DC power supply over-voltage and under-voltage diagnosis method and system based on mutation data analysis

Abstract

The invention provides a direct current power supply over-voltage and under-voltage diagnosis method and system based on mutation data analysis, and relates to the field of direct current power supply monitoring. The method collects voltage signal data of a direct current power supply in real time through a sensing acquisition device, sets a data instantaneous acquisition frequency and records the acquisition time, and obtains a real-time change trend of the voltage signal data; extracts characteristic parameters of mutation amplitude, duration, frequency component and energy distribution from mutation detection results, inputs the extracted characteristic parameters into a pre-trained fault type recognition model, and recognizes the fault type in the voltage signal; analyzes the fault diagnosis result based on the real-time change trend, the recognized fault type and the characteristic parameter, establishes a real-time monitoring system for uninterrupted monitoring of the direct current power supply, sets a multi-level alarm mechanism, triggers an alarm according to the fault diagnosis result, and can quickly and accurately handle overvoltage and undervoltage faults of the direct current power supply system.

Description

Technical Field

[0001] The present invention relates to the field of direct current power supply monitoring, and in particular to a direct current power supply overvoltage and undervoltage diagnosis method and system based on mutation data analysis. Background Art

[0002] In modern society, the widespread application of electronic equipment and systems has put forward higher and higher requirements on the stability and reliability of power supply. As the core part of many key equipment and applications, the voltage stability of DC power supply system is directly related to the normal operation and life of the equipment. However, in the actual operation process, DC power supply system often faces various voltage abnormality problems, such as overvoltage and undervoltage problems. Among them, the overvoltage phenomenon caused by overvoltage problem refers to the situation where the power supply voltage exceeds the rated value. Overvoltage may be caused by power grid fluctuations, control system failure or power overload. Excessive voltage can damage electronic equipment, shorten its life, and even cause instantaneous damage to the equipment. Although traditional overvoltage protection devices can cut off the power supply when the voltage is abnormal, they often cannot detect and record the specific situation and cause of overvoltage in time. Among them, undervoltage caused by undervoltage problem refers to the phenomenon that the power supply voltage is lower than the rated value, which is usually caused by factors such as power grid voltage fluctuations, transmission line losses or insufficient power output capacity. Continuous undervoltage can cause the equipment to fail to work properly or even shut down. The severity of undervoltage is usually related to its duration. Short-term undervoltage may have little effect on the equipment, but long-term undervoltage can cause equipment performance degradation or damage.

[0003] The traditional DC power supply fault diagnosis method mainly relies on manual detection and simple protection devices, which have the disadvantages of delayed detection, inaccurate diagnosis and lack of comprehensive analysis. The traditional DC power supply fault diagnosis method can often only detect the fault after it occurs, and cannot achieve real-time monitoring and rapid response; due to the lack of detailed analysis of voltage signals, it is difficult for traditional methods to accurately determine the fault type and specific cause, and usually only detects a single fault type (such as overvoltage or undervoltage), and cannot comprehensively analyze multiple fault characteristics.

[0004] With the development of big data and artificial intelligence technology, fault diagnosis methods based on data analysis have gradually attracted attention. These methods can more accurately identify fault characteristics and causes through comprehensive analysis of large amounts of historical data and real-time data. However, existing data analysis methods still face challenges when processing mutant data, especially in terms of real-time performance and accuracy. Summary of the invention

[0005] In view of this, in order to solve the problems of delayed detection, inaccurate diagnosis and lack of comprehensive analysis capabilities in the existing DC power supply system when facing overvoltage and undervoltage faults, the purpose of the present invention is to propose a DC power supply over- and undervoltage diagnosis method and system based on mutation data analysis, which is used to quickly and accurately handle overvoltage and undervoltage faults of the DC power supply system. By performing multi-dimensional mutation detection and comprehensive analysis on the voltage signal, the overvoltage and undervoltage faults in the DC power supply system can be quickly and accurately diagnosed. The present invention not only improves the real-time and accuracy of fault detection, but also can effectively distinguish different types of faults, providing a solid guarantee for the stable operation of the power supply system.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] Based on the above purpose, in a first aspect, the present invention provides a DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis, comprising the following steps:

[0008] The voltage signal data of the DC power supply is collected in real time through the sensor collection equipment, the instantaneous data collection frequency is set and the collection time is recorded to obtain the real-time change trend of the voltage signal data;

[0009] After preprocessing the voltage signal data, multi-dimensional mutation detection is performed to analyze the time domain, frequency domain and time-frequency domain characteristics of the voltage signal, identify the mutation points and abnormal fluctuations in the voltage signal, and obtain the mutation detection results;

[0010] Extract characteristic parameters of mutation amplitude, duration, frequency component and energy distribution from mutation detection results, input the extracted characteristic parameters into the pre-trained fault type recognition model to identify the fault type in the voltage signal;

[0011] Analyze fault diagnosis results based on real-time change trends, identified fault types and characteristic parameters, including fault occurrence time, duration, fault location and fault cause;

[0012] Establish a real-time monitoring system for uninterrupted monitoring of DC power supply, update the data and analysis results of voltage signals in real time, set up a multi-level alarm mechanism, and trigger alarms based on fault diagnosis results.

[0013] As a further solution of the present invention, obtaining the real-time change trend of the voltage signal data includes the following steps:

[0014] Set the sampling frequency, where the sampling frequency is set to at least twice the highest frequency component of the voltage signal, that is:

[0015]

[0016] In the formula, is the sampling frequency; is the highest frequency component of the voltage signal;

[0017] Use NTP server for time calibration and clock synchronization. Record the current timestamp in real time every time voltage signal data is collected.

[0018] Use Apache Flink for data stream processing, obtain and process voltage signal data according to the current timestamp, and use sliding windows to perform trend analysis on the real-time collected voltage signal data, where the instantaneous change trend is calculated based on the data points in the sliding window;

[0019] The obtained real-time voltage signal data is plotted into a graph to show the trend of voltage change.

[0020] As a further solution of the present invention, multi-dimensional mutation detection is performed on the voltage signal data after preprocessing, including the following steps:

[0021] The Daubechies wavelet basis function is selected to perform wavelet decomposition on the preprocessed voltage signal data to obtain wavelet coefficients at different scales, wherein the wavelet coefficients include approximate coefficients and detail coefficients, the approximate coefficients are the low-frequency part, and the detail coefficients are the high-frequency part;

[0022] Search for abnormal coefficient values ​​in the wavelet coefficients of each scale. The abnormal coefficient values ​​correspond to the mutation points of the voltage signal. The positions of the signal mutation points are located according to the set threshold and through the time series distribution.

[0023] The inverse wavelet transform is performed using the detail coefficients and approximate coefficients that retain the mutation point information, the signal is reconstructed, the reconstructed signal is analyzed, and the mutation point is verified.

