Power grid safety fault voltage drop detector

Abstract

The invention relates to the field of power grid safety monitoring, in particular to a power grid safety fault voltage drop detector. The method comprises the steps of multi-modal sensing unit deployment, adaptive filtering and spectrum feature extraction, amplitude out-of-limit detection and impedance comparison, multi-source data fusion and comprehensive diagnosis, fault area determination and optimal positioning, digital simulation verification and the like. According to the method, the data accuracy in a complex environment can be improved through adaptive filtering and nonlinear transformation technologies, multi-source data are fused to construct an equipment association diagram to realize comprehensive diagnosis, a fault area is accurately positioned by using a consistency algorithm and numerical optimization, and meanwhile, the algorithm reliability is verified through digital simulation, so that the fault diagnosis accuracy is improved. The method improves the efficiency and precision of power grid fault detection, and is higher in adaptability and expansibility.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power system safety monitoring, and particularly relates to a power grid safety fault voltage drop detector. BACKGROUND

[0002] In modern power systems, it is crucial to ensure the safe and stable operation of the power grid. However, traditional voltage drop detection techniques often perform poorly in terms of data accuracy and stability under extreme conditions when faced with complex operating environments. In particular, in high-altitude, strong electromagnetic interference, high-temperature, high-humidity, and other environments, the adaptability of existing technologies is relatively limited, which can significantly affect the detection results. This not only reduces the reliability of fault detection, but also increases the risk of misjudgment, thereby affecting the overall operational efficiency of the power grid.

[0003] With the continuous expansion of the power grid and technological progress, multiple faults may occur simultaneously in actual operation. Existing technologies mostly focus on the study of single faults, and have weak capabilities in handling multiple faults collaboratively. This limitation makes it difficult to quickly and comprehensively identify and locate problems in complex fault scenarios, which can lead to prolonged fault handling time and affect the safety of the power grid and the power consumption experience of users.

[0004] In addition, the power grid is composed of transmission lines and equipment of different voltage levels, while existing technologies are usually designed for specific voltage levels, lacking unified detection and analysis solutions across voltage levels. This design limits its applicability in multi-level power grid structures and makes it difficult to achieve global fault monitoring and analysis. For example, in complex power grids involving multiple voltage levels, traditional methods may not be able to efficiently integrate data from different levels, thereby affecting the overall diagnostic effectiveness.

[0005] In summary, there is a need for an improved power grid safety fault voltage drop detector to address the shortcomings in existing technologies. SUMMARY

[0006] The present application provides a power grid safety fault voltage drop detector to solve the problems raised in the background art.

[0007] To achieve the above-mentioned purposes, the present application adopts the following technical solutions: A power grid safety fault voltage drop detector, the detection method of the power grid safety fault voltage drop detector comprising the following steps: Step S1: Deploy multi-modal sensing units at key nodes of the power grid and construct a distributed communication network architecture to obtain real-time electrical parameter curves and abnormal alarm information from the power grid automation monitoring system and advanced measurement system; Step S2: The collected electrical signal is preprocessed by using an adaptive filtering algorithm, the high-frequency component in the voltage drop signal is separated through a spectrum feature extraction module, the transient feature is extracted by combining a nonlinear transformation technology, and a feature vector is generated; Step S3: Based on the feature vector, the voltage amplitude out-of-limit condition is detected by using a multi-dimensional data analysis method, the line impedance is calculated and compared with a preset threshold range, the fault section is determined, the multi-source data provided by the intelligent terminal and the remote monitoring system are fused, the power grid equipment correlation graph is constructed, and the comprehensive diagnosis result is output; Step S4: The fault area is determined by using a distributed consistency algorithm through the fault indication information of adjacent nodes, the key switch position is labeled combined with the established power grid topology model, the optimal fault point coordinates are solved based on the optimization objective function, and the positioning topology graph is generated by summarizing the results; Step S5: The accuracy of the algorithm is verified and the fault report is generated by injecting a preset fault signal through a digital simulation platform.

