Hybrid XLSTM-based electric power system operation safety early warning method and device

Abstract

The invention discloses a hybrid XLSTM-based power system operation safety early warning method and device, a storage medium and electronic equipment. The method comprises the following steps: acquiring a characteristic signal sequence of a power system; the characteristic signal sequence comprises a voltage signal, a power angle signal and a phase signal, and missing values in the characteristic signal sequence are filled; constructing a topological structure matrix based on the topological structure of the power system; inputting the filled feature signal sequence into a mixed XLSTM model for feature extraction to obtain a predicted future feature signal sequence; and inputting the predicted future feature sequence and the topological structure matrix into a hybrid XLSTM model for classification to obtain a classification result, the classification result including a risk and a risk-free result. According to the invention, more powerful comprehensive risk early warning performance of the operation safety of the power system can be realized.

Description

Technical Field

[0001] The present application relates to the field of power system operation control technology, and in particular to a power system operation safety early warning method, device, storage medium and electronic equipment based on hybrid XLSTM. Background Art

[0002] In the context of today's energy transition, early warning technologies for power system security and sufficiency have become particularly important. As renewable energy sources, particularly intermittent energy sources like wind and solar, increase their share of the power mix, their inherent uncertainty poses new challenges to the power system's supply and demand balance. The fluctuating output of these energy sources requires the power system to possess greater adaptability and flexibility to maintain a stable power supply. Furthermore, the structure of the power system is also undergoing transformation. The integration of distributed energy resources has transformed the power grid from a traditional passive network to an active one, resulting in power flows shifting from unidirectional to complex bidirectional or multidirectional flows. This change increases the complexity of power system operation and control and places higher demands on system stability and reliability.

[0003] At the same time, the digitalization and intelligentization of power systems present new cybersecurity threats. While the application of new technologies such as artificial intelligence and big data improves power system efficiency, it also introduces new security risks. Therefore, the power system requires more advanced early warning technologies to address these challenges and ensure the security of critical infrastructure. The frequent occurrence of extreme weather events is also a reality that the power system must face. These events can damage power facilities and impact the stability of power supply. Therefore, the power system needs to use early warning technologies to predict and adapt to these extreme events to minimize the impact on power supply. In the power market, with the liberalization and marketization of power trading, the power system needs to respond quickly to market changes, optimize energy supply costs, and reduce transmission line congestion. This requires the power system to have efficient early warning and response mechanisms to enable the rapid transaction and settlement of energy, information, and data.

[0004] Current power system risk assessments are mostly achieved by monitoring current or voltage signals, which makes it difficult to predict and prevent risks. Summary of the Invention

[0005] The embodiments of the present application provide a hybrid XLSTM-based power system operation safety early warning method, device, storage medium and electronic equipment, which can predict risks and take timely precautions.

[0006] The embodiment of the present application provides a hybrid XLSTM-based power system operation safety early warning method, including: Acquire a characteristic signal sequence of the power system; the characteristic signal sequence includes a voltage signal, a power angle signal, and a phase signal; Filling missing values in the characteristic signal sequence; Construct a topology matrix based on the topology of the power system; The padded feature signal sequence is input into the hybrid XLSTM model for feature extraction to obtain the predicted future feature signal sequence; The predicted future feature sequence and the topological structure matrix are input into a hybrid XLSTM model for classification to obtain classification results, which include risky and non-risky.

[0007] Furthermore, the above-mentioned electric power system operation safety early warning method based on hybrid XLSTM, wherein, before the step of filling the missing values in the characteristic signal sequence, includes: The data in the characteristic signal sequence are analyzed to obtain the distribution characteristics of the data, the pattern of missing data and the correlation between the data.

[0008] Furthermore, in the above-mentioned hybrid XLSTM-based power system operation safety early warning method, the missing values in the characteristic signal sequence are filled, including: Performing frequency domain analysis on sequence data in the characteristic signal sequence by Fourier transform to extract periodic information; Folding the characteristic signal sequence based on the period information to convert it into a two-dimensional tensor, and obtaining a two-dimensional time series change representation based on the two-dimensional tensor; Inputting the two-dimensional time series variation representation into a two-dimensional convolutional neural network for feature extraction to obtain periodic information; Expanding the two-dimensional time series variation representation to obtain a one-dimensional time series, and adaptively fusing the one-dimensional time series and the period to obtain a fused one-dimensional time series; The missing values in the characteristic signal sequence are filled based on the fused one-dimensional time series.