[0024] As a further solution of the present invention, short-time Fourier transform (STFT) is used to verify the mutation point. When verifying the mutation point, the following steps are included:

[0025] Select the Hanning window, set the window overlap rate to 70%, and select the window length L to calculate the short-time Fourier transform;

[0026] The STFT calculation is performed on the preprocessed voltage signal, and the calculation formula is:

[0027]

[0028] in, is a voltage signal; It is the function window; is the window length; the overlap rate is 70%, that is, the overlap length is ;

[0029] Generate a time-frequency matrix based on the STFT calculation results, where each column corresponds to the spectrum information in a time window;

[0030] Use Matplotlib to draw the time-frequency diagram of the voltage signal, where the horizontal axis is time and the vertical axis is frequency. The color / intensity represents the amplitude of the frequency component. The mutation points are identified in the time-frequency diagram. According to the mutation characteristics in the time-frequency diagram, the time position of the mutation point is located, and the position of the mutation point identified in the Daubechies wavelet basis function is compared and verified.

[0031] As a further solution of the present invention, multi-dimensional mutation detection is performed on the voltage signal, time domain, frequency domain and time-frequency domain characteristics of the voltage signal are analyzed, and mutation points and abnormal fluctuations are identified, including:

[0032] (1) Time domain analysis:

[0033] Use the sliding window to calculate the standard deviation of the signal within the sliding window and detect the mutation point;

[0034] At each identified mutation point, the voltage difference before and after the mutation is calculated to obtain the mutation amplitude:

[0035] Mutation amplitude = |voltage value (after mutation) - voltage value (before mutation)|;

[0036] (2) Frequency domain analysis:

[0037] Convert the signal from the time domain to the frequency domain through Fourier transform (FFT) to analyze the frequency components;

[0038] Calculate the power spectral density of the signal and identify the frequency components and their intensities;

[0039] Analyze the changes in frequency domain characteristics to determine whether there are sudden changes or abnormal fluctuations in frequency components;

[0040] (3) Time-frequency domain analysis:

[0041] Calculate the time-frequency diagram of the signal using the short-time Fourier transform (STFT);

[0042] Perform wavelet transform on the signal to analyze the changing characteristics at different scales;

[0043] Combine the time domain, frequency domain and time-frequency domain features to identify mutation points and abnormal fluctuation locations.

[0044] As a further solution of the present invention, characteristic parameters of mutation amplitude, duration, frequency component and energy distribution are extracted from the mutation detection results, including the following steps:

[0045] (1) Mutation point detection and mutation amplitude calculation:

[0046] Use the sliding window to calculate the standard deviation of the signal within the sliding window and identify the mutation point of the voltage signal;

[0047] At each identified mutation point, the difference in voltage value before and after the mutation is calculated to obtain the mutation amplitude: mutation amplitude = |voltage value (after mutation) - voltage value (before mutation)|;

[0048] (2) Extraction duration:

[0049] Determine the timestamps before and after the mutation point based on time domain analysis;

[0050] Identify the time period when the voltage signal is stable before and after the mutation and obtain the duration:

[0051] Duration = mutation end time - mutation start time;

[0052] (3) Frequency component extraction:

[0053] Through frequency domain and time-frequency domain analysis, the frequency spectrum information is obtained by Fourier transform (FFT) of a voltage signal data before and after the mutation;

[0054] Identify the main frequency components through the spectrum diagram;

[0055] Perform STFT on the entire voltage signal to generate a time-frequency matrix, extract the spectrum information at the mutation point from the time-frequency matrix, identify the frequency components, and extract the frequency components;

[0056] (4) Energy distribution extraction:

[0057] The power spectrum density of the voltage signal before and after the mutation point is calculated and the energy distribution is extracted.

[0058] As a further solution of the present invention, the fault diagnosis result includes the fault occurrence time, duration, fault location and fault cause, wherein:

[0059] When determining the time of fault occurrence, determine the timestamp of the mutation point as the time of fault occurrence. If there are multiple mutation points, record the timestamp of each mutation point.

[0060] When determining the fault duration, the time period during which the voltage signal is stable before and after each mutation point is calculated to determine the fault duration;

[0061] When locating a fault, the fault location is located by combining the grid topology and characteristic parameters of the fault voltage signal;

[0062] When inferring the cause of a fault, the cause of the fault is inferred based on the fault type, spectrum characteristics, and energy distribution.

[0063] As a further solution of the present invention, when a real-time monitoring system is established, the real-time monitoring system includes:

[0064] Sensor layer: Install sensors for measuring voltage and current parameters at key nodes of the DC power supply and set the data collection frequency;

[0065] Data acquisition layer: Use data acquisition equipment to obtain voltage signal data from sensors;

[0066] Data transmission layer: used to transmit the voltage signal data acquired by the data acquisition layer to the monitoring center;

[0067] Data storage and processing layer: Build servers to perform data storage, processing and analysis, perform time domain, frequency domain and time-frequency domain feature analysis and fault diagnosis on data; locate the fault location based on the grid topology and fault feature parameters, and analyze to obtain fault diagnosis results;

[0068] User interface layer: Design a human-machine interface to display monitoring data and alarm information in real time.

[0069] In a second aspect, the present invention provides a DC power supply over-voltage and under-voltage diagnosis system based on mutation data analysis, comprising:

[0070] Sensor acquisition equipment: used to collect voltage signal data of DC power supply in real time, set the instantaneous data acquisition frequency and record the acquisition time, and obtain the real-time change trend of voltage signal data;

[0071] Multidimensional mutation detection module: performs multidimensional mutation detection on voltage signal data after preprocessing, analyzes the time domain, frequency domain and time-frequency domain characteristics of the voltage signal, and identifies mutation points and abnormal fluctuations in the voltage signal;

[0072] Feature extraction module: used to extract feature parameters from mutation detection results, including mutation amplitude, duration, frequency component and energy distribution;

[0073] Fault type identification module: used to input the extracted characteristic parameters into the pre-trained fault type identification model to identify the fault type in the voltage signal;

[0074] Fault diagnosis module: Based on real-time change trends, identified fault types and characteristic parameters, analyze fault diagnosis results, including fault occurrence time, duration, fault location and fault cause;

[0075] Real-time monitoring module: used to establish a real-time monitoring system for uninterrupted monitoring of DC power supply, and update the data and analysis results of voltage signals in real time;

[0076] Alarm mechanism module: set up a multi-level alarm mechanism to trigger the corresponding level of alarm according to the fault diagnosis results;

[0077] Data storage module: used to store all collected voltage signal data, processing results and alarm records;

[0078] User interface module: provides a user interface for real-time display of voltage signal data, analysis results and alarm information.

[0079] In another aspect of the present invention, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, any one of the above-mentioned DC power supply over-voltage and under-voltage diagnosis methods based on mutation data analysis according to the present invention is executed.

[0080] In another aspect of the present invention, a computer-readable storage medium is provided, storing computer program instructions, which, when executed, implement any of the above-mentioned DC power supply over-voltage and under-voltage diagnosis methods based on mutation data analysis according to the present invention.