[0008] Optionally, the step S2 is implemented by the following way: Step A1: Signal acquisition and preprocessing, the current or voltage signals at both ends of the power grid line are synchronously sampled to obtain an original data set; Step A2: Adaptive filtering processing, the adaptive filtering algorithm is applied to the original data set to remove noise interference and retain effective signals; Step A3: Spectrum feature extraction, the filtered signal is subjected to spectrum analysis, the high-frequency component is extracted, and the corresponding spectrum feature is generated; Step A4: Nonlinear transformation, the spectrum feature is subjected to nonlinear transformation, the transient feature is extracted, and the feature vector is generated; Step A5: According to the change trend of the feature vector and the nonlinear transformation result, it is judged whether there is voltage drop abnormality, if the feature vector exceeds the threshold range of the normal operating state, it is determined that there is voltage drop abnormality.

[0009] Optionally, the spectrum feature extraction formula in the step A3 is:

[0010] In the formula, F(ω) is a spectrum coefficient, k is a conversion factor, s(t) is an original signal, w(τ) is a window function, ω is a frequency parameter, τ is a time offset parameter, and ε is an error term.

[0011] Optionally, the nonlinear transformation in the step A4 is calculated by the following way:

[0012] In the formula, qi is the weight proportion of the i th feature component, Hi is the energy value of the i th feature component, and M is the total number of feature components.

[0013] Optionally, the step S3 is implemented by the following way: Step B1: Data collection, real-time collection of voltage, current phasor data, acquisition of discrete event information such as switch state, protection action signal, etc. Step B2: Amplitude out-of-limit detection, for each line, set the upper and lower limits of the current and voltage, compare the real-time detected current and voltage values with the pre-set upper and lower limits; Step B3: Impedance calculation and comparison, calculate the line impedance using Kirchhoff's law, compare the calculated impedance value with the impedance threshold value in the normal operating range defined in advance, and judge whether it is abnormal; Step B4: Based on the impedance calculation result and the phase angle difference information provided by the intelligent terminal, the fault location is located by calculation, and the specific fault section is determined in combination with the switch state information of the remote monitoring system; Step B5: Integrate the continuous time series data provided by the intelligent terminal with the discontinuous event information provided by the remote monitoring system, process the fused data, and accurately identify the fault; Step B6: Build a power grid equipment association graph, establish the connection relationship graph between the power grid equipment according to the network topology structure, and mark the position and its influence range of the fault in the relationship graph; Step B7: Comprehensive analysis result, generate a report containing fault type, location, severity and recommended operation measures.

[0014] Optionally, the step S4 is implemented by the following way: Step C1: Construct a topology model according to the physical connection relationship of the power grid system; Step C2: Fault indication information analysis, collect fault indication information from each node, and record whether each node detects a fault signal; Step C3: Based on the consensus algorithm of message passing, exchange fault indication information between adjacent nodes, gradually converge to a consistent fault area judgment, and update the state of the node by weighted average or logical operation of the fault state of the neighbor node; Step C4: Label all key switch positions on the topology model, define an optimization objective function, solve the objective function using numerical optimization method, get the optimal fault point coordinates, map the obtained fault point coordinates back to the power grid topology model, mark the specific fault location, use graphical tools to draw the final positioning topology graph, and show the fault point and its influence range; Step C5: Summarize all analysis results, including but not limited to the exact position of the fault point, the affected area, and the recommended repair measures, and generate a detailed report document.

[0015] Optionally, the step S5 is implemented by the following way: Step D1: Establish a simulation model, build a detailed topology model of the power grid in the digital simulation platform, and make the model parameters consistent with the actual system; Step D2: Inject faults and collect data, use the fault injection function of the digital simulation platform to inject a pre-set fault signal into the model at a specified time point, export the time series data of the key electrical quantities from the platform, and record the fault indication information of each node and branch; Step D3: Apply the algorithm for fault location, input the collected data into the fault location algorithm, update the node state based on the consistency algorithm, and gradually converge to a consistent fault area judgment; Step D4: Define a target function to minimize the weighted sum of the distances between the fault point and the fault indication nodes, use an optimization algorithm to solve the optimal fault point coordinates, verify the correctness of the algorithm, and output the fault report.