[0009] Furthermore, the above-mentioned power system operation safety early warning method based on hybrid XLSTM, wherein the padded feature signal sequence is input into the hybrid XLSTM model for feature extraction to obtain the predicted future feature signal sequence, includes: Input multiple padded feature signal sequences into the hybrid XLSTM model for feature extraction and then fusion to obtain the first fusion feature; The first fusion feature is input into the hybrid XLSTM model for prediction to obtain a future feature signal sequence.

[0010] Furthermore, the above-mentioned power system operation safety early warning method based on hybrid XLSTM, wherein the hybrid XLSTM model includes an LSTM layer, an attention enhancement module, a residual connection and a normalization layer; The hollow LSTM layer includes at least one LSTM unit, wherein the LSTM unit is an exponential gated scalar memory unit, or the LSTM unit is a matrix memory unit; An attention enhancement module connected to the atrous LSTM layer; A residual connection and normalization layer, comprising a plurality of residual blocks, each of which is integrated with an xLSTM unit; The fully connected layer is used to map the output of the xLSTM unit to the label space of the classification task and perform multi-category classification through the softmax function.

[0011] Furthermore, the above-mentioned hybrid XLSTM-based power system operation safety early warning method, wherein the predicted future feature sequence and the topological structure matrix are input into the hybrid XLSTM model for feature extraction and classification, includes: In the atrous LSTM layer, convolution is performed on the future feature sequence and the topological structure matrix respectively to obtain future time series features and topological features, and the future time series features and the topological features are fused to obtain a second fused feature; In the attention enhancement module, the attention weights of the historical hidden states are dynamically calculated to generate context-aware sequence representations. The weighted feature vectors are output and concatenated or added with the hidden states of the dilated LSTM layer to obtain weighted features. In the residual connection and normalization layer, the stabilized weighted features are output; In the fully connected layer, the weighted features are classified to obtain a classification result.

[0012] Furthermore, the above-mentioned power system operation safety early warning method based on hybrid XLSTM, wherein the method also includes: The characteristic signal sequence is collected through a SCADA system.

[0013] The embodiment of the present application also provides a hybrid XLSTM-based power system operation safety early warning device, including: A signal acquisition module is used to acquire a characteristic signal sequence of the power system; the characteristic signal sequence includes a voltage signal, a power angle signal and a phase signal; A missing value filling module is used to fill the missing values in the characteristic signal sequence; A topology matrix building module is used to build a topology structure matrix based on the topology structure of the power system; The prediction module is used to input the padded feature signal sequence into the hybrid XLSTM model for feature extraction to obtain the predicted future feature signal sequence; The classification module is used to input the predicted future feature sequence and the topological structure matrix into the hybrid XLSTM model for classification to obtain a classification result, which includes risky and non-risky.

[0014] An embodiment of the present application also provides a computer-readable storage medium, which stores multiple instructions, and the instructions are suitable for being loaded by a processor to execute any of the above-mentioned hybrid XLSTM-based power system operation safety early warning methods.

[0015] An embodiment of the present application also provides an electronic device, including a processor and a memory, wherein the processor is electrically connected to the memory, the memory is used to store instructions and data, and the processor is used for the steps in the power system operation safety early warning method based on hybrid XLSTM as described in any of the above items.

[0016] This application provides a hybrid XLSTM-based power system operation safety early warning method, device, storage medium, and electronic device. This application uses a SCADA data acquisition system to collect signals of some power system features, uses the Timsesnet algorithm to preprocess the features to fill missing values, predicts the selected features through hybrid XLSTM, and uses the XLSTM algorithm to classify the prediction results. The classification results are used to provide a comprehensive risk early warning technology for future power system operation safety. This method can achieve more powerful comprehensive risk early warning performance for power system operation safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The following detailed description of the specific embodiments of the present application in conjunction with the accompanying drawings will make the technical solutions and other beneficial effects of the present application apparent.