[0081] Compared with the prior art, the DC power supply over-voltage and under-voltage diagnosis method and system based on mutation data analysis proposed in the present invention has the following beneficial effects:

[0082] 1. High real-time performance and accuracy: The voltage signal data of the DC power supply is collected in real time through the sensor collection equipment, and the NTP server is used for time calibration and clock synchronization to ensure the time accuracy of data collection; Apache Flink is used for data stream processing, and the sliding window technology is used to perform trend analysis on the real-time collected voltage signal data, which can timely discover the real-time change trend of the voltage.

[0083] 2. Multi-dimensional mutation detection: Combine the time domain, frequency domain and time-frequency domain features to comprehensively analyze the voltage signal and identify the mutation points and abnormal fluctuations in the voltage signal; use wavelet transform (Daubechies wavelet basis function) and short-time Fourier transform (STFT) for multi-dimensional mutation detection to ensure the accuracy and precision of mutation point identification.

[0084] 3. Feature extraction and fault identification: Extract characteristic parameters such as mutation amplitude, duration, frequency component and energy distribution from the mutation detection results to comprehensively describe the changing characteristics of the voltage signal; use the pre-trained fault type identification model to identify the fault type in the voltage signal based on the extracted characteristic parameters to improve the accuracy of fault identification.

[0085] 4. Fault diagnosis and alarm mechanism: Based on real-time change trends, identified fault types and characteristic parameters, analyze the fault diagnosis results, including the time of occurrence, duration, location and cause of the fault, and provide a comprehensive fault diagnosis report; set up a multi-level alarm mechanism to trigger the corresponding level of alarm according to the fault diagnosis results to ensure timely response and handling of various fault situations.

[0086] 5. Scalability and ease of use: By establishing a real-time monitoring system for uninterrupted monitoring of the DC power supply, the data and analysis results of the voltage signal are updated in real time to ensure the real-time and continuity of the system; a user interface module is provided to display the data, analysis results and alarm information of the voltage signal in real time, so that users can monitor and manage conveniently.

[0087] In summary, the present invention realizes accurate diagnosis of overvoltage and undervoltage of DC power supply through multi-dimensional mutation detection and feature extraction combined with a pre-trained fault type recognition model, and ensures the practicality and efficiency of the system through real-time monitoring and multi-level alarm mechanism. It can be widely used in power systems, industrial automation and other fields.

[0088] These and other aspects of the present application will be more concise and understandable in the following description of the embodiments. It should be understood that the above general description and the following detailed description are only exemplary and explanatory and cannot limit the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or related technologies, the following briefly introduces the drawings required for use in the exemplary embodiments or related technical descriptions. The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation to the present invention. In the drawings:

[0090] Figure 1 The present invention is a flowchart of a DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis according to an embodiment of the present invention.

[0091] Figure 2 The present invention is a flowchart of obtaining the real-time change trend of voltage signal data in the DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis in an embodiment of the present invention.

[0092] Figure 3 The present invention is a flowchart of multi-dimensional mutation detection in a DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis in an embodiment of the present invention.

[0093] Figure 4 The present invention is a flowchart for verifying mutation points in a DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis according to an embodiment of the present invention. DETAILED DESCRIPTION

[0094] Below, the present application is further described in conjunction with the accompanying drawings and specific implementation methods. It should be noted that, under the premise of no conflict, the various embodiments or technical features described below can be arbitrarily combined to form a new embodiment.

[0095] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the embodiments of the present invention are further described in detail below in combination with specific embodiments and with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0096] It should be noted that all expressions using "first" and "second" in the embodiments of the present invention are intended to distinguish two non-identical entities or non-identical parameters with the same name. It can be seen that "first" and "second" are only for the convenience of expression and should not be understood as limitations on the embodiments of the present invention. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, other steps or units inherent to a process, method, system, product or device that includes a series of steps or units.

[0097] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0098] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0099] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0100] In view of the problems of detection lag, inaccurate diagnosis and lack of comprehensive analysis capabilities when facing overvoltage and undervoltage faults in existing DC power supply systems, the present invention proposes a DC power supply over- and undervoltage diagnosis method and system based on mutation data analysis. By performing multi-dimensional mutation detection and comprehensive analysis on voltage signals, overvoltage and undervoltage faults in DC power supply systems can be diagnosed quickly and accurately. The present invention not only improves the real-time and accuracy of fault detection, but also can effectively distinguish different types of faults, providing a solid guarantee for the stable operation of the power supply system.

[0101] See also Figure 1 As shown, an embodiment of the present invention provides a DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis, the method comprising the following steps:

[0102] Step S10: collect voltage signal data of the DC power supply in real time through a sensor collection device, set the instantaneous data collection frequency and record the collection time, and obtain the real-time change trend of the voltage signal data.

[0103] In this step, when collecting data, the voltage signal data can be collected in real time at the DC power supply output port through a high-precision voltage sensor, the instantaneous data collection frequency can be set to 1000Hz, and the timestamp of each collection is recorded to ensure that the data has high time resolution and synchronization. The collected voltage signal data is transmitted to the central data processing unit in real time, and is preliminarily recorded and stored. This step ensures high-precision and high-frequency collection of voltage signals.

[0104] Step S20: perform multi-dimensional mutation detection on the voltage signal data after preprocessing, analyze the time domain, frequency domain and time-frequency domain characteristics of the voltage signal, identify the mutation points and abnormal fluctuations in the voltage signal, and obtain the mutation detection result.

[0105] In this step, when performing multi-dimensional mutation detection, the collected voltage signal data is first pre-processed, including filtering to remove high-frequency noise and low-frequency interference to obtain a clean signal. Then, multi-dimensional mutation detection is performed, including time domain analysis, frequency domain analysis, and time-frequency domain analysis. In the time domain analysis, a sliding window is used to calculate the time domain characteristics of the voltage signal to identify the mutation point of the voltage signal; in the frequency domain analysis, the fast Fourier transform (FFT) is used to analyze the spectrum of the voltage signal, identify the changes in frequency components, and detect abnormal frequency components; in the time-frequency domain analysis, short-time Fourier transform (STFT) and wavelet transform are used to perform time-frequency domain analysis on the voltage signal to identify the instantaneous frequency changes and energy distribution changes of the voltage signal. By integrating the analysis results of the time domain, frequency domain, and time-frequency domain, the mutation points and abnormal fluctuations in the voltage signal are automatically identified to obtain mutation detection results.

[0106] Step S30: extract characteristic parameters of mutation amplitude, duration, frequency component and energy distribution from the mutation detection result, input the extracted characteristic parameters into a pre-trained fault type recognition model, and identify the fault type in the voltage signal.

[0107] In this step, when performing feature extraction and fault identification, characteristic parameters such as mutation amplitude, duration, frequency component and energy distribution are extracted from the mutation detection results, and the extracted characteristic parameters are input into the pre-trained fault type identification model. The input characteristic parameters are analyzed by the model to identify the fault type in the voltage signal.

[0108] Step S40: Analyze the fault diagnosis results based on the real-time change trend, the identified fault type and the characteristic parameters, including the fault occurrence time, duration, fault location and fault cause.