[0016] A power grid safety fault voltage drop detection system adopts a power grid safety fault voltage drop detector, characterized by comprising a multi-modal sensing unit configuration module, an adaptive filtering and spectral feature extraction module, an amplitude overrun detection and impedance comparison module, a multi-source data fusion and comprehensive diagnosis module, a fault area determination and optimization positioning module, and a digital simulation verification module. The multi-modal sensing unit configuration module is responsible for installing and configuring multi-modal sensing units on key nodes of the power grid and building a distributed communication network architecture. The adaptive filtering and spectral feature extraction module is used for signal collection and preprocessing of the power grid, extracting spectral features and generating feature vectors. The amplitude overrun detection and impedance comparison module is used for detecting current and voltage amplitude overrun and comparing impedance. The multi-source data fusion and comprehensive diagnosis module is used to combine data provided by intelligent terminals and remote monitoring systems, perform multi-source data fusion, build a power grid device correlation graph, and output comprehensive diagnosis results. The fault area determination and optimization positioning module is used to determine the fault area based on the consistency algorithm and optimize the positioning. The digital simulation verification module is used to establish a detailed power grid simulation model, inject faults in the digital simulation platform, and apply the aforementioned algorithm for verification, and output a fault report.

[0017] Optionally, the adaptive filtering and spectral feature extraction module includes a signal acquisition unit, an adaptive filtering unit, and a spectral analysis unit. The signal acquisition unit is used for synchronous sampling of current or voltage signals at both ends of the power grid line. The adaptive filtering unit is used to apply an adaptive filtering algorithm to the original data set to remove noise interference and retain valid signals. a spectrum analysis unit configured to perform spectrum analysis on the filtered signal, extract high-frequency components, and generate corresponding spectrum features.

[0018] Optionally, the fault region determination and optimization positioning module comprises a fault indication information analysis unit, a consistency algorithm unit, and a numerical optimization solving unit. The fault indication information analysis unit is configured to collect fault indication information from each node. The consistency algorithm unit is configured to exchange fault indication information between adjacent nodes and gradually converge to a consistent fault region determination. The numerical optimization solving unit is configured to define an optimization objective function, solve the optimal fault point coordinates using a numerical optimization method, and generate a positioning topology graph.

[0019] Compared with the prior art, the present application has the following advantages: 1. In the present application, the collected electrical signals are preprocessed using an adaptive filtering algorithm, and the feature vectors are generated by combining spectrum feature extraction and nonlinear transformation technology, which can more accurately extract high-frequency components and transient features in the voltage drop signal. This method helps to improve the accuracy and stability of the data in complex operating environments and reduce the risk of misjudgment caused by extreme conditions.

[0020] 2. In the present application, multi-source data provided by intelligent terminals and remote monitoring systems are fused to construct a power grid device association graph. This comprehensive analysis method can more comprehensively and accurately perform fault diagnosis and provide more detailed fault information reports. Based on the consistency algorithm and numerical optimization solving method, the fault region can be more efficiently and accurately determined in complex power grid structures, and the optimal fault point coordinates can be output to generate a positioning topology graph, thereby guiding the on-site staff to take prompt measures.

[0021] 3. In the present application, a detailed power grid model is built in the digital simulation platform and a preset fault signal is injected, which can comprehensively verify the proposed fault detection and diagnosis algorithm and ensure its reliability and effectiveness in actual application. Modular design is adopted, including a multi-modal sensing unit configuration module, an adaptive filtering and spectrum feature extraction module, etc., which makes the entire system have high flexibility and scalability, and can better adapt to changes in power grid structure and new operating conditions.