[0018] Figure 1 A flowchart of a hybrid XLSTM-based power system operation safety early warning method provided in an embodiment of the present application.

[0019] Figure 2 A flowchart for generating a future characteristic signal sequence provided in an embodiment of the present application.

[0020] Figure 3 A schematic diagram of the structure of the hybrid XLSTM model provided in an embodiment of the present application.

[0021] Figure 4 A structural diagram of the power system operation safety early warning device based on hybrid XLSTM provided in an embodiment of the present application.

[0022] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0024] The present invention provides a method, device, storage medium, and electronic device for early warning of power system operation safety based on hybrid XLSTM. The present invention provides an early warning device for power system operation safety based on hybrid XLSTM, which can be integrated into an electronic device, such as a terminal or server. The terminal can include a tablet computer, a laptop computer, a personal computer (PC), a microprocessor box, or other devices.

[0025] See also Figure 1 , Figure 1 A flowchart of a power system operation safety early warning method based on hybrid XLSTM provided in an embodiment of the present application, which is applied to an electronic device, includes the following steps: S1, obtain the characteristic signal sequence of the power system; the characteristic signal sequence includes voltage signal, power angle signal and phase signal.

[0026] Specifically, a SCADA (Supervisory Control and Data Acquisition) system is used to collect partial voltage signals, partial power angle signals, and partial phase signals of the power system.

[0027] S2, fill in the missing values in the feature signal sequence.

[0028] In one embodiment, before the step of filling the missing values in the characteristic signal sequence, the method includes: analyzing the data in the characteristic signal sequence to obtain the distribution characteristics of the data, the pattern of the missing data and the correlation between the data.

[0029] In one embodiment, step S2 includes the following steps: S21, performing frequency domain analysis on sequence data in the characteristic signal sequence through Fourier transform to extract periodic information.

[0030] By applying Fast Fourier Transform (FFT) to perform frequency domain analysis on the sequence data in the input feature signal sequence, significant periodic features in the data can be effectively extracted. This process involves converting the time series from the time domain to the frequency domain, thereby revealing the hidden periodic patterns in the data and providing valuable information for further analysis and processing; S22, folding the characteristic signal sequence based on the periodic information, converting it into a two-dimensional tensor, and obtaining a two-dimensional time series change representation based on the two-dimensional tensor.

[0031] The one-dimensional feature signal sequence data is reconstructed based on the extracted periodic information and converted into a two-dimensional tensor. In this tensor, each row corresponds to a complete period, and each column represents a specific time point within that period. This conversion yields a two-dimensional representation of time series changes, facilitating in-depth analysis of the periodic characteristics of time series data.

[0032] S23, inputting the two-dimensional time series change representation into a two-dimensional convolutional neural network for feature extraction to obtain periodic information.

[0033] By applying a 2D convolutional neural network to the transformed 2D tensor for deep feature extraction, we can effectively capture time series variations within and between cycles. The 2D convolutional layer can identify local features within periodic patterns. By stacking multiple convolutional layers, the model can learn higher-level abstract representations, thereby better understanding and modeling the periodic structure of time series data.

[0034] S24, expanding the two-dimensional time series change representation to obtain a one-dimensional time series, and performing adaptive fusion based on the one-dimensional time series and the period to obtain a fused one-dimensional time series.

[0035] The transformed two-dimensional time series representation is re-expanded into a one-dimensional time series, and an adaptive fusion strategy is used to integrate multi-period information. In this process, fusion weights are assigned based on the relative importance of each period, ensuring that key period features are highlighted while the influence of minor periods is reduced.

[0036] S25, filling missing values in the feature signal sequence based on the fused one-dimensional time series.

[0037] S3, constructs a topology matrix based on the topology of the power system.

[0038] The topology matrix represents the connection structure of the power system. In the topology matrix, 1 represents connection and 0 represents disconnection.

[0039] S4, inputs the padded feature signal sequence into the hybrid XLSTM model for feature extraction to obtain the predicted future feature signal sequence.