[0109] In this step, when diagnosing a fault, the voltage signal fluctuation is analyzed in combination with the real-time change trend, and the fault type, mutation characteristic parameters and real-time trend are integrated to generate the fault diagnosis result. The diagnosis result includes the fault occurrence time, duration, fault location and cause analysis, which is used to provide maintenance personnel with a detailed fault analysis report to facilitate rapid response and fault handling.

[0110] Step S50: Establish a real-time monitoring system for uninterrupted monitoring of the DC power supply, update the data and analysis results of the voltage signal in real time, set a multi-level alarm mechanism, and trigger an alarm according to the fault diagnosis result.

[0111] In this step, for real-time monitoring and alarm, a real-time monitoring system for uninterrupted monitoring of the DC power supply is established to update the data and analysis results of the voltage signal in real time. A multi-level alarm mechanism is set up according to the fault diagnosis results:

[0112] Warning level: Slight overvoltage / undervoltage, the system prompts attention.

[0113] Severity level: significant overvoltage / undervoltage, the system issues an alarm and recommends inspection.

[0114] Emergency level: Severe overvoltage / undervoltage, the system issues an emergency alarm and automatically triggers protection measures (such as power off protection).

[0115] According to the above multi-level alarm mechanism, the corresponding level of alarm is automatically triggered according to the diagnosis result, ensuring that the maintenance personnel can respond and handle it in time.

[0116] The DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis of the present invention establishes a comprehensive real-time monitoring and alarm mechanism by real-time acquisition of voltage signal data, multi-dimensional mutation detection, extraction of characteristic parameters, identification of fault types and generation of fault diagnosis results. This method not only improves the accuracy and timeliness of fault detection, but also provides maintenance personnel with a detailed fault analysis report, which is convenient for rapid response and handling of faults, and ensures the safe and stable operation of the DC power supply system.

[0117] When diagnosing over-voltage and under-voltage of a DC power supply based on mutation data analysis of the present invention, in the data acquisition stage of step S10, a high-precision voltage sensor can be installed at the output port of the DC power supply to ensure that the sensor can collect voltage signals at a high frequency (such as 1000 times per second), set the instantaneous data acquisition frequency to 1000 Hz, and record the timestamp of each acquisition to ensure that the collected data has high time resolution and synchronization, and transmit the collected voltage signal data to the central data processing unit in real time, and perform preliminary recording and storage. In the multi-dimensional mutation detection stage of step S20, the collected voltage signal data is filtered to remove high-frequency noise and low-frequency interference to obtain a clean signal. Then, in the time domain analysis, the sliding window is used to calculate the time domain characteristics (mean, variance, peak value, etc.) of the voltage signal to identify the mutation point. In the frequency domain analysis, the fast Fourier transform (FFT) is used to analyze the spectrum of the voltage signal to identify the changes in frequency components and detect abnormal frequency components. In the time-frequency domain analysis, the short-time Fourier transform (STFT) and wavelet transform are used to perform time-frequency domain analysis on the voltage signal to identify the instantaneous frequency changes and energy distribution changes of the voltage signal. In the mutation detection, the analysis results of the time domain, frequency domain and time-frequency domain are integrated to automatically identify the mutation points and abnormal fluctuations in the voltage signal to obtain the mutation detection results.

[0118] In the feature extraction and fault identification stage of step S30, when extracting characteristic parameters, characteristic parameters such as mutation amplitude, duration, frequency component and energy distribution are extracted from the mutation detection results, and the extracted characteristic parameters are input into the pre-trained fault type identification model, which is obtained based on historical data and convolutional neural network training. During fault identification, the model analyzes the input characteristic parameters and identifies the fault type in the voltage signal, such as overvoltage, undervoltage, transient transition, etc. In the fault diagnosis stage of step S40, the fluctuation of the voltage signal is analyzed in combination with the real-time change trend, and the fault type, mutation characteristic parameters and real-time trend are comprehensively considered to generate fault diagnosis results, including the time of occurrence, duration, fault location and cause analysis of the fault. Finally, in the real-time monitoring and alarm stage, a real-time monitoring system for uninterrupted monitoring of the DC power supply is established, and the data and analysis results of the voltage signal are updated in real time. According to the fault diagnosis results, a multi-level alarm mechanism is set up:

[0119] Warning level: Slight overvoltage / undervoltage, the system prompts attention.

[0120] Severity level: significant overvoltage / undervoltage, the system issues an alarm and recommends inspection.

[0121] Emergency level: Severe overvoltage / undervoltage, the system issues an emergency alarm and automatically triggers protection measures (such as power off protection).

[0122] The system automatically triggers alarms of corresponding levels based on the diagnosis results, ensuring that maintenance personnel can respond and handle them in a timely manner.

[0123] Exemplarily, when the DC power supply overvoltage and undervoltage diagnosis method based on mutation data analysis of the present invention is used to detect a DC power supply overvoltage fault, the specific steps are as follows:

[0124] 1. Data collection: At 14:35:27 on June 25, 2024, the voltage sensor collected a signal of an instantaneous voltage increase, and the voltage value instantly increased from 220V to 250V.

[0125] 2. Multi-dimensional mutation detection: Time domain analysis found that the voltage mean value increased significantly in a short period of time, frequency domain analysis found that the high-frequency component was enhanced, and time-frequency domain analysis showed that the voltage signal energy was concentrated at the mutation point.

[0126] 3. Feature extraction and fault identification: The extracted mutation amplitude is 30V, the duration is 2 seconds, and the frequency component shows high-frequency enhancement. These features are input into the fault identification model and identified as an overvoltage fault.

[0127] 4. Fault diagnosis: The system generates a diagnostic report, indicating that the overvoltage fault occurred at 14:35:27 on June 25, 2024, and lasted for 2 seconds. The fault location is at the power supply output end. The cause may be an instantaneous change in load or an internal fault in the power supply. The system triggers a severe level alarm, prompting maintenance personnel to check and recommend taking protective measures.

[0128] In this embodiment, see Figure 2 As shown, obtaining the real-time change trend of voltage signal data includes the following steps:

[0129] Step S101, setting the sampling frequency:

[0130] According to the Nyquist sampling theorem, set the sampling frequency At least the highest frequency component of the voltage signal twice, that is:

[0131]

[0132] In the formula, is the sampling frequency; is the highest frequency component of the voltage signal;

[0133] For example, if the highest frequency component of the voltage signal is 500 Hz, the sampling frequency should be set to at least 1000 Hz. In order to improve the accuracy and anti-interference ability of signal sampling, a higher sampling frequency than the theoretical value can be selected in actual operation, such as 4 to 5 times the highest frequency component. In addition, an anti-aliasing filter can be used to pre-process the signal before sampling to filter out high-frequency noise and ensure the quality of the sampled signal.

[0134] Step S102: time calibration and clock synchronization:

[0135] In this step, NTP server calibration is used. Use the Network Time Protocol (NTP) server for time calibration and clock synchronization. Each time the voltage signal data is collected, the current timestamp is recorded in real time to ensure the time consistency of the data.