[0022] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, and to implement the content of the specification, the following will be described in detail with the preferred embodiments of the present application and the accompanying drawings. The specific embodiments of the present application are described in detail by the following examples and their accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0023] The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings: Figure 1 The flowchart of the power grid safety fault voltage drop detector provided by the embodiment of the application.

[0024] Figure 2 The working principle block of the adaptive filtering and spectrum feature extraction module in the embodiment of the application details the process of signal acquisition, filtering processing and spectrum analysis.

[0025] Figure 3 The function structure diagram of the fault area determination and optimization positioning module in the embodiment of the application. DETAILED DESCRIPTION

[0026] The principles and features of the application are described below in conjunction with the accompanying drawings, and the examples are used only to explain the application and not to limit the scope of the application. In the following paragraphs, the application is described in more detail by way of example with reference to the accompanying drawings. The advantages and features of the application will be clearer from the following description and claims. It should be noted that the drawings are very simplified and use non-precise proportions, only to facilitate and clearly assist in explaining the purpose of the embodiment of the application.

[0027] The application provides a power grid safety fault voltage drop detector, and a detection method thereof includes the following steps: A multi-modal sensing unit is installed and configured on a key node of the power grid, and a distributed communication network architecture is constructed to realize real-time transmission of data. The multi-modal sensing unit is connected to current or voltage signal collection points at both ends of the power grid line through a physical connection mode, for synchronous sampling of original electrical signals. These signals are transmitted to an adaptive filtering unit through a cable or a wireless communication module, and the adaptive filtering unit applies an adaptive filtering algorithm to the original data set to remove noise interference and retain effective signals. In order to further improve the signal quality, the LMS (Least Mean Square Error) algorithm is used in the adaptive filtering unit to dynamically adjust the filtering parameters, so that the filtered signals can more accurately reflect the actual power grid operating state. The filtered signals are transmitted to a spectrum analysis unit, which performs spectrum analysis on the signals and extracts high-frequency components to generate corresponding spectrum features. The spectrum feature extraction formula is:

[0028] Wherein, F(w) is a frequency spectrum coefficient, k is a conversion factor, s(t) is an original signal, w(t) is a window function, w is a frequency parameter, t is a time offset parameter, and e is an error term. The transient feature is calculated by performing a nonlinear transformation on the spectral feature to generate a feature vector, and the nonlinear transformation formula is , wherein is a weight proportion of the i-th feature component, is an energy value of the i-th feature component, and M is the total number of feature components.

[0029] The generated feature vector is transmitted to the amplitude out-of-limit detection and impedance comparison module through a communication link. The module is responsible for collecting voltage, current phasor data, and discrete event information such as switch state and protection action signals in real time. For each line, the upper and lower limits of the current and voltage are set, and the real-time detected current and voltage values are compared with the preset upper and lower limits. If an out-of-limit condition is detected, the line impedance is calculated using Kirchhoff's law, and the calculated impedance value is compared with the impedance threshold value in the normal operating range to determine whether it is abnormal. In this process, the phase angle difference information provided by the intelligent terminal and the remote monitoring system is introduced as an auxiliary judgment basis, and the switch state information is combined to further determine the specific fault section. Subsequently, the continuous time series data provided by the intelligent terminal and the intermittent event information provided by the remote monitoring system are integrated, and the fused data is processed to accurately identify the fault. Based on the analysis results of the integrated data, a power grid device correlation graph is constructed, and the location of the fault and its impact range are marked in the relationship graph. Finally, a comprehensive diagnostic report containing the fault type, location, severity, and recommended operation measures is generated.

