provided by the present invention.

[0174] The present invention further provides a real-time inspection and intelligent diagnosis system 20 for power grid operation mode, which is applied to the above-mentioned real-time inspection and intelligent diagnosis method for power grid operation mode, and comprises:

[0175] Processor 201;

[0176] The memory 202 stores computer-readable instructions, and when the computer-readable instructions are executed by the processor 201, the real-time inspection and intelligent diagnosis method of the power grid operation mode as described in the method embodiment is implemented.

[0177] The real-time inspection and intelligent diagnosis system 20 for power grid operation mode provided by the present invention can execute the above-mentioned real-time inspection and intelligent diagnosis method for power grid operation mode, and achieve the same or similar technical effects. To avoid repetition, the present invention will not go into details.

[0178] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0179] (1) In the present invention, wavelet transform is used to extract multiple feature graphs from the voltage prediction signal, which can more comprehensively reflect the changes in the power grid state. The disturbance parameters are calculated based on the feature graphs. The disturbance parameters may include frequency disturbance, voltage fluctuation, harmonic distortion, etc., which can more accurately quantify the disturbance in the power grid. Then, the power grid operation quality can be detected in real time based on the disturbance parameters. It can more accurately reflect the power grid operation status when a large number of wind power systems are connected, adapt to the intermittent and uncertain nature of the wind power system, reduce false alarms, and improve detection accuracy.

[0180] (2) In the present invention, the voltage signal is predicted by the improved GRU-RNN model, which can more accurately predict the voltage change trend in the future period, carry out status warning, and respond to the changes in the grid operation status in a timely manner, which helps to make corresponding adjustments and measures in advance before changes occur in the grid.

[0181] (3) In the present invention, when the grid operation quality is poor, the improved GRU-RNN is used for intelligent diagnosis, which can help determine the possible problems or faults in the grid, help to quickly locate the root cause of the problem, and guide the operation and maintenance personnel to perform effective maintenance and processing, which can effectively monitor and manage the grid operation status and improve the reliability, safety and economy of the grid.

[0182] It should be understood that the processor in the embodiment of the present invention may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0183] It should also be understood that the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0184] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0185] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0186] In the present invention, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can be represented by: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0187] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0188] Those skilled in the art will appreciate that the units and algorithm steps of each example 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 performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0189] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0190] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0191] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0192] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0193] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0194] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein when the program is executed by a processor, the method for real-time inspection and intelligent diagnosis of power grid operation mode as described in the method embodiment is implemented.

[0195] A computer-readable storage medium provided by the present invention can implement the steps and effects of the real-time inspection and intelligent diagnosis method of the power grid operation mode of the above-mentioned method embodiment. To avoid repetition, the present invention will not go into details.

[0196] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0197] (1) In the present invention, wavelet transform is used to extract multiple feature graphs from the voltage prediction signal, which can more comprehensively reflect the changes in the power grid state. The disturbance parameters are calculated based on the feature graphs. The disturbance parameters may include frequency disturbance, voltage fluctuation, harmonic distortion, etc., which can more accurately quantify the disturbance in the power grid. Then, the power grid operation quality can be detected in real time based on the disturbance parameters. It can more accurately reflect the power grid operation status when a large number of wind power systems are connected, adapt to the intermittent and uncertain nature of the wind power system, reduce false alarms, and improve detection accuracy.

[0198] (2) In the present invention, the voltage signal is predicted by the improved GRU-RNN model, which can more accurately predict the voltage change trend in the future period, carry out status warning, and respond to the changes in the grid operation status in a timely manner, which helps to make corresponding adjustments and measures in advance before changes occur in the grid.

[0199] (3) In the present invention, when the grid operation quality is poor, the improved GRU-RNN is used for intelligent diagnosis, which can help determine the possible problems or faults in the grid, help to quickly locate the root cause of the problem, and guide the operation and maintenance personnel to perform effective maintenance and processing, which can effectively monitor and manage the grid operation status and improve the reliability, safety and economy of the grid.

[0200] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0201] There are a few points to note:

[0202] (1) The drawings of the embodiments of the present invention only relate to the structures related to the embodiments of the present invention, and other structures may refer to the general design.

[0203] (2) For the sake of clarity, in the drawings used to describe the embodiments of the present invention, the thickness of the layers or regions is exaggerated or reduced, that is, these drawings are not drawn according to the actual scale. It is understood that when an element such as a layer, film, region or substrate is referred to as being "on" or "under" another element, the element may be "directly" "on" or "under" the other element or there may be intermediate elements.

[0204] (3) In the absence of conflict, the embodiments of the present invention and the features therein may be combined with each other to obtain new embodiments.

[0205] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A real-time inspection and intelligent diagnosis method for the operation mode of a power grid with large-scale access of wind power, characterized in that: include: S1: Acquire a power grid operation state sequence, wherein the power grid operation state sequence includes a voltage signal sequence; S2: According to the voltage signal sequence, based on the improved GRU-RNN, voltage prediction is performed to obtain a voltage prediction signal; specifically including: S201: In the input layer of the GRU-RNN, obtaining the voltage signal sequence; S202: In the update gate of the GRU-RNN, the output vector of the update gate is calculated according to the following formula: ， in, z t express t Update the output vector of the gate at all times, σ ( ) represents the activation function, W hz Indicates that the update gate is t Moment and t -1 self-connection weight matrix between moments, h t-1 express t The hidden state vector at time -1, b z represents the bias term of the update gate; S203: In the reset gate of the GRU-RNN, the output vector of the reset gate is calculated according to the following formula: ， in, r t express t The output vector of the reset gate at time t, W hr Indicates that the reset gate is t Moment and t -1 self-connection weight matrix between moments, b r represents the bias term of the reset gate; S204: In the candidate layer of the GRU-RNN, the output vector of the candidate layer is calculated according to the following formula: ， in, c t express t The output vector of the candidate layer at time t, W uc represents the weight matrix between the input layer and the candidate layer, u t express t The voltage signal at the moment, W hc Indicates that the candidate layer is t Moment and t -1 self-connection weight matrix between moments, b c represents the bias term of the candidate layer, Represents element-wise product operation; S205: In the hidden layer of GRU-RNN, the hidden state vector at the moment is calculated according to the following formula: ， in, h t express t Hidden state vector at time t; S206: In the output layer of the GRU-RNN, the voltage prediction signal is calculated according to the following formula: ， in, express t The voltage prediction signal at the moment, W yh represents the weight matrix between the hidden layer and the output layer, b y Represents the bias term of the output layer; S3: extracting a plurality of feature graphs from the voltage prediction signal through wavelet transform; S4: Calculate disturbance parameters according to the characteristic diagram; the disturbance parameters specifically include: voltage harmonic distortion, amplitude disturbance, amplitude disturbance time, absolute value and disturbance, absolute value and disturbance time, power frequency disturbance and power frequency disturbance time; S5: Detecting the power grid operation quality in real time according to the disturbance parameters; S6: When the power grid operation quality is poor, intelligent diagnosis is performed based on the improved GRU-RNN; the intelligent diagnosis based on the improved GRU-RNN specifically includes: S601: Sequence the disturbance parameters in the preset time period before the current moment x , input into the improved GRU-RNN; S602: In the update gate of the GRU-RNN, the output vector of the update gate is calculated according to the following formula: ， in, z t express t Update the output vector of the gate at all times, σ ( ) represents the activation function, W hz Indicates that the update gate is t Moment and t -1 self-connection weight matrix between moments, h t-1 express t The hidden state vector at time -1, b z represents the bias term of the update gate; S603: In the reset gate of the GRU-RNN, the output vector of the reset gate is calculated according to the following formula: ， in, r t express t The output vector of the reset gate at time t, W hr Indicates that the reset gate is t Moment and t -1 self-connection weight matrix between moments, b r represents the bias term of the reset gate; S604: In the candidate layer of the GRU-RNN, the output vector of the candidate layer is calculated according to the following formula: ， in, c t express t The output vector of the candidate layer at time t, W xc represents the weight matrix between the input layer and the candidate layer, x t express t The disturbance parameters at time are combined, W hc Indicates that the candidate layer is t Moment and t -1 self-connection weight matrix between moments, b c represents the bias term of the candidate layer, Represents element-wise product operation; S605: In the hidden layer of the GRU-RNN, the hidden state vector at the moment is calculated according to the following formula: ， in, h t express t Hidden state vector at time t; S606: In the output layer of the GRU-RNN, the voltage prediction signal is calculated according to the following formula: ， ， in, express t The intelligent diagnosis result at the moment, Softmax( ) represents the Softmax activation function, W yh represents the weight matrix between the hidden layer and the output layer, b y represents the bias term of the output layer, y t ( j ) indicates classification to j The probability of the state category, , m Indicates the total number of status categories.

2. The real-time inspection and intelligent diagnosis method of power grid operation mode according to claim 1 is characterized in that: The S3 specifically includes: S301: performing wavelet transform on the voltage prediction signal, decomposing the voltage prediction signal into different frequency components, and providing the temporal distribution of these components; S302: Determine an amplitude curve graph, an absolute value and curve graph, and a power frequency curve graph according to the coefficients after wavelet decomposition.

3. The real-time inspection and intelligent diagnosis method of power grid operation mode according to claim 2 is characterized in that: The S4 specifically includes: S401: Determine the voltage harmonic distortion degree according to the fundamental wave component and harmonic component obtained by wavelet transform; S402: Determine the amplitude disturbance degree and the amplitude disturbance time according to the amplitude curve graph; S403: Determine the absolute value and the disturbance degree and the absolute value and the disturbance time according to the absolute value and the curve graph; S404: Determine the power frequency disturbance degree and the power frequency disturbance time according to the power frequency curve.

4. The method for real-time inspection and intelligent diagnosis of power grid operation mode according to claim 3 is characterized in that: The voltage harmonic distortion is specifically: ， in, THD Indicates the voltage harmonic distortion, V 1 represents the fundamental voltage value, Indicates each harmonic voltage value; The amplitude disturbance degree is determined in the following manner: ， in, δ A represents the amplitude disturbance, A max Indicates the maximum amplitude in the amplitude curve graph, A min Indicates the minimum amplitude value in the amplitude curve graph; The amplitude disturbance time is determined in the following manner: ， ， in, T A Indicates the amplitude disturbance time, specifically the sum of the moments when the absolute value of the slope in the amplitude curve is greater than the preset amplitude slope. t A Indicates the moment when the absolute value of the slope in the amplitude curve is greater than the preset amplitude slope. A represents the amplitude, Indicates the amplitude curve t A The slope at time, k A Indicates the preset amplitude slope; The absolute value and the disturbance degree are determined in the following manner: ， in, δ S represents the absolute value and the degree of disturbance, S max Indicates the absolute value and the maximum value in the curve graph. S min Indicates the absolute value and the minimum value in the absolute value and curve graph; The absolute value and disturbance time are determined in the following manner: ， ， in, T S Represents the absolute value and disturbance time, specifically the sum of the absolute value and the time when the absolute value of the slope in the curve is greater than the preset absolute value and slope. t S Indicates the moment when the absolute value and the absolute value of the slope in the curve are greater than the preset absolute value and slope. S represents the absolute value and, Indicates absolute value and curve t S The slope at time, k S Indicates preset absolute value and slope; The power frequency disturbance degree is determined in the following manner: ， in, δ f represents the power frequency disturbance, f max Indicates the maximum power frequency in the power frequency curve, f min Indicates the minimum power frequency in the power frequency curve; The power frequency disturbance time is determined in the following manner: ， ， in, T f Indicates the power frequency disturbance time, specifically the sum of the moments when the absolute value of the slope in the power frequency curve is greater than the preset power frequency slope. t f Indicates the moment when the absolute value of the slope in the power-frequency curve is greater than the preset power-frequency slope. f represents the power frequency, Indicates the power frequency curve t f The slope at time, k f Indicates the preset power frequency slope.

5. The real-time inspection and intelligent diagnosis method of power grid operation mode according to claim 3 is characterized in that: The S5 is specifically: According to the following formula, the power grid operation quality parameters are calculated: ， in, Q Indicates the power grid operation quality parameters, THD Indicates the voltage harmonic distortion, λ THD Represents the weight coefficient of voltage harmonic distortion, δ A represents the amplitude disturbance, T A represents the amplitude disturbance time, λ A represents the weight coefficient of the amplitude disturbance, δ S represents the absolute value and the degree of disturbance, T S represents the absolute value and disturbance time, λ S represents the weight coefficient of absolute value and perturbation degree, δ f represents the power frequency disturbance, T f represents the power frequency disturbance time, λ f The weight coefficient representing the power frequency disturbance degree.

6. The method for real-time inspection and intelligent diagnosis of power grid operation mode according to claim 5, characterized in that: The S6 is specifically: When the grid operation quality parameters are higher than the preset values, it indicates that the grid operation quality is poor, and intelligent diagnosis is performed based on the improved GRU-RNN.

7. A real-time inspection and intelligent diagnosis system for power grid operation mode, characterized in that: include: processor; A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the method for real-time inspection and intelligent diagnosis of power grid operation mode according to any one of claims 1 to 6 is implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the real-time inspection and intelligent diagnosis method of the power grid operation mode as described in any one of claims 1 to 6 is implemented.

Images