[0136] Step S103: Data stream processing:

[0137] In this step, Apache Flink is used for data stream processing. The voltage signal data is obtained and processed according to the current timestamp. Flink can process high-throughput, low-latency data streams and is suitable for real-time data processing. During the data stream processing, the Apache Kafka distributed computing framework can be combined to buffer and transmit data to ensure the high availability and reliability of the data stream. The data stream is transmitted to Flink for processing through Kafka, and the results of Flink processing are then passed to the subsequent system through Kafka to achieve efficient data stream transmission and processing architecture. At the same time, Flink's fault-tolerant mechanism of Checkpoints can be used to ensure the robustness of real-time data processing.

[0138] Step S104: Sliding window analysis:

[0139] Sliding window configuration: Configure the sliding window parameters: window size and sliding step size. Set the sliding window size to 1 second and the sliding step size to 0.1 second. Use the sliding window to perform trend analysis on the real-time collected voltage signal data. Calculate the instantaneous change trend of the voltage signal in each window, that is, calculate the instantaneous change trend based on the data points in the sliding window.

[0140] The parameter settings of the sliding window can be adjusted according to actual needs. For monitoring of changes with higher frequencies, the window size and step size can be reduced, for example, the window size is set to 0.5 seconds and the step size is set to 0.05 seconds. In addition, the window types of the counting window and session window can be combined to use the average, maximum, and minimum values ​​for data analysis to more comprehensively understand the changing trend of the voltage signal.

[0141] Step S105, drawing a voltage change trend graph:

[0142] Graphics drawing: Use the graphical tool Matplotlib to draw real-time voltage signal data into graphs to show the trend of voltage changes and dynamically display the latest voltage changes.

[0143] By setting an appropriate sampling frequency, using NTP time synchronization, and using Apache Flink for data stream processing and sliding window analysis, the change trend of voltage signal data can be obtained and plotted in real time. This method provides high-precision, high-time resolution voltage data, which can monitor the change trend of voltage signals in real time, help to timely discover and diagnose faults, and ensure the safe and stable operation of the DC power supply system.

[0144] In this embodiment, see Figure 3 As shown, multi-dimensional mutation detection is performed on the voltage signal data after preprocessing, including the following steps:

[0145] Step S201: Wavelet decomposition

[0146] The Daubechies wavelet basis function is selected to perform wavelet decomposition on the preprocessed voltage signal data to obtain wavelet coefficients at different scales, wherein the wavelet coefficients include approximate coefficients (low-frequency part) and detail coefficients (high-frequency part);

[0147] Step S202: Abnormal coefficient value search

[0148] Search for abnormal coefficient values ​​in the wavelet coefficients of each scale. The abnormal coefficient values ​​correspond to the mutation points of the voltage signal. The positions of the signal mutation points are located according to the set threshold and through the time series distribution.

[0149] Step S203: Signal reconstruction and verification

[0150] The inverse wavelet transform is performed using the detail coefficients and approximate coefficients that retain the mutation point information, the signal is reconstructed, the reconstructed signal is analyzed, and the mutation point is verified.

[0151] Exemplarily, for a set of acquired voltage signal data, detecting the mutation point therein, the specific process is as follows:

[0152] (1) The original signal data is as follows:

[0153]

[0154] Here, the signal is There was a mutation at the second;

[0155] (2) Wavelet decomposition: Select the Daubechies wavelet basis function db4 to decompose the preprocessed voltage signal to obtain approximate coefficients and detail coefficients at different scales;

[0156] (3) Search for abnormal coefficient values: In the decomposed wavelet coefficients, the abnormal coefficient values ​​are searched in combination with the set threshold (the threshold is 3 times the standard deviation), and the detail coefficients are found to be There is a significant change in the second;

[0157] (4) Signal reconstruction and verification: Use the detail coefficients and approximate coefficients containing the mutation point information to perform inverse wavelet transform and reconstruct the signal. The reconstructed signal should be consistent with the original signal in The mutation at the second is consistent, thus verifying the accuracy of the mutation point.

[0158] Through the above steps, the mutation points in the voltage signal can be effectively detected and verified. Through multi-scale analysis, the mutation points can be accurately located. It has strong robustness and strong suppression ability against noise and interference, and is suitable for various forms of voltage signal analysis.

[0159] In this embodiment, see Figure 4 As shown, short-time Fourier transform (STFT) is used to verify the mutation point. When verifying the mutation point, the following steps are included:

[0160] Step S301, select window function and set parameters: select Hanning Window as window function, set window overlap rate to 70%, select window length L, and calculate short-time Fourier transform; overlap rate is 70%, that is, the window overlap length is .

[0161] Step S302, calculate short-time Fourier transform: perform STFT calculation on the pre-processed voltage signal, and the calculation formula is:

[0162]

[0163] in, is a voltage signal; It is the function window; is the window length; the overlap rate is 70%, that is, the overlap length is ;

[0164] Step S303, generating a time-frequency matrix: generating a time-frequency matrix according to the STFT calculation result, wherein each column corresponds to the spectrum information in a time window;

[0165] Step S304, draw a time-frequency diagram: use Matplotlib to draw a time-frequency diagram of the voltage signal, where the horizontal axis is time, the vertical axis is frequency, and the color / intensity represents the amplitude of the frequency component, and identify the mutation point in the time-frequency diagram, locate the time position of the mutation point according to the mutation characteristics in the time-frequency diagram, and compare and verify the mutation point position identified in the Daubechies wavelet basis function.

[0166] For example, for a set of voltage signal data as mentioned above, the mutation point is verified by using STFT, and the specific process is as follows:

[0167] (1) The original signal data is as follows:

[0168]

[0169] Here, the signal is There was a mutation at the second;

[0170] (2) Select window function and set parameters

[0171] Select Hanning window and set the window length (The signal sampling rate is 1 kHz, that is, the window length is 0.256 seconds), and the window overlap rate is 70%.

[0172] (3) Calculate the short-time Fourier transform

[0173] Perform STFT calculation on the preprocessed voltage signal to generate a time-frequency matrix.

[0174] (4) Draw a time-frequency diagram and verify the comparison

[0175] Through the time-frequency diagram generated by the above code, it can be observed that The frequency spectrum changes at 1000 s, identify and locate the mutation point, and compare the mutation point position identified in the time-frequency diagram with the mutation point position identified using the Daubechies wavelet basis function. If the two positions are consistent, it means that the result of the mutation point detection is accurate.

[0176] In this embodiment, multi-dimensional mutation detection is performed on the voltage signal, and the time domain, frequency domain and time-frequency domain characteristics of the voltage signal are analyzed to identify mutation points and abnormal fluctuations, including:

[0177] Step 1) Time domain analysis:

[0178] (1) Sliding window standard deviation detection mutation point: Use the sliding window to calculate the standard deviation of the signal within the sliding window and detect the mutation point; set the window length to ; The standard deviation formula of the signal in the sliding window is:

[0179]

[0180] In the formula, is the mean value of the signal in the window, is the voltage value within the window.