[0030] In the fault area determination and optimization positioning module, a topology model is first constructed according to the physical connection relationship of the power grid system and all key switch positions are labeled. The fault indication information analysis unit is responsible for collecting fault indication information from each node, and records whether each node detects a fault signal. The consistency algorithm unit gradually converges to a consistent fault area judgment by exchanging fault indication information between adjacent nodes. Specifically, adjacent nodes update their own states through weighted averaging or logical operations until consistency is achieved. The numerical optimization solving unit defines an optimization objective function and uses numerical optimization methods to solve the objective function to obtain the optimal fault point coordinates. The design of the objective function aims to minimize the weighted sum of the distances between the fault point and the fault indication nodes, thereby ensuring the accuracy of the fault point coordinates. The obtained fault point coordinates are mapped back to the power grid topology model to mark the specific fault location, and a graphical tool is used to draw the final positioning topology graph to display the fault point and its impact range. All analysis results including but not limited to the exact location of the fault point, the affected area, and the recommended repair measures are summarized and a detailed report document is generated.

[0031] A digital simulation verification module is used to further verify the effectiveness and reliability of the above method. A detailed topology model of the power grid is built in the digital simulation platform to make the model parameters consistent with the actual system. Through the fault injection function of the digital simulation platform, a preset fault signal is injected into the model at a specified time point, and the time series data of the key electrical quantities are exported to record the fault indication information of each node and branch. The collected data is input into the fault location algorithm, and the node state is updated based on the consistency algorithm to gradually converge to a consistent fault area judgment. A target function is defined to minimize the weighted sum of the distance between the fault point and the fault indication node, and an optimization algorithm is used to solve the optimal fault point coordinates to verify the correctness of the algorithm and output the fault report.

[0032] As can be seen from the above embodiments, the power grid safety fault voltage drop detector and system provided by the present application has the following characteristics. First, in the signal acquisition and preprocessing stage, through the cooperation of the multi-modal sensing unit and the adaptive filtering unit, efficient filtering processing of the original signal is realized, and through the frequency spectrum analysis unit, high-frequency components and transient characteristics are extracted to form a complete feature vector. Secondly, in the fault diagnosis and positioning stage, through the combination of amplitude overrun detection and impedance ratio module and multi-source data fusion technology, accurate positioning of the fault is realized, and through the construction of the power grid equipment correlation graph, comprehensive fault information is provided. Finally, in the fault area determination and optimization positioning module, through the cooperative work of the consistency algorithm unit and the numerical optimization solving unit, the coordinates of the fault point are accurately solved, and the reliability and effectiveness of the algorithm are further verified through the digital simulation verification module.

[0033] In order to better enable relevant persons in the technical field to fully understand and implement the present application, the specific implementation principles of the present application are further described below in conjunction with a specific application scenario.

[0034] In the actual operation scenario of the power grid, first, multi-modal sensing units are installed on key nodes, including substation outlets, power line branch points, and important load access points. The signal acquisition end of the multi-modal sensing unit is connected to the current or voltage signal acquisition points at both ends of the power line for synchronous sampling of the original electrical signal. These signals are transmitted to the adaptive filtering unit through a cable or a wireless communication module. The adaptive filtering unit uses the LMS algorithm to dynamically filter the original data set, the core of which is to remove noise interference and retain effective signals by adjusting the filter parameters. For example, in a high electromagnetic interference environment, the LMS algorithm can dynamically adjust the filter coefficients according to the real-time signal characteristics, thereby ensuring the quality of the output signal. The filtered signal is transmitted to the frequency spectrum analysis unit, which extracts high-frequency components based on the formula and generates corresponding spectral features. In this process, the choice of the window function w(t) directly affects the spectral resolution, and through the nonlinear transformation formula Further transient features are extracted to form a complete feature vector.

[0035] Subsequently, the generated feature vector is transmitted to the amplitude out-of-limit detection and impedance comparison module. This module collects voltage, current phasor data and switch state information from the smart terminal and remote monitoring system in real time. For each line, the upper and lower limits of the current and voltage are set, and whether the detected values exceed the preset range is compared. If an out-of-limit condition is detected, the line impedance is calculated using Kirchhoff's law, and the calculation result is compared with the impedance threshold value in the normal operation range defined in advance. In this process, the phase angle difference information provided by the smart terminal is used as an auxiliary judgment basis, and the switch state information is further used to determine the specific fault section. For example, when the impedance value of a certain line deviates significantly from the normal range and the phase angle difference is abnormal, the system will determine that this line is a potential fault area. Subsequently, the continuous time series data and the intermittent event information are integrated, and the fault is accurately identified through multi-source data fusion technology. Based on the integrated data analysis results, a power grid equipment correlation diagram is constructed, and the location of the fault and its impact range are marked in the relationship diagram, and finally a comprehensive diagnostic report containing the fault type, location, severity and recommended operation measures is generated.

[0036] In the fault area determination and optimization positioning module, first, a topological model is constructed according to the physical connection relationship of the power grid system, and all key switch positions are marked. The fault indication information analysis unit 4 is responsible for collecting fault indication information from each node, recording whether each node detects a fault signal. The consistency algorithm unit gradually converges to a consistent fault area judgment by exchanging fault indication information between adjacent nodes. For example, adjacent nodes update their own state through weighted averaging until consistency is achieved. The numerical optimization solving unit defines an objective function, aiming to minimize the weighted sum of the distance between the fault point and the fault indication node. The optimal fault point coordinates are obtained by solving the objective function through numerical optimization method. The obtained fault point coordinates are mapped back to the power grid topological model, and the specific fault location is marked, and a graphical tool is used to draw the final positioning topological graph, showing the fault point and its impact range. All analysis results are summarized to generate a detailed report document.

[0037] The digital simulation verification module is used for further verifying the effectiveness and reliability of the above method. A detailed topology model of the power grid is built in the digital simulation platform, so that the model parameters are consistent with the actual system. Through the fault injection function of the digital simulation platform, a preset fault signal is injected into the model at a specified time point, and the time series data of the key electrical quantities are exported, and the fault indication information of each node and branch is recorded. The collected data is input into the fault location algorithm, the node state is updated based on the consistency algorithm, and the fault area is gradually converged to the consistent fault area. The objective function is defined to minimize the weighted sum of the distance between the fault point and the fault indication node, and the optimal fault point coordinates are solved by using the optimization algorithm, the correctness of the algorithm is verified, and the fault report is output.

[0038] As can be seen from the above steps, in the signal acquisition stage, efficient filtering processing is realized by cooperation of the multi-modal sensing unit and the adaptive filtering unit, and high-frequency components and transient characteristics are extracted by the frequency spectrum analysis unit to form a complete feature vector. In the fault diagnosis and positioning stage, accurate positioning is realized by combining the amplitude out-of-limit detection and impedance comparison module with multi-source data fusion technology, and comprehensive fault information is provided by constructing the power grid equipment correlation graph. Finally, in the fault area determination and optimization positioning module, the fault point coordinates are accurately solved by the cooperative work of the consistency algorithm unit and the numerical optimization solving unit, and the reliability and effectiveness of the algorithm are further verified by the digital simulation verification module.

[0039] The above is only a preferred embodiment of the present application, and does not limit the present application in any form; any person skilled in the art can easily implement the present application according to the drawings and the above description; however, any equivalent changes, modifications and evolution made by those skilled in the art without departing from the scope of the technical solution of the present application, using the above disclosed technical content, are equivalent embodiments of the present application; at the same time, any equivalent changes, modifications and evolution of the above embodiments according to the essential technology of the present application, are still within the protection scope of the technical solution of the present application.

Claims

1. A voltage drop detector for power grid safety faults, characterized in that, The detection method of this power grid safety fault voltage drop detector includes the following steps: Step S1: Deploy multimodal sensing units at key nodes of the power grid and build a distributed communication network architecture to obtain real-time electrical parameter curves and abnormal alarm information from the power grid automation monitoring system and advanced measurement system; Step S2: The acquired electrical signal is preprocessed using an adaptive filtering algorithm. The high-frequency components in the voltage drop signal are separated by the spectrum feature extraction module. Transient features are extracted by combining nonlinear transformation technology to generate feature vectors. Step S3: Based on feature vectors, use multi-dimensional data analysis methods to detect voltage amplitude exceeding limits, calculate line impedance and compare it with preset threshold range to determine fault sections, integrate multi-source data provided by intelligent terminals and remote monitoring systems, construct a power grid equipment correlation diagram, and output comprehensive diagnostic results; Step S4: Using the fault indication information of adjacent nodes, the fault area is determined by the distributed consensus algorithm. Combined with the established power grid topology model, the location of key switches is marked. The optimal fault point coordinates are solved based on the optimization objective function. The results are summarized to generate a location topology map. Step S5: Inject a preset fault signal through a digital simulation platform to verify the accuracy of the algorithm and generate a fault report.

2. The power grid safety fault voltage drop detector according to claim 1, characterized in that, Step S2 is implemented in the following manner: Step A1: Signal acquisition and preprocessing. The current or voltage signals at both ends of the power grid line are sampled synchronously to obtain the raw dataset. Step A2: Adaptive filtering process, applying an adaptive filtering algorithm to the original dataset to remove noise interference and retain the effective signal; Step A3: Spectral feature extraction. Perform spectral analysis on the filtered signal to extract high-frequency components and generate corresponding spectral features. Step A4: Nonlinear transformation, perform nonlinear transformation on the spectral features, extract transient features and generate feature vectors; Step A5: Based on the changing trend of the feature vector and the result of the nonlinear transformation, determine whether there is an abnormal voltage drop. If the feature vector exceeds the threshold range of normal operation, it is determined to be an abnormal voltage drop.

3. The power grid safety fault voltage drop detector according to claim 2, characterized in that, The formula for extracting spectral features in step A3 is as follows: ； In the formula, F(ω) is the spectral coefficient, k is the conversion factor, s(t) is the original signal, w(t) is the window function, ω is the frequency parameter, τ is the time offset parameter, and ε is the error term.

4. The power grid safety fault voltage drop detector according to claim 2, characterized in that, The nonlinear transformation in step A4 is calculated in the following way: In the formula, The weight percentage of the i-th feature component. Let be the energy value of the i-th feature component, and M be the total number of feature components.

5. The power grid safety fault voltage drop detector according to claim 1, characterized in that, Step S3 is implemented in the following manner: Step B1: Data acquisition, real-time collection of voltage and current phasor data, and acquisition of discrete event information such as switch status and protection action signals; Step B2: Amplitude over-limit detection. For each line, set safe upper and lower limits for current and voltage, and compare whether the real-time detected current and voltage values ​​exceed the preset upper and lower limits. Step B3: Impedance calculation and comparison. Kirchhoff's laws are used to calculate the line impedance. The calculated impedance value is compared with the impedance threshold within the predefined normal operating range to determine if there is any abnormality. Step B4: Based on the impedance calculation results and the phase angle difference information provided by the smart terminal, the fault location is located by calculation, and the specific fault section is determined by combining the switch status information of the remote monitoring system. Step B5: Integrate the continuous time series data provided by the smart terminal with the intermittent event information provided by the remote monitoring system, process the fused data, and accurately identify the fault; Step B6: Construct a power grid equipment interconnection diagram. Based on the network topology, establish a connection relationship diagram between power grid equipment and mark the location of the fault and its impact range in the diagram. Step B7: Analyze the results and generate a report that includes the fault type, location, severity, and recommended actions.

6. The power grid safety fault voltage drop detector according to claim 1, characterized in that, Step S4 is implemented in the following manner: Step C1: Construct a topology model based on the physical connections of the power grid system; Step C2: Fault indication information analysis. Collect fault indication information from each node. For each node, record whether it has detected a fault signal. Step C3: Based on the message-passing consensus algorithm, neighboring nodes exchange fault indication information and gradually converge to a consistent fault area judgment. The weighted average or logical operation of the fault status of neighboring nodes is used to update the state of the node itself. Step C4: Mark all key switch locations on the topology model, define an optimization objective function, solve the objective function using numerical optimization methods to obtain the optimal fault point coordinates, map the obtained fault point coordinates back to the power grid topology model, mark the specific fault location, and use graphical tools to draw the final location topology map to show the fault point and its impact range. Step C5: Summarize all analysis results, including but not limited to the exact location of the fault, the affected area, and recommended remedial measures, and generate a detailed report document.

7. The power grid safety fault voltage drop detector according to claim 1, characterized in that, Step S5 is implemented in the following manner: Step D1: Establish a simulation model. Build a detailed topology model of the power grid in a digital simulation platform, ensuring that the model parameters are consistent with the actual system. Step D2: Inject faults and collect data. Use the fault injection function of the digital simulation platform to inject preset fault signals into the model at specified time points, export the time series data of key electrical quantities from the platform, and record the fault indication information of each node and branch. Step D3: Apply the algorithm to locate the fault. Take the collected data as input and pass it to the fault location algorithm. Update the node status based on the consensus algorithm and gradually converge to a consistent fault area judgment. Step D4: Define an objective function to minimize the weighted sum of distances between the fault point and the fault indication node, use an optimization algorithm to solve for the optimal fault point coordinates, verify the correctness of the algorithm, and output a fault report.

8. A power grid safety fault voltage drop detection system, employing the power grid safety fault voltage drop detector as described in any one of claims 1 to 6, characterized in that, It includes a multimodal sensing unit configuration module, an adaptive filtering and spectral feature extraction module, an amplitude over-limit detection and impedance comparison module, a multi-source data fusion and comprehensive diagnosis module, a fault area determination and optimized location module, and a digital simulation verification module; The multimodal sensing unit configuration module is responsible for installing and configuring multimodal sensing units at key nodes of the power grid and building a distributed communication network architecture. The adaptive filtering and spectral feature extraction module is used to acquire and preprocess signals from the power grid, extract spectral features, and generate feature vectors. The amplitude over-limit detection and impedance comparison module is used to detect and compare the impedance of current and voltage amplitude over-limits. The multi-source data fusion and comprehensive diagnosis module is used to combine data provided by smart terminals and remote monitoring systems to perform multi-source data fusion, construct a power grid equipment correlation diagram, and output comprehensive diagnostic results. The fault area determination and optimization positioning module is used to determine the fault area based on a consensus algorithm and optimize its positioning. The digital simulation verification module is used to establish a detailed power grid simulation model, inject faults into the digital simulation platform, apply the aforementioned algorithms for verification, and output a fault report.

9. The power grid safety fault voltage drop detection system according to claim 8, characterized in that, The adaptive filtering and spectral feature extraction module includes a signal acquisition unit, an adaptive filtering unit, and a spectral analysis unit; The signal acquisition unit is used to synchronously sample the current or voltage signals at both ends of the power grid line; The adaptive filtering unit is used to apply an adaptive filtering algorithm to the original dataset to remove noise interference and retain the effective signal. The spectrum analysis unit is used to perform spectrum analysis on the filtered signal, extract high-frequency components, and generate corresponding spectrum features.

10. The power grid safety fault voltage drop detection system according to claim 8, characterized in that, The fault area determination and optimization positioning module includes a fault indication information analysis unit, a consistency algorithm unit, and a numerical optimization solution unit. The fault indication information analysis unit is used to collect fault indication information from various nodes; The consensus algorithm unit is used to gradually converge to a consistent fault region judgment by exchanging fault indication information between adjacent nodes. The numerical optimization solution unit is used to define the optimization objective function, solve for the optimal fault point coordinates using numerical optimization methods, and generate a location topology map.

Images