Receiving end system transient voltage stability evaluation method and device

Abstract

The invention provides a receiving end system transient voltage stability evaluation method and device, and belongs to the field of power system operation control. The method comprises the following steps: firstly, acquiring voltage time sequence data after a fault of a receiving-end system based on a sliding time window sampling method, secondly, performing dimension reduction optimization processing on the acquired voltage time sequence data, removing noise and redundant information of the time sequence data, improving feature extraction efficiency and quality, and further improving feature extraction accuracy. And inputting the voltage time sequence data subjected to dimension reduction optimization processing into a trained transient voltage prediction model to obtain a transient voltage prediction value of a next moment output by the transient voltage prediction model, and finally performing voltage instability evaluation on the receiving end system according to the transient voltage prediction value of the next moment. And the efficiency and the accuracy of voltage instability evaluation are improved.

Description

Technical Field

[0001] The present application relates to the field of power system operation control, and particularly to a method and device for evaluating the transient voltage stability of a receiving-end system. Background Art

[0002] With the continuous increase in the grid connection scale of new energy power sources such as wind and light, the "hollowing out" of the load center caused by the withdrawal of traditional generating units, and the investment in a high proportion of power electronic devices, the operating characteristics of the new power grid have changed greatly, making the transient voltage stability problem more prominent, and the existing transient voltage stability evaluation methods are not applicable.

[0003] Therefore, how to quickly and accurately evaluate the transient voltage stability of the existing power system is an urgent problem to be solved at present. Summary of the Invention

[0004] Embodiments of the present application provide a method and device for evaluating the transient voltage stability of a receiving-end system to solve the problems of low efficiency and poor accuracy in voltage instability evaluation.

[0005] In a first aspect, embodiments of the present application provide a method for evaluating the transient voltage stability of a receiving-end system, including:

[0006] Collecting voltage time-series data after a fault in the receiving-end system based on the sliding time window sampling method;

[0007] Performing dimensionality reduction and optimization processing on the voltage time-series data;

[0008] Inputting the voltage time-series data after dimensionality reduction and optimization processing into a trained transient voltage prediction model to obtain the transient voltage prediction value at the next moment output by the transient voltage prediction model; wherein, the transient voltage prediction model is trained based on the voltage time-series data at different moments and the corresponding transient voltage at the next moment;

[0009] Evaluating the voltage instability of the receiving-end system according to the transient voltage prediction value at the next moment.

[0010] In a possible implementation manner, the performing dimensionality reduction and optimization processing on the voltage time-series data includes:

[0011] Normalizing the voltage time-series data, calculating the covariance matrix of the normalized data, solving the eigenvalues and corresponding eigenvectors of the covariance matrix, sorting the eigenvalues from large to small, and selecting the eigenvectors corresponding to the first preset number of eigenvalues as new basis vectors, and obtaining the voltage time-series data after dimensionality reduction based on the new basis vectors;

[0012] Input the voltage time series data after dimensionality reduction into the trained feature optimization model to obtain the voltage time series data after dimensionality reduction and optimization processed output by the feature optimization model; wherein, the feature optimization model optimizes the input data through a deep network constructed by stacking multiple DAE layer by layer.

[0013] In a possible implementation manner, before inputting the voltage time series data after dimensionality reduction and optimization processing into the trained transient voltage prediction model, it further includes:

[0014] Obtain the original feature set;

[0015] Perform feature dimensionality reduction and feature optimization processing on the original feature set to determine the initial feature set;

[0016] Train the transient voltage prediction model according to the initial feature set to obtain the trained transient voltage prediction model; wherein, the transient voltage prediction model includes a forgetting gate, an input gate, and an output gate.

[0017] In a possible implementation manner, the voltage instability assessment of the receiving-end system according to the predicted value of the transient voltage at the next moment includes:

[0018] Obtain the pre-stored correspondence between different transient voltages and different integral weights;

[0019] Based on the correspondence, determine the integral weight corresponding to the predicted value of the transient voltage at the next moment, and determine the transient voltage assessment index value according to the integral weight corresponding to the predicted value of the transient voltage at the next moment;

[0020] Perform voltage instability assessment on the receiving-end system based on the transient voltage assessment index value.

[0021] In a possible implementation manner, the determining the transient voltage assessment index value according to the integral weight corresponding to the predicted value of the transient voltage at the next moment includes:

[0022] Obtain the predicted value of the transient voltage, the true transient voltage, and the rated voltage at the current moment;

[0023] Based on the predicted value of the transient voltage and the true transient voltage at the current moment, the predicted value of the transient voltage at the next moment, the rated voltage, and the integral weight corresponding to the predicted value of the transient voltage at the next moment, perform numerical integration calculation to obtain the transient voltage assessment index.

[0024] In a possible implementation manner, performing numerical integration calculation based on the transient voltage prediction value and the true transient voltage at the current moment, the transient voltage prediction value at the next moment, the rated voltage, and the integration weight value corresponding to the transient voltage prediction value at the next moment to obtain the transient voltage evaluation index, including:

[0025] According to Determine the transient voltage evaluation index;

[0026] where η n+2 is the transient voltage evaluation index at t=n + 2, η n+1 is the transient voltage evaluation index at t=n + 1, U N is the rated voltage of the receiving-end system, U(t n ) is the true transient voltage at t=n, is the transient voltage prediction value at t=n, Kn is the integration weight value corresponding to the transient voltage prediction value at t=n, is the transient voltage prediction value at t=n + 1, K n+1 is the integration weight value corresponding to the transient voltage prediction value at t=n + 1.

[0027] In a possible implementation manner, performing voltage instability assessment on the receiving-end system based on the transient voltage evaluation index value, including:

[0028] Comparing the transient voltage evaluation index value with a preset voltage instability assessment index threshold;

[0029] If the transient voltage evaluation index value exceeds the voltage instability assessment index threshold, it is determined that voltage instability occurs.

[0030] In a possible implementation manner, after inputting the voltage time series data after dimensionality reduction optimization processing into the trained transient voltage prediction model and obtaining the transient voltage prediction value at the next moment output by the transient voltage prediction model, it further includes:

[0031] If the transient voltage prediction value at the next moment is inconsistent with the true transient voltage at the next moment, adjust the transient voltage prediction model so that the difference between the transient voltage prediction value output by the adjusted transient voltage prediction model and the corresponding true transient voltage is within a preset difference range.

[0032] In a second aspect, an embodiment of the present application provides a receiving-end system transient voltage stability assessment device, and the device includes:

[0033] An acquisition unit, configured to collect voltage time series data after a fault of the receiving-end system based on the sliding time window sampling method;

[0034] The first processing unit is configured to perform dimensionality reduction and optimization processing on the voltage time series data;

[0035] The second processing unit is configured to input the voltage time series data after dimensionality reduction and optimization processing into a trained transient voltage prediction model to obtain a predicted value of the transient voltage at the next moment output by the transient voltage prediction model; wherein, the transient voltage prediction model is trained based on the voltage time series data at different moments and the corresponding transient voltage at the next moment;

[0036] The evaluation unit is configured to perform voltage instability evaluation on the receiving-end system according to the predicted value of the transient voltage at the next moment.

[0037] In a possible implementation manner, the first processing unit is specifically configured to:

[0038] Standardize the voltage time series data, calculate the covariance matrix of the standardized data, solve the eigenvalues and corresponding eigenvectors of the covariance matrix, sort the eigenvalues from large to small, and select the eigenvectors corresponding to the first preset number of eigenvalues as new basis vectors, and based on the new basis vectors, obtain the voltage time series data after dimensionality reduction;

[0039] Input the voltage time series data after dimensionality reduction into a trained feature optimization model to obtain the voltage time series data after dimensionality reduction and optimization processing output by the feature optimization model; wherein, the feature optimization model optimizes the input data through a deep network constructed by stacking multiple DAE layers by layer.

[0040] In the embodiments of the present application, first, based on the sliding time window sampling method, the voltage time series data after the fault of the receiving-end system is collected, which not only effectively reduces the amount of data, but also can well capture the dynamic changes of the voltage data in the time series. Secondly, the obtained voltage time series data is subjected to dimensionality reduction and optimization processing to remove the noise and redundant information in the time series data, improve the efficiency and quality of feature extraction, effectively reduce the consumption of computing resources at the same time, and accelerate the model convergence speed; furthermore, the voltage time series data after dimensionality reduction and optimization processing is input into a trained transient voltage prediction model to obtain a predicted value of the transient voltage at the next moment output by the transient voltage prediction model. Finally, according to the predicted value of the transient voltage at the next moment, the voltage instability evaluation of the receiving-end system is performed, which improves the efficiency and accuracy of the voltage instability evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0042] Figure 1 It is a flowchart of the implementation of the transient voltage stability evaluation method for the receiving-end system provided by the embodiments of the present application;

[0043] Figure 2 It is a flowchart of the implementation of another transient voltage stability evaluation method for the receiving-end system provided by the embodiments of the present application;

[0044] Figure 3 It is a schematic structural diagram of the transient voltage stability evaluation device for the receiving-end system provided by the embodiments of the present application. Detailed implementation manners

[0045] In the following description, specific details such as specific system structures and technologies are presented for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of the present application. However, those skilled in the art should clearly understand that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present application. Summary of the invention

[0047] The applicant of the present invention has found that with the continuous growth of the grid-connected scale of new energy power sources such as wind and light, the "hollowing out" of the load center caused by the withdrawal of traditional generating units, and the investment in a high proportion of power electronic devices, the operating characteristics of the new power grid have changed greatly, making the transient voltage stability problem more prominent. The existing transient voltage stability evaluation methods are not applicable. Therefore, how to quickly and accurately evaluate the transient voltage stability of the existing power system is an urgent problem to be solved at present.

[0048] With the idea of improving the efficiency and accuracy of voltage instability evaluation, in the embodiments of the present application, first, the voltage time-series data after a fault is processed by dimensionality reduction and optimization to remove the noise and redundant information in the voltage time-series data, improve the efficiency and quality of subsequent feature extraction. Then, the voltage time-series data after dimensionality reduction and optimization is input into the trained transient voltage prediction model to obtain the transient voltage prediction value at the next moment output by the transient voltage prediction model. Finally, based on the transient voltage prediction value at the next moment, the voltage instability of the receiving-end system is evaluated, improving the efficiency and accuracy of voltage instability evaluation.

[0049] To make the purpose, technical solutions, and advantages of the present application clearer, the following will be described through specific embodiments with reference to the accompanying drawings.

[0050] Figure 1 It is a flowchart of the implementation of the transient voltage stability evaluation method for the receiving-end system provided by the embodiments of the present application, and is described in detail as follows:

[0051] Step 101: Collect the voltage time series data after the fault of the receiving-end system based on the sliding time window sampling method.

[0052] Exemplarily, when a power system fault occurs, the voltage will undergo a series of complex changes. Based on this, in this embodiment, the sliding time window sampling method is used to collect the voltage time series data after the fault based on a preset window size and a preset sliding step.

[0053] In one example, the fault in this embodiment can be a short-circuit fault (such as three-phase short circuit, two-phase short circuit, etc.), generator trip, line switching, etc.

[0054] Step 102: Perform dimensionality reduction and optimization processing on the voltage time series data.

[0055] Exemplarily, in this embodiment, data dimensionality reduction processing such as dimensionality compression of the collected voltage time series data is performed. At the same time, information such as redundancy and noise in the data can also be removed to achieve the refinement of the voltage time series data.

[0056] Step 103: Input the voltage time series data after dimensionality reduction and optimization processing into the trained transient voltage prediction model to obtain the transient voltage prediction value at the next moment output by the transient voltage prediction model; wherein, the transient voltage prediction model is trained based on the voltage time series data at different moments and the corresponding transient voltage at the next moment.

[0057] In one example, in this embodiment, the model is pre-trained based on the voltage time series data at different moments and the corresponding transient voltage at the next moment until the value of the loss function is within a preset range, and a trained transient voltage prediction model is obtained.

[0058] Exemplarily, in this embodiment, the voltage time series data after dimensionality reduction and optimization processing is used as the model input data and input into the trained transient voltage prediction model. The trained transient voltage prediction model analyzes and processes the voltage time series data and outputs the transient voltage prediction value at the next moment.

[0059] Step 104: Perform voltage instability assessment on the receiving-end system according to the transient voltage prediction value at the next moment.

[0060] Exemplarily, in this embodiment, it is judged whether the power system is unstable according to the simulated transient voltage prediction value.

[0061] In one example, in this embodiment, the fluctuation of the system voltage is predicted based on the transient voltage prediction value at the next moment to achieve voltage instability assessment. For example, if the voltage in the system can return to the normal operation range after a period of time, the system is not unstable. On the contrary, if the voltage continues to be lower or higher than the normal range, the system is unstable.

[0062] In summary, in the embodiments of the present application, first, based on the sliding time window sampling method, the voltage time series data after the fault of the receiving-end system is collected, which not only effectively reduces the data volume, but also can well capture the dynamic changes of the voltage data in the time series. Secondly, the obtained voltage time series data is processed by dimensionality reduction optimization to remove the noise and redundant information in the time series data, improve the efficiency and quality of feature extraction, effectively reduce the consumption of computing resources, and accelerate the model convergence speed. Furthermore, the voltage time series data after dimensionality reduction optimization is input into the trained transient voltage prediction model to obtain the transient voltage prediction value at the next moment output by the transient voltage prediction model. Finally, according to the transient voltage prediction value at the next moment, the voltage instability assessment of the receiving-end system is carried out, improving the efficiency and accuracy of the voltage instability assessment.

[0063] To improve the calculation convergence speed of the transient voltage prediction model, reduce the consumption of computing resources, and at the same time, improve the efficiency and quality of feature extraction, the embodiments of the present application perform dimensionality reduction optimization processing on the obtained voltage time series data to remove the noise and redundant information in the time series data. At the same time, in order to more accurately evaluate voltage instability, different integral weights are assigned to different voltage drop intervals in the present application. Based on different integral weights, the transient voltage prediction value and the true value at the current moment, and the transient voltage prediction value at the next moment, the transient voltage evaluation index is calculated. Combining the transient voltage conditions at multiple moments to calculate the voltage evaluation index improves the rationality of the evaluation index. Then, based on the magnitude of the transient voltage evaluation index value, the voltage instability assessment of the receiving-end system is carried out, improving the accuracy of the voltage instability assessment.

[0064] Figure 2 FIG. is a flowchart for implementing another transient voltage stability assessment method for the receiving-end system provided by the embodiments of the present application, which is described in detail as follows:

[0065] Step 201: Collect the voltage time series data after the fault of the receiving-end system based on the sliding time window sampling method.

[0066] Exemplarily, refer to the relevant description in Figure 1 the embodiments for this step, and details are not described herein again.

[0067] Step 202: Standardize the voltage time series data, calculate the covariance matrix of the standardized data, solve the eigenvalues and corresponding eigenvectors of the covariance matrix, sort the eigenvalues from large to small, and select the eigenvectors corresponding to the first preset number of eigenvalues as the new basis vectors. Based on the new basis vectors, the voltage time series data after dimensionality reduction is obtained.

[0068] Exemplarily, in this embodiment, the voltage time series data is first standardized as follows, so that the mean of each feature is 0 and the variance is 1:

[0069]

[0070] where X ij is the element at the i-th row and j-th column of the data set, and μ j is the mean of the j-th feature, and σ j is the variance of the j-th feature; then calculate the covariance matrix of the standardized data as follows:

[0071]

[0072] where: n is the number of samples; X’ is the sample data; ∑ is the covariance matrix.

[0073] Solve for the eigenvalues and corresponding eigenvectors of the covariance matrix as follows:

[0074] ∑V i =λ i V i

[0075] where: λ i is the i-th eigenvalue of the covariance matrix, and V i is the corresponding eigenvector.

[0076] Finally, sort the eigenvalues from largest to smallest, and select the eigenvectors corresponding to the top preset number of eigenvalues as the new basis vectors. Based on the new basis vectors, obtain the voltage time series data W after dimensionality reduction as follows:

[0077] W=[v1,v2,v3,…,v k

[0078] In the formula: W is the voltage time series data after dimensionality reduction, and V i is the eigenvector.

[0079] Step 203, input the voltage time series data after dimensionality reduction into the trained feature optimization model to obtain the voltage time series data after dimensionality reduction and optimization processed by the output of the feature optimization model; wherein, the feature optimization model optimizes the input data through a deep network constructed by stacking multiple DAE layers by layer.

[0080] Exemplarily, in this embodiment, the voltage time series data after dimensionality reduction is input into the trained feature optimization model, and the deep network constructed by stacking multiple DAE layers by layer in the feature optimization model is used to optimize the input data, and the voltage time series data after dimensionality reduction and optimization processed by the output of the feature optimization model is obtained.

[0081] ​In one example, in this embodiment, the dimension-reduced voltage time-series data x is added with noise to obtain x'. The first DAE is trained to obtain the first-layer hidden representation h1. The hidden representation h1 of the first layer is used as the input of the second layer, and noise is added again to obtain h1', and the second DAE is trained to obtain the second-layer hidden representation h2. More layers are continuously stacked until the predetermined network depth is reached. The input of each layer is the output of the previous layer.

[0082] The hidden layer representation h1 is obtained through the encoding process mapping, and then the reconstructed z0 is obtained through the decoding mapping. By continuously optimizing the parameters, the error L between the reconstructed z0 and the input of x' is made small enough. Its calculation method is as follows:

[0083]

[0084] Among them, the encoding and decoding processes are as follows:

[0085]

[0086] In the formula: W and W' are weight matrices; b and b' are biases, and s is an activation function.

[0087] Step 204: Input the voltage time-series data after dimension reduction and optimization processing into the trained transient voltage prediction model to obtain the transient voltage prediction value at the next moment output by the transient voltage prediction model. Among them, the transient voltage prediction model is trained based on the voltage time-series data at different moments and the corresponding transient voltage at the next moment.

[0088] In one example, before step 204, this embodiment further includes:

[0089] Obtain the original feature set; perform feature dimension reduction and feature optimization processing on the original feature set to determine the initial feature set; train the transient voltage prediction model according to the initial feature set to obtain the trained transient voltage prediction model. Among them, the transient voltage prediction model includes a forgetting gate, an input gate, and an output gate.

[0090] In one example, after step 204, this embodiment further includes:

[0091] If the transient voltage prediction value at the next moment is inconsistent with the true transient voltage at the next moment, the transient voltage prediction model is adjusted so that the difference between the transient voltage prediction value output by the adjusted transient voltage prediction model and the corresponding true transient voltage is within the preset difference range.

[0092] Exemplarily, this embodiment first trains the model to obtain the transient voltage prediction model.

[0093] First, obtain the original feature set, such as simulation data sets under different operating modes and different fault conditions; perform feature dimensionality reduction and feature optimization on the original feature set, such as noise removal and other processing, to determine the initial feature set; according to the initial feature set, train the transient voltage prediction model to obtain a trained transient voltage prediction model; wherein, the transient voltage prediction model includes a forget gate, an input gate, and an output gate.

[0094] In one example, by introducing three gate control structures: a forget gate, an input gate, and an output gate, the long-term and short-term dependence problem between data is solved. The state update process of the transient voltage prediction model is as follows:

[0095]

[0096] Wherein, f t , i t and o t are the forget gate, the input gate, and the output gate, ct is the state of the hidden layer neuron at time t, xt is the input at time t, σ is the activation function, [ht-1,xt] is the concatenated vector of the output of the hidden layer at time t-1 and the input at time t, W is the weight, and b is the bias.

[0097] In a feasible implementation manner, input the voltage time series data after dimensionality reduction and optimization processing into the trained transient voltage prediction model to obtain the transient voltage prediction value at the next moment output by the transient voltage prediction model.

[0098] Meanwhile, if the transient voltage prediction value at the next moment is inconsistent with the true transient voltage at the next moment, adjust the transient voltage prediction model so that the difference between the transient voltage prediction value output by the adjusted transient voltage prediction model and the corresponding true transient voltage is within a preset difference range, realizing the accuracy of the transient voltage prediction value output by the transient voltage prediction model.

[0099] Step 205, obtain the corresponding relationship between different pre-stored transient voltages and different integration weights.

[0100] Exemplarily, to make the instability assessment more accurate, in this embodiment, different voltage drop intervals correspond to different integration weights. Before performing the instability assessment, obtain the corresponding relationship between different pre-stored transient voltages and different integration weights.

[0101] Step 206, based on the corresponding relationship, determine the integration weight corresponding to the transient voltage prediction value at the next moment, and determine the transient voltage assessment index value according to the integration weight corresponding to the transient voltage prediction value at the next moment.

[0102] Exemplarily, based on the corresponding relationship, this embodiment determines the integration weight value corresponding to the predicted transient voltage value at the next moment, and determines the transient voltage evaluation index value according to the integration weight value corresponding to the predicted transient voltage value at the next moment. For example, the predicted transient voltage value and the true transient voltage at the current moment, as well as the rated voltage, are obtained; based on the predicted transient voltage value and the true transient voltage at the current moment, the predicted transient voltage value at the next moment, the rated voltage, and the integration weight value corresponding to the predicted transient voltage value at the next moment, numerical integration calculation is performed to obtain the transient voltage evaluation index.

[0103] For example, this embodiment can calculate to obtain the transient voltage evaluation index; where η n+2 is the transient voltage evaluation index at t = n + 2, η n+1 is the transient voltage evaluation index at t = n + 1, U N is the rated voltage of the receiving-end system, U(t n ) is the true transient voltage at t = n, is the predicted transient voltage value at t = n, Kn is the integration weight value corresponding to the predicted transient voltage value at t = n, is the predicted transient voltage value at t = n + 1, K n+1 is the integration weight value corresponding to the predicted transient voltage value at t = n + 1.

[0104] Step 207, based on the transient voltage evaluation index value, perform voltage instability assessment on the receiving-end system.

[0105] In one example, step 207 includes the following steps:

[0106] Compare the transient voltage evaluation index value with a preset voltage instability evaluation index threshold; if the transient voltage evaluation index value exceeds the voltage instability evaluation index threshold, it is determined that voltage instability occurs.

[0107] Exemplarily, after calculating the transient voltage evaluation index value, this embodiment compares the transient voltage evaluation index value with a preset voltage instability evaluation index threshold; if the transient voltage evaluation index value exceeds the voltage instability evaluation index threshold, it is determined that voltage instability occurs.

[0108] In a feasible implementation manner, after determining voltage instability, this embodiment gives a corresponding warning to notify the technical personnel to take measures. If the voltage is not unstable, continue with the steps of collecting voltage time-series data based on the sliding time window sampling method and subsequent instability assessment based on the voltage time-series data until the voltage returns to normal or becomes unstable.

[0109] It should be understood that the sequence numbers of the steps in the above embodiments do not indicate the order of execution, and the execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0110] The following is the device embodiment of the present application. For the details not described in detail, reference can be made to the corresponding method embodiments above.

[0111] Figure 3 The structural schematic diagram of the receiving-end system transient voltage stability evaluation device provided by the embodiments of the present application is shown. For the convenience of description, only the parts related to the embodiments of the present application are shown and are described in detail as follows:

[0112] As Figure 3 shown, the receiving-end system transient voltage stability evaluation device includes:

[0113] An acquisition unit 31, configured to collect voltage time series data after a fault in the receiving-end system based on a sliding time window sampling method.

[0114] A first processing unit 32, configured to perform dimensionality reduction and optimization processing on the voltage time series data.

[0115] A second processing unit 33, configured to input the voltage time series data after dimensionality reduction and optimization processing into a trained transient voltage prediction model, and obtain a transient voltage prediction value at the next moment output by the transient voltage prediction model. Among them, the transient voltage prediction model is trained based on voltage time series data at different moments and the corresponding transient voltage at the next moment.

[0116] An evaluation unit 34, configured to perform voltage instability evaluation on the receiving-end system according to the transient voltage prediction value at the next moment.

[0117] In a feasible implementation manner, the first processing unit 32 is specifically configured to:

[0118] Normalize the voltage time series data, calculate the covariance matrix of the normalized data, solve the eigenvalues and corresponding eigenvectors of the covariance matrix, sort the eigenvalues from large to small, and select the eigenvectors corresponding to the first preset number of eigenvalues as new basis vectors, and obtain the voltage time series data after dimensionality reduction based on the new basis vectors.

[0119] Input the voltage time series data after dimensionality reduction into a trained feature optimization model, and obtain the voltage time series data after dimensionality reduction and optimization processing output by the feature optimization model. Among them, the feature optimization model optimizes the input data through a deep network constructed by stacking multiple DAE layers by layer.

[0120] In a feasible implementation manner, before the second processing unit 33, the device further includes a training unit, and the training unit is specifically configured to:

[0121] Obtain the original feature set.

[0122] Perform feature dimensionality reduction and feature optimization on the original feature set to determine the initial feature set.

[0123] Train the transient voltage prediction model according to the initial feature set to obtain a trained transient voltage prediction model. Among them, the transient voltage prediction model includes a forgetting gate, an input gate, and an output gate.

[0124] In a feasible implementation manner, the evaluation unit 34 is specifically configured to:

[0125] Obtain the corresponding relationship between different transient voltages and different integral weights stored in advance.

[0126] Based on the corresponding relationship, determine the integral weight corresponding to the predicted value of the transient voltage at the next moment, and determine the transient voltage evaluation index value according to the integral weight corresponding to the predicted value of the transient voltage at the next moment.

[0127] Perform voltage instability assessment on the receiving-end system based on the transient voltage evaluation index value.

[0128] In a feasible implementation manner, the evaluation unit 34 is specifically configured to:

[0129] Obtain the predicted value of the transient voltage, the true transient voltage, and the rated voltage at the current moment.

[0130] Perform numerical integration calculation based on the predicted value of the transient voltage, the true transient voltage at the current moment, the predicted value of the transient voltage at the next moment, the rated voltage, and the integral weight corresponding to the predicted value of the transient voltage at the next moment to obtain the transient voltage evaluation index.

[0131] In a feasible implementation manner, the evaluation unit 34 is specifically configured to:

[0132] According to Determine the transient voltage evaluation index.

[0133] Among them, η n+2 is the transient voltage evaluation index at t = n + 2, η n+1 is the transient voltage evaluation index at t = n + 1, U N is the rated voltage of the receiving-end system, U(t n ) is the true transient voltage at t = n, is the predicted value of the transient voltage at t = n, Kn is the integral weight corresponding to the predicted value of the transient voltage at t = n, is the predicted value of the transient voltage at t = n + 1, K n+1 is the integral weight corresponding to the predicted value of the transient voltage at t = n + 1.

[0134] In a feasible implementation manner, the evaluation unit 34 is specifically configured to:

[0135] Compare the transient voltage evaluation index value with a preset voltage instability evaluation index threshold.

[0136] If the transient voltage evaluation index value exceeds the voltage instability evaluation index threshold, it is determined that voltage instability occurs.

[0137] In a feasible implementation manner, after the second processing unit 33, the device further includes: a third processing unit, which is specifically configured to:

[0138] If the predicted transient voltage value at the next moment is inconsistent with the true transient voltage at the next moment, adjust the transient voltage prediction model so that the difference between the transient voltage prediction value output by the adjusted transient voltage prediction model and the corresponding true transient voltage is within a preset difference range.

[0139] In the above embodiments, the descriptions of the various embodiments have their own focuses. For the parts not detailed or recorded in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0140] Those of ordinary skill in the art can realize that the templates, units, and algorithm steps of the examples described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0141] If a module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, to implement all or part of the processes of the above method embodiments, the present application can also be completed by a computer program instructing relevant hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above method embodiments for evaluating the transient voltage stability of each receiving-end system can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory, random access memory, electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0142] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A transient voltage stability assessment method for a receiving-end system, characterized in that, The method includes: Collecting voltage time series data after the fault of the receiving-end system based on the sliding time window sampling method; Performing dimensionality reduction and optimization processing on the voltage time series data; Inputting the voltage time series data after dimensionality reduction and optimization processing into the trained transient voltage prediction model to obtain the predicted value of the transient voltage at the next moment output by the transient voltage prediction model; wherein, the transient voltage prediction model is trained based on the voltage time series data at different moments and the corresponding transient voltage at the next moment; Performing voltage instability assessment on the receiving-end system according to the predicted value of the transient voltage at the next moment.

2. The transient voltage stability assessment method for the receiving-end system according to claim 1, wherein The performing dimensionality reduction and optimization processing on the voltage time series data includes: Normalizing the voltage time series data, calculating the covariance matrix of the normalized data, solving the eigenvalues and corresponding eigenvectors of the covariance matrix, sorting the eigenvalues from large to small, and selecting the eigenvectors corresponding to the first preset number of eigenvalues as the new basis vectors, and obtaining the voltage time series data after dimensionality reduction based on the new basis vectors; Inputting the voltage time series data after dimensionality reduction into the trained feature optimization model to obtain the voltage time series data after dimensionality reduction and optimization processing output by the feature optimization model; wherein, the feature optimization model optimizes the input data through a deep network constructed by stacking multiple DAE layer by layer.

3. The transient voltage stability assessment method for the receiving-end system according to claim 1, characterized in that Before inputting the voltage time series data after dimensionality reduction and optimization processing into the trained transient voltage prediction model, it further includes: Obtaining the original feature set; Performing feature dimensionality reduction and feature optimization processing on the original feature set to determine the initial feature set; Training the transient voltage prediction model according to the initial feature set to obtain the trained transient voltage prediction model; wherein, the transient voltage prediction model includes a forgetting gate, an input gate, and an output gate.

4. The transient voltage stability assessment method for the receiving-end system according to any one of claims 1 to 3, characterized in that The performing voltage instability assessment on the receiving-end system according to the predicted value of the transient voltage at the next moment includes: Obtaining the corresponding relationship between different transient voltages and different integration weights stored in advance; Based on the corresponding relationship, determining the integration weight corresponding to the predicted value of the transient voltage at the next moment, and determining the transient voltage assessment index value according to the integration weight corresponding to the predicted value of the transient voltage at the next moment; Performing voltage instability assessment on the receiving-end system based on the transient voltage assessment index value.

5. The transient voltage stability assessment method for the receiving-end system according to claim 4, characterized in that The determining the transient voltage assessment index value according to the integration weight corresponding to the predicted value of the transient voltage at the next moment includes: Obtaining the predicted value of the transient voltage and the true transient voltage at the current moment, and the rated voltage; Performing numerical integration calculation based on the predicted value of the transient voltage and the true transient voltage at the current moment, the predicted value of the transient voltage at the next moment, the rated voltage, and the integration weight corresponding to the predicted value of the transient voltage at the next moment to obtain the transient voltage assessment index.

6. The transient voltage stability assessment method for the receiving-end system according to claim 5, wherein The performing numerical integration calculation based on the predicted value of the transient voltage and the true transient voltage at the current moment, the predicted value of the transient voltage at the next moment, the rated voltage, and the integration weight corresponding to the predicted value of the transient voltage at the next moment to obtain the transient voltage assessment index includes: According to determine the transient voltage evaluation index; Among them, η n+2 is the transient voltage evaluation index at t = n + 2, and η n+1 is the transient voltage evaluation index at t = n + 1, U N is the rated voltage of the receiving-end system, U(t n ) is the true transient voltage at t = n, is the predicted value of the transient voltage at t = n, Kn is the integral weight corresponding to the predicted value of the transient voltage at t = n, is the predicted value of the transient voltage at t = n + 1, K n+1 is the integral weight corresponding to the predicted value of the transient voltage at t = n + 1.

7. The transient voltage stability assessment method for the receiving-end system according to claim 4, characterized in that Based on the transient voltage evaluation index value, performing voltage instability evaluation on the receiving-end system, including: Comparing the transient voltage evaluation index value with a preset voltage instability evaluation index threshold; If the transient voltage evaluation index value exceeds the voltage instability evaluation index threshold, determining voltage instability.

8. The transient voltage stability assessment method for the receiving-end system according to claim 1, characterized in that After inputting the voltage time series data after dimensionality reduction and optimization processing into the trained transient voltage prediction model to obtain the transient voltage prediction value at the next moment output by the transient voltage prediction model, it further includes: If the transient voltage prediction value at the next moment is inconsistent with the true transient voltage at the next moment, adjusting the transient voltage prediction model so that the difference between the transient voltage prediction value output by the adjusted transient voltage prediction model and the corresponding true transient voltage is within a preset difference range.

9. A transient voltage stability assessment device for a receiving-end system, characterized in that, The device includes: An acquisition unit, configured to acquire voltage time series data after a fault in the receiving-end system based on the sliding time window sampling method; A first processing unit, configured to perform dimensionality reduction and optimization processing on the voltage time series data; A second processing unit, configured to input the voltage time series data after dimensionality reduction and optimization processing into the trained transient voltage prediction model to obtain the transient voltage prediction value at the next moment output by the transient voltage prediction model; wherein, the transient voltage prediction model is trained based on voltage time series data at different moments and the corresponding transient voltage at the next moment; An evaluation unit, configured to perform voltage instability evaluation on the receiving-end system according to the transient voltage prediction value at the next moment.

10. The transient voltage stability evaluation device for the receiving-end system according to claim 9, wherein The first processing unit is specifically configured to: Standardize the voltage time series data, calculate the covariance matrix of the standardized data, solve the eigenvalues and corresponding eigenvectors of the covariance matrix, sort the eigenvalues from large to small, and select the eigenvectors corresponding to the first preset number of eigenvalues as new basis vectors, and based on the new basis vectors, obtain the voltage time series data after dimensionality reduction; Input the voltage time series data after dimensionality reduction into the trained feature optimization model to obtain the voltage time series data after dimensionality reduction and optimization processing output by the feature optimization model; wherein, the feature optimization model optimizes the input data through a deep network constructed by stacking multiple DAE layers by layer.

Images