[0181] Assume that the voltage signal data is:

[0182]

[0183] Use sliding window length The code for analyzing and executing the sliding window standard deviation to detect mutation points is as follows:

[0184] #Generate voltage signal data

[0185] fs=1000

[0186] t = np.arange(0, 10, 1 / fs)

[0187] v=np.where(t < 5, 120 + 5*np.sin(2*np.pi*t), 130 + 5*np.sin(2*np.pi*t))

[0188] #Calculate sliding window standard deviation

[0189] N=100

[0190] std_dev=[np.std(v[i:i+N]) for i in range(len(v)-N)]

[0191] # Detect mutation points

[0192] threshold=np.mean(std_dev) + 2 * np.std(std_dev)

[0193] change_points=[i for i, val in enumerate(std_dev) if val > threshold]

[0194] print("Detected change points:", change_points)

[0195] (2) Calculate the mutation amplitude: At each identified mutation point, calculate the difference in voltage value before and after the mutation to obtain the mutation amplitude:

[0196] Mutation amplitude = |voltage value (after mutation) - voltage value (before mutation)|.

[0197] Step 2), frequency domain analysis:

[0198] (1) Fourier transform (FFT): The signal is converted from the time domain to the frequency domain through Fourier transform (FFT) to analyze the frequency components;

[0199] (2) Calculate power spectral density: Calculate the power spectral density of the signal and identify the frequency components and intensities;

[0200] (3) Analyze changes in frequency domain characteristics: By comparing the power spectrum density in different time periods, determine whether there is a sudden change or abnormal fluctuation in the frequency components.

[0201] Step 3), time-frequency domain analysis:

[0202] (1) Short-time Fourier transform (STFT): Use short-time Fourier transform (STFT) to calculate the time-frequency diagram of the signal;

[0203] (2) Wavelet transform: Perform wavelet transform on the signal to analyze the changing characteristics at different scales.

[0204] Finally, the mutation points and abnormal fluctuations are identified by combining the time domain, frequency domain and time-frequency domain features. Combining the above multi-dimensional analysis methods, the mutation points and abnormal fluctuations in the voltage signal can be detected more comprehensively and accurately, improving the reliability and accuracy of the detection.

[0205] In this embodiment, extracting characteristic parameters of mutation amplitude, duration, frequency component and energy distribution from the mutation detection result includes the following steps:

[0206] (1) Mutation point detection and mutation amplitude calculation:

[0207] Use the sliding window to calculate the standard deviation of the signal within the sliding window and identify the mutation point of the voltage signal;

[0208] At each identified mutation point, the difference in voltage values ​​before and after the mutation is calculated, and the mutation amplitude is calculated: mutation amplitude = |voltage value (after mutation) - voltage value (before mutation)|.

[0209] The voltage signal data is divided into sliding windows of fixed size, the standard deviation is calculated for each window, and compared with the standard deviation of the previous window. If the change of the standard deviation exceeds a predetermined threshold, it is marked as a mutation point. At each mutation point, the difference in voltage values ​​before and after the mutation is calculated.

[0210] For example, assume that the voltage signal sampling frequency is 1000Hz and the sliding window size is 100ms. When the standard deviation change in the sliding window exceeds the set threshold, the mutation point is identified. At a certain mutation point, the voltage value before the mutation is 220V, and the voltage value after the mutation is 200V, then the mutation amplitude is:

[0211] |220V-200V|=20V.

[0212] (2) Extraction duration:

[0213] Through time domain analysis, the timestamps before and after the mutation point are determined, and the time period during which the voltage signal is stable is identified to calculate the duration:

[0214] Duration = mutation end time - mutation start time.

[0215] The timestamp of the mutation point is recorded as the fault start time, and the voltage signal after the mutation is observed until the voltage signal stabilizes. The timestamp when it stabilizes is recorded as the fault end time, and the duration is calculated.

[0216] For example, in a certain mutation detection, the mutation point timestamp is 0.5 seconds, and the voltage signal returns to stability at 0.7 seconds after the mutation. The duration is:

[0217] 0.7 seconds - 0.5 seconds = 0.2 seconds.

[0218] (3) Frequency component extraction:

[0219] Fourier transform and short-time Fourier transform (STFT) are used to analyze the voltage signal spectrum before and after the mutation and extract the main frequency components. Specifically:

[0220] Through frequency domain and time-frequency domain analysis, the frequency spectrum information is obtained by Fourier transform (FFT) of a voltage signal data before and after the mutation;

[0221] Identify the main frequency components through the spectrum diagram;

[0222] Perform STFT on the entire voltage signal to generate a time-frequency matrix, extract the spectrum information at the mutation point from the time-frequency matrix, identify the frequency components, and extract the frequency components.

[0223] During operation, a section of voltage signal data before and after the mutation is subjected to Fourier transform (FFT) to obtain spectrum information, the main frequency components are identified from the spectrum diagram, the entire voltage signal is subjected to STFT to generate a time-frequency matrix, the spectrum information at the mutation point is extracted from the time-frequency matrix, and the frequency components are identified.

[0224] Exemplarily, the voltage signal data from 0.1 seconds before the mutation to 0.1 seconds after the mutation is subjected to FFT analysis to obtain spectrum information. In the spectrum diagram, the main frequency components are 50 Hz and 150 Hz. The entire signal is subjected to STFT analysis to confirm that the main frequency component at the mutation point is 50 Hz.

[0225] (4) Energy distribution extraction:

[0226] The energy distribution is extracted by calculating the power spectrum density of the voltage signal before and after the mutation point. Specifically:

[0227] The power spectrum density of the voltage signal before and after the mutation point is calculated and the energy distribution is extracted.

[0228] Among them, the power spectral density (PSD) of the voltage signal before and after the mutation point is calculated, and the PSD before and after the mutation is compared to analyze the change in energy distribution. For example, the power spectral density of the voltage signal before the mutation point is mainly concentrated around 50Hz. After the mutation, the energy distribution increases significantly around 50Hz and 150Hz.

[0229] In this embodiment, the fault diagnosis result includes the fault occurrence time, duration, fault location and fault cause, among which:

[0230] When determining the time of fault occurrence, determine the timestamp of the mutation point as the time of fault occurrence. If there are multiple mutation points, record the timestamp of each mutation point.

[0231] When determining the fault duration, the time period during which the voltage signal is stable before and after each mutation point is calculated to determine the fault duration;

[0232] When locating a fault, the fault location is located by combining the grid topology and characteristic parameters of the fault voltage signal;

[0233] When inferring the cause of a fault, the cause of the fault is inferred based on the fault type, spectrum characteristics, and energy distribution.

[0234] For example, a voltage signal mutation is detected in a certain power grid, with a mutation point timestamp of 1.2 seconds and a duration of 0.3 seconds. Through spectrum analysis, the main frequency components are 50Hz and 150Hz, and the energy distribution increases significantly at 150Hz. Combined with the grid topology, the fault is located near transformer T1. By analyzing the spectrum characteristics and energy distribution, it is inferred that the cause of the fault may be a transformer winding fault.