[0040] Figure 2 The flow chart for generating future characteristic signal sequences provided in the embodiment of the present application is as follows: Figure 2 As shown, step S4 includes the following steps: S41, inputting the multiple padded feature signal sequences into the hybrid XLSTM model for feature extraction and then fusing them to obtain a first fused feature; S42, input the first fused feature into the hybrid XLSTM model for prediction to obtain a future feature signal sequence.

[0041] S5, input the predicted future feature sequence and topological structure matrix into the hybrid XLSTM model for classification to obtain the classification results, which include risky and non-risky.

[0042] Figure 3 A structural diagram of the hybrid XLSTM model provided in the embodiment of the present application is shown as follows: Figure 3 As shown in Figure 2, the hybrid XLSTM model includes an LSTM layer, an attention enhancement module, a residual connection, and a normalization layer.

[0043] The atrous LSTM layer includes at least one LSTM unit, where the LSTM unit is an exponentially gated scalar memory unit, or the LSTM unit is a matrix memory unit.

[0044] Specifically, a scalar xLSTM (sLSTM) with an exponential gating mechanism or a matrix xLSTM (mLSTM) with matrix memory is used as an LSTM unit to receive the input sequence. The sLSTM controls the information flow through an exponential activation function, while the mLSTM enhances the storage capacity through matrix operations.

[0045] In mLSTM, matrix memory is used instead of scalar memory units, and matrix operations are used to enhance the storage capacity and parallel processing capabilities of the model, allowing more complex data relationships and patterns to be captured within a single time step.

[0046] Attention enhancement module, connected to the hollow LSTM layer.

[0047] The residual connection and normalization layer includes multiple residual blocks, in which xLSTM units are integrated.

[0048] Specifically, layer normalization (LayerNorm) is used to stabilize the input of each layer. Causal Conv1D is introduced to ensure the temporal order of information and prevent future information leakage. Residual connections are implemented to increase the stability of the model when processing deep networks. xLSTM units are integrated into residual blocks, and skip connections (SkipConnection) are used to directly pass the input to the block output and add it to the main path output to alleviate the problem of vanishing gradients.

[0049] The fully connected layer is used to map the output of the xLSTM unit to the label space of the classification task and perform multi-category classification through the softmax function.

[0050] At the last time step of the sequence, the output of xLSTM is mapped to the label space of the classification task through a fully connected layer, and multi-category classification is performed through the softmax function.

[0051] Step S5 specifically includes the following steps: S51, in the dilated LSTM layer, convolving the future feature sequence and the topological structure matrix respectively to obtain future time series features and topological features, fusing the future time series features and the topological features (fusing information about changes in the previous and subsequent topological structures into the two-dimensional data through convolution) to obtain a second fused feature; S52, in the attention enhancement module, dynamically calculate the attention weights of the historical hidden states, generate context-aware sequence representations, output the weighted feature vectors and concatenate or add them with the hidden states of the hollow LSTM layer to obtain weighted features; S53, in the residual connection and normalization layer, outputs the stabilized weighted features; S54: Classify the weighted features at the fully connected layer to obtain a classification result.

[0052] According to the method described in the above embodiment, this embodiment will be further described from the perspective of an electric power system operation safety early warning device based on hybrid XLSTM. The electric power system operation safety early warning device based on hybrid XLSTM can be implemented as an independent entity or integrated into an electronic device, which can be a terminal, server and other equipment. The terminal may include a tablet computer, a laptop computer, a personal computer (PC, Personal Computer), a micro processing box, or other equipment.

[0053] See also Figure 4 , Figure 4The present invention specifically describes an electric power system operation safety early warning device based on hybrid XLSTM provided in an embodiment of the present application, which is applied to electronic equipment. The electric power system operation safety early warning device based on hybrid XLSTM may include: A signal acquisition module is used to acquire a characteristic signal sequence of the power system; the characteristic signal sequence includes a voltage signal, a power angle signal and a phase signal; A missing value filling module is used to fill the missing values in the characteristic signal sequence; A topology matrix building module is used to build a topology structure matrix based on the topology structure of the power system; The prediction module is used to input the padded feature signal sequence into the hybrid XLSTM model for feature extraction to obtain the predicted future feature signal sequence; The classification module is used to input the predicted future feature sequence and the topological structure matrix into the hybrid XLSTM model for classification to obtain a classification result, which includes risky and non-risky.

[0054] During specific implementation, the above modules and / or units can be implemented as independent entities, or can be arbitrarily combined to be implemented as the same or several entities. The specific implementation of the above modules and / or units can refer to the previous method embodiments. The specific beneficial effects that can be achieved can also be found in the beneficial effects in the previous method embodiments, which will not be repeated here.

[0055] In addition, an embodiment of the present application further provides an electronic device, which may be a computer, tablet computer, or other device. The electronic device can implement the steps of any embodiment of the power system operation safety early warning method based on hybrid XLSTM provided in the embodiment of the present application, and thus can achieve the beneficial effects that can be achieved by any power system operation safety early warning method based on hybrid XLSTM provided in the embodiment of the present application. For details, please refer to the previous embodiment and will not be repeated here.

[0056] Figure 5 The following figure shows a block diagram of the specific structure of an electronic device provided in an embodiment of the present invention. This electronic device can be used to implement the hybrid XLSTM-based power system operation safety early warning method provided in the above embodiments. The electronic device 500 can be a terminal, server, or other device. The terminal can include a tablet computer, laptop computer, personal computer (PC), microprocessor box, or other device.

[0057] RF circuit 510 is used to receive and transmit electromagnetic waves, converting them into electrical signals, thereby enabling communication with a communications network or other devices. RF circuit 510 may include various existing circuit components for performing these functions, such as an antenna, a radio frequency transceiver, a digital signal processor, an encryption / decryption chip, a subscriber identity module (SIM) card, memory, and the like. RF circuit 510 can communicate with various networks, such as the Internet, an intranet, or a wireless network, or with other devices via a wireless network. These wireless networks may include cellular telephone networks, wireless local area networks, or metropolitan area networks. The wireless networks may utilize various communication standards, protocols, and technologies, including but not limited to Global System for Mobile Communication (GSM), Enhanced Data GSM Environment (EDGE), Wideband Code Division Multiple Access (WCDMA), Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Wireless Fidelity (Wi-Fi) (such as Institute of Electrical and Electronics Engineers standards IEEE 802.11a, IEEE 802.11b, IEEE802.11g, and / or IEEE802.11n), Voice over Internet Protocol (VoIP), Worldwide Interoperability for Microwave Access (Wi-Max), other protocols for email, instant messaging, and short messaging, and any other suitable communication protocols, including those currently undeveloped.

[0058] The memory 520 can be used to store software programs and modules, such as the corresponding program instructions / modules in the above-mentioned embodiments. The processor 580 executes various functional applications and data processing by running the software programs and modules stored in the memory 520, that is, realizing functions such as taking pictures with the front camera, processing the captured images, and switching the display color of the displayed content on the display screen. The memory 520 may include a high-speed random access memory and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 520 may further include a memory remotely located relative to the processor 580, and these remote memories may be connected to the electronic device 500 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0059] The input unit 530 may be used to receive input digital or character information, and generate a keyboard and a mouse related to user settings and function control. The display unit 540 can be used to display information input by the user or information provided to the user, as well as various graphical user interfaces. These graphical user interfaces can be composed of graphics, text, icons, videos, or any combination thereof. The display unit 540 may include a display panel 541. Optionally, the display panel 541 can be configured in the form of an LCD (Liquid Crystal Display), an OLED (Organic Light-Emitting Diode), or the like.

[0060] Audio circuit 560, speaker 561, and microphone 562 provide an audio interface between the user and electronic device 500. Audio circuit 560 converts received audio data into electrical signals and transmits them to speaker 561, which then converts them into sound signals for output. Microphone 562, on the other hand, converts collected sound signals into electrical signals, which are then received by audio circuit 560 and converted into audio data. The audio data is then processed by output processor 580 and transmitted via RF circuit 510 to, for example, another terminal. Alternatively, the audio data may be output to memory 520 for further processing. Audio circuit 560 may also include an earphone jack to allow communication between external headphones and electronic device 500.

[0061] Electronic device 500, through a transmission module 570 (e.g., a Wi-Fi module), can help users receive requests, send information, and so on, providing users with wireless broadband Internet access. Although the figure shows transmission module 570, it is understood that it is not a required component of electronic device 500 and can be omitted as needed without changing the essence of the invention.

[0062] Processor 580 is the control center of electronic device 500. It connects all components of the phone using various interfaces and circuits. By running or executing software programs and / or modules stored in memory 520 and accessing data stored in memory 520, it executes various functions of electronic device 500 and processes data, thereby providing overall monitoring of the electronic device. Optionally, processor 580 may include one or more processing cores. In some embodiments, processor 580 may integrate an application processor and a modem processor. The application processor primarily handles the operating system, user interface, and application programs, while the modem processor primarily handles wireless communications. It is understood that the modem processor may not be integrated into processor 580.

[0063] Electronic device 500 also includes a power supply 590 (e.g., a battery) for powering various components. In some embodiments, the power supply can be logically connected to processor 580 via a power management system, thereby enabling the power management system to manage charging, discharging, and power consumption. Power supply 590 can also include any components, such as one or more DC or AC power supplies, a recharging system, a power failure detection circuit, a power converter or inverter, and a power status indicator.

[0064] Although not shown, the electronic device 500 also includes a camera (such as a front camera and a rear camera), a Bluetooth module, etc., which will not be described in detail here. Specifically, in this embodiment, the display unit of the electronic device is a touch screen display, and the mobile terminal also includes a memory and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by one or more processors. The one or more programs include instructions for performing the following operations: Acquire a characteristic signal sequence of the power system; the characteristic signal sequence includes a voltage signal, a power angle signal, and a phase signal; Filling missing values in the characteristic signal sequence; Construct a topology matrix based on the topology of the power system; The padded feature signal sequence is input into the hybrid XLSTM model for feature extraction to obtain the predicted future feature signal sequence; The predicted future feature sequence and the topological structure matrix are input into a hybrid XLSTM model for classification to obtain classification results, which include risky and non-risky.

[0065] During specific implementation, the above modules can be implemented as independent entities, or can be arbitrarily combined and implemented as the same or several entities. The specific implementation of the above modules can be found in the previous method embodiments and will not be repeated here.

[0066] Those skilled in the art will appreciate that all or part of the steps in the various methods of the above embodiments can be accomplished through instructions, or by controlling related hardware through instructions. The instructions can be stored in a computer-readable storage medium and loaded and executed by a processor. To this end, an embodiment of the present invention provides a storage medium storing a plurality of instructions that can be loaded by a processor to execute the steps of any embodiment of the hybrid XLSTM-based power system operation safety early warning method provided in the embodiment of the present invention.

[0067] The computer-readable storage medium may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0068] Since the instructions stored in the storage medium can execute the steps in any embodiment of the power system operation safety early warning method based on hybrid XLSTM provided in the embodiments of the present invention, the beneficial effects that can be achieved by any power system operation safety early warning method based on hybrid XLSTM provided in the embodiments of the present invention can be achieved. Please see the previous embodiments for details and will not be repeated here.

[0069] The above is a detailed introduction to the electric power system operation safety warning method, device, storage medium and electronic device based on hybrid XLSTM provided in the embodiments of the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, based on the ideas of the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A hybrid XLSTM-based power system operation safety early warning method, characterized in that: The method comprises: Acquire a characteristic signal sequence of the power system; the characteristic signal sequence includes a voltage signal, a power angle signal, and a phase signal; Filling missing values in the characteristic signal sequence; Construct a topology matrix based on the topology of the power system; The padded feature signal sequence is input into the hybrid XLSTM model for feature extraction to obtain the predicted future feature signal sequence; The predicted future feature sequence and the topological structure matrix are input into a hybrid XLSTM model for classification to obtain classification results, which include risky and non-risky.

2. The electric power system operation safety early warning method based on hybrid XLSTM according to claim 1 is characterized in that: Before the step of filling the missing values in the characteristic signal sequence, the method includes: The data in the characteristic signal sequence are analyzed to obtain the distribution characteristics of the data, the pattern of missing data and the correlation between the data.

3. The electric power system operation safety early warning method based on hybrid XLSTM according to claim 2 is characterized in that: Filling the missing values in the characteristic signal sequence includes: Performing frequency domain analysis on sequence data in the characteristic signal sequence by Fourier transform to extract periodic information; Folding the characteristic signal sequence based on the period information to convert it into a two-dimensional tensor, and obtaining a two-dimensional time series change representation based on the two-dimensional tensor; Inputting the two-dimensional time series variation representation into a two-dimensional convolutional neural network for feature extraction to obtain periodic information; Expanding the two-dimensional time series variation representation to obtain a one-dimensional time series, and adaptively fusing the one-dimensional time series and the period to obtain a fused one-dimensional time series; The missing values in the characteristic signal sequence are filled based on the fused one-dimensional time series.

4. The electric power system operation safety early warning method based on hybrid XLSTM according to claim 1 is characterized in that: The padded feature signal sequence is input into the hybrid XLSTM model for feature extraction to obtain the predicted future feature signal sequence, including: Input multiple padded feature signal sequences into the hybrid XLSTM model for feature extraction and then fusion to obtain the first fusion feature; The first fusion feature is input into the hybrid XLSTM model for prediction to obtain a future feature signal sequence.

5. The electric power system operation safety early warning method based on hybrid XLSTM according to claim 1 is characterized in that: The hybrid XLSTM model includes an LSTM layer, an attention enhancement module, a residual connection and a normalization layer; The hollow LSTM layer includes at least one LSTM unit, wherein the LSTM unit is an exponential gated scalar memory unit, or the LSTM unit is a matrix memory unit; An attention enhancement module connected to the atrous LSTM layer; A residual connection and normalization layer, comprising a plurality of residual blocks, each of which is integrated with an xLSTM unit; The fully connected layer is used to map the output of the xLSTM unit to the label space of the classification task and perform multi-category classification through the softmax function.

6. The electric power system operation safety early warning method based on hybrid XLSTM according to claim 5 is characterized in that: The predicted future feature sequence and the topological structure matrix are input into the hybrid XLSTM model for feature extraction and classification, including: In the atrous LSTM layer, convolution is performed on the future feature sequence and the topological structure matrix respectively to obtain future time series features and topological features, and the future time series features and the topological features are fused to obtain a second fused feature; In the attention enhancement module, the attention weights of the historical hidden states are dynamically calculated to generate context-aware sequence representations. The weighted feature vectors are output and concatenated or added with the hidden states of the dilated LSTM layer to obtain weighted features. In the residual connection and normalization layer, the stabilized weighted features are output; In the fully connected layer, the weighted features are classified to obtain a classification result.

7. The electric power system operation safety early warning method based on hybrid XLSTM according to claim 1 is characterized in that: The method further comprises: The characteristic signal sequence is collected through a SCADA system.

8. A hybrid XLSTM-based power system operation safety early warning device, characterized in that: include: A signal acquisition module, used to acquire characteristic signal sequences of the power system; The characteristic signal sequence includes a voltage signal, a power angle signal and a phase signal; A missing value filling module is used to fill the missing values in the characteristic signal sequence; A topology matrix building module is used to build a topology structure matrix based on the topology structure of the power system; The prediction module is used to input the padded feature signal sequence into the hybrid XLSTM model for feature extraction to obtain the predicted future feature signal sequence; The classification module is used to input the predicted future feature sequence and the topological structure matrix into the hybrid XLSTM model for classification to obtain a classification result, which includes risky and non-risky.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a plurality of instructions, which are suitable for being loaded by a processor to execute the electric power system operation safety early warning method based on hybrid XLSTM as described in any one of claims 1 to 7.

10. An electronic device, characterized in that: It includes a processor and a memory, the processor is electrically connected to the memory, the memory is used to store instructions and data, and the processor is used to execute the steps in the power system operation safety early warning method based on hybrid XLSTM as described in any one of claims 1 to 7.

Images