[0235] Through the above steps, the present invention can systematically extract the mutation characteristic parameters in the voltage signal, and perform fault diagnosis based on these parameters, including the determination of the fault occurrence time, duration, fault location and fault cause. The method of the present invention effectively improves the accuracy and rapid response capability of power grid fault diagnosis.

[0236] In this embodiment, when establishing a real-time monitoring system, it includes a sensor layer, a data acquisition layer, a data transmission layer, a data storage and processing layer, and a user interface layer. Specifically, the real-time monitoring system includes:

[0237] Sensor layer: Install sensors for measuring voltage and current parameters at key nodes of the DC power supply, and set the data acquisition frequency; install voltage sensors and current sensors at key nodes of the DC power supply (transformer, distribution cabinet, busbar), and set the data acquisition frequency of the sensors according to system requirements and signal characteristics. For example, set it to 1000Hz to capture rapidly changing voltage signals. For example, install a voltage sensor (e.g., LemLV 25-P) and a current sensor (e.g., Lem LA 55-P) at a grid node, and set the data acquisition frequency to 1000Hz.

[0238] Data acquisition layer: Use data acquisition equipment to obtain voltage signal data from sensors and convert these signals into digital signals for processing; when applied, connect the sensor to the data acquisition device (NI DAQ data acquisition card), the data acquisition device collects voltage and current signals at the set frequency, and converts the analog signals into digital signals. When applied, use NI DAQ (such as NI USB-6356) to connect the sensor, set the acquisition frequency to 1000Hz, and obtain voltage and current signals in real time.

[0239] Data transmission layer: used to transmit the voltage signal data acquired by the data acquisition layer to the monitoring center to ensure the real-time and reliability of data transmission. Specifically, the data acquisition device is connected to the monitoring center server via Ethernet or wireless network, and the data transmission protocol is used to ensure the integrity and real-time of the data. Specifically, the NI DAQ device is connected to the monitoring center server via Ethernet, and the TCP / IP protocol is used to transmit data to ensure that 1000 data points are transmitted per second.

[0240] Data storage and processing layer: Build a server to store, process and analyze data, perform time domain, frequency domain and time-frequency domain feature analysis and fault diagnosis on the data; locate the fault location and analyze the fault diagnosis results based on the grid topology and fault feature parameters. Specifically, use Python for data processing, use the Pandas library to store data, use the SciPy library for FFT and STFT analysis, use custom algorithms for fault diagnosis, and locate the fault based on the grid topology.

[0241] User interface layer: Design a human-machine interface for real-time display of monitoring data and alarm information. Specifically, use Ignition to design the HMI interface to display voltage and current waveforms, spectrum diagrams, fault diagnosis results in real time, and set alarm thresholds. When the voltage suddenly changes and exceeds the set value, an alarm is triggered.

[0242] Through the design and implementation of the above layers, a real-time monitoring system was established, which can obtain the voltage and current data of key nodes in real time, collect, transmit, store and process data, perform fault diagnosis in combination with the grid topology, and display monitoring data and alarm information in real time through the user interface. The implementation of this system improves the reliability and fault response of the grid.

[0243] In summary, the embodiments of the present invention provide a DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis, which collects the voltage signal data of the DC power supply in real time through the sensing acquisition equipment, and uses the NTP server for time calibration and clock synchronization to ensure the time accuracy of data acquisition; Apache Flink is used for data stream processing, and the sliding window technology is used to perform trend analysis on the real-time collected voltage signal data, so as to timely discover the real-time change trend of the voltage; the time domain, frequency domain and time-frequency domain features are combined to comprehensively analyze the voltage signal, and identify the mutation points and abnormal fluctuations in the voltage signal; Daubechies wavelet basis function and short-time Fourier transform are used to perform multi-dimensional mutation detection to ensure the accuracy and precision of mutation point identification; characteristic parameters such as mutation amplitude, duration, frequency component and energy distribution are extracted from the mutation detection results to comprehensively describe the change characteristics of the voltage signal; a pre-trained fault type recognition model is used to identify the faults in the voltage signal based on the extracted characteristic parameters. The fault type can be identified to improve the accuracy of fault identification; based on the real-time change trend, identified fault type and characteristic parameters, the fault diagnosis results are analyzed, including the fault occurrence time, duration, fault location and fault cause, and a comprehensive fault diagnosis report is provided; a multi-level alarm mechanism is set up to trigger the corresponding level of alarm according to the fault diagnosis results to ensure timely response and handling of various fault conditions; by establishing a real-time monitoring system for uninterrupted monitoring of the DC power supply, the data and analysis results of the voltage signal are updated in real time to ensure the real-time and continuity of the system; a user interface module is provided to display the data, analysis results and alarm information of the voltage signal in real time, so that users can monitor and manage conveniently.

[0244] The present invention realizes accurate diagnosis of overvoltage and undervoltage of DC power supply through multi-dimensional mutation detection and feature extraction combined with a pre-trained fault type recognition model, and ensures the practicality and efficiency of the system through real-time monitoring and multi-level alarm mechanism. It can be widely used in power systems, industrial automation and other fields.

[0245] It should be noted that the above figures are only schematic illustrations of the processes included in the method according to an exemplary embodiment of the present invention, and are not intended to be limiting. It is easy to understand that the processes shown in the above figures do not indicate or limit the time sequence of these processes. In addition, it is also easy to understand that these processes can be performed synchronously or asynchronously, for example, in multiple modules.

[0246] It should be understood that, although described in a certain order, these steps are not necessarily performed in sequence in the above order. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be performed in other orders. Moreover, a part of the steps of the present embodiment may include a plurality of steps or a plurality of stages, and these steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a part of the steps or stages in other steps or other steps.

[0247] In a second aspect of the embodiments of the present invention, the present invention further provides a DC power supply over-voltage and under-voltage diagnosis system based on mutation data analysis, comprising:

[0248] Sensing and acquisition equipment: used to collect voltage signal data of DC power supply in real time, set the instantaneous data acquisition frequency and record the acquisition time, and obtain the real-time change trend of voltage signal data; the sensing and acquisition equipment is a voltage sensor and data acquisition equipment installed at the key nodes of the DC power supply, and the acquisition frequency is set to 1000Hz to ensure that the real-time change trend of the voltage signal is captured.

[0249] Multidimensional mutation detection module: perform multidimensional mutation detection on the voltage signal data after preprocessing, analyze the time domain, frequency domain and time-frequency domain characteristics of the voltage signal, and identify the mutation points and abnormal fluctuations in the voltage signal; use wavelet transform for preprocessing, use Canny edge detection to analyze the time domain, frequency domain and time-frequency domain characteristics of the voltage signal, identify the mutation points and abnormal fluctuations. Specifically, use Matlab to implement signal preprocessing and mutation detection.

[0250] Feature extraction module: used to extract characteristic parameters from mutation detection results, including mutation amplitude, duration, frequency component and energy distribution; use Python to write feature extraction algorithm to extract key features of voltage signals.

[0251] Fault type identification module: used to input the extracted feature parameters into the pre-trained fault type identification model to identify the fault type in the voltage signal; specifically, through a trained Scikit-learn-based support vector machine model, overvoltage, undervoltage, transient faults, etc. are identified.

[0252] Fault diagnosis module: Based on real-time change trends, identified fault types and characteristic parameters, analyze fault diagnosis results, including fault occurrence time, duration, fault location and fault cause;

[0253] Real-time monitoring module: used to establish a real-time monitoring system for uninterrupted monitoring of DC power supply, and update the data and analysis results of voltage signals in real time;

[0254] Alarm mechanism module: set up a multi-level alarm mechanism to trigger the corresponding level of alarm according to the fault diagnosis results;

[0255] Data storage module: used to store all collected voltage signal data, processing results and alarm records;

[0256] User interface module: provides a user interface for real-time display of voltage signal data, analysis results and alarm information.

[0257] The training steps of the Scikit-learn-based Support Vector Machine (SVM) model are as follows:

[0258] 1. Install Scikit-learn: First install the Scikit-learn library.

[0259] 2. Import necessary libraries: Before you start writing code, import the required Scikit-learn library,

[0260] 3. Load and prepare data: Divide the data to be loaded into feature variables (X) and target variables (y).

[0261] Example data loading:

[0262] data = pd.read_csv('voltage_fault_data.csv')

[0263] X = data.drop('fault_type', axis=1) # Feature variable

[0264] y = data['fault_type'] # Target variable.

[0265] 4. Split the dataset: Divide the dataset into a training set and a test set, where the test set accounts for 20%-30% of the dataset.

[0266] 5. Train the SVM model: Select an SVM classifier and train it with the training data, use the trained model to predict the test data, and evaluate the model performance.

[0267] Through the above steps, a Scikit-learn-based SVM model is trained and evaluated.

[0268] Among them, the fault diagnosis module analyzes the fault diagnosis results based on the real-time change trend, the identified fault type and the characteristic parameters, including the fault occurrence time, duration, fault location and fault cause. The python codes for real-time data analysis, real-time data collection and analysis, and alarm triggering are as follows:

[0269] Real-time data analysis:

[0270] def diagnose_fault(real_time_data, model):

[0271] fault_type = model.predict(real_time_data['features'])

[0272] fault_time = real_time_data['timestamp']

[0273] fault_duration = real_time_data['duration']

[0274] fault_location = real_time_data['location']

[0275] fault_reason = "Analysis of the specific reason" # Further rules or models are needed to determine the cause of the fault

[0276] return {

[0277] 'fault_type': fault_type,

[0278] 'fault_time': fault_time,

[0279] 'fault_duration': fault_duration,

[0280] 'fault_location': fault_location,

[0281] 'fault_reason': fault_reason

[0282] }

[0283] Real-time data collection and analysis:

[0284] def real_time_monitoring():

[0285] while True:

[0286] real_time_data = get_real_time_data() # Function to get real-time data

[0287] diagnosis = diagnose_fault(real_time_data, svm_model)

[0288] update_dashboard(diagnosis) # Update the user interface

[0289] time.sleep(1) # Update once every second, can be adjusted as needed.

[0290] Alarm trigger:

[0291] def trigger_alarm(diagnosis):

[0292] if diagnosis['fault_type'] == 'Serious fault':

[0293] send_alert('urgent', diagnosis)

[0294] elif diagnosis['fault_type'] == 'Medium fault':

[0295] send_alert('Warning', diagnosis)

[0296] elif diagnosis['fault_type'] == 'Minor fault':

[0297] send_alert('prompt', diagnosis)

[0298] Through the above detailed steps and code examples, the DC power supply over-voltage and under-voltage diagnosis system based on mutation data analysis of the present invention is used to execute the steps of the DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis in the above embodiment, which will not be repeated here. The DC power supply early warning operation and maintenance system of the present invention can accurately detect the occurrence of overvoltage or undervoltage by analyzing the mutation point of voltage data. This real-time detection capability helps to timely warn in the early stage of the problem to avoid further deterioration of the problem. By using edge detection and wavelet transform of mutation data analysis, a large amount of voltage data can be quickly processed and analyzed, and efficient data processing capabilities can significantly improve the response speed of the diagnostic system. Mutation data analysis can also more accurately identify abnormal changes in voltage and reduce false alarms caused by noise or other non-fault factors, which is crucial to improving the reliability of the system and user trust. The diagnostic system can monitor the DC power supply voltage in real time, and when overvoltage or undervoltage is detected, it immediately takes corresponding adjustment measures, which helps to maintain the stable operation of the power supply system, protect equipment and extend service life. The DC power supply overvoltage and undervoltage diagnosis system based on mutation data analysis of the present invention can not only improve the accuracy and timeliness of fault detection, but also significantly improve the overall stability and reliability of the power supply system, while reducing maintenance costs, bringing significant economic and operational benefits.

[0299] According to a third aspect of an embodiment of the present invention, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the method of any one of the above embodiments is implemented.

[0300] The computer device includes a processor and a memory, and may also include an input system and an output system. The processor, memory, input system and output system may be connected via a bus or other means, and the input system may receive input digital or character information, and generate signal input related to the migration of DC power supply over-voltage and under-voltage diagnosis based on mutation data analysis. The output system may include display devices such as a display screen.

[0301] As a non-volatile computer-readable storage medium, the memory can be used to store non-volatile software programs, non-volatile computer executable programs and modules, such as the program instructions / modules corresponding to the DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis in the embodiment of the present application. The memory may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created by the use of the DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the local module 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.

[0302] The processor may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip in some embodiments. The processor is generally used to control the overall operation of the computer device. In this embodiment, the processor is used to run the program code stored in the memory or process data. The processors of multiple computer devices of the computer device of this embodiment execute various functional applications and data processing of the server by running non-volatile software programs, instructions and modules stored in the memory, that is, implementing the steps of the DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis of the above method embodiment.

[0303] It should be understood that, to the extent that they do not conflict with each other, all the embodiments, features and advantages described above for the DC power supply over- and under-voltage diagnosis method based on mutation data analysis according to the present invention are also applicable to the DC power supply over- and under-voltage diagnosis and storage medium based on mutation data analysis according to the present invention.

[0304] It will also be appreciated by those skilled in the art that various exemplary logic blocks, modules, circuits and algorithm steps described in conjunction with the disclosure herein can be implemented as electronic hardware, computer software or a combination of the two. In order to clearly illustrate this interchangeability of hardware and software, a general description has been given to the functions of various schematic components, blocks, modules, circuits and steps. Whether this function is implemented as software or hardware depends on specific applications and the design constraints imposed on the entire system. Those skilled in the art can implement the function in various ways for each specific application, but this implementation decision should not be interpreted as causing a departure from the disclosed scope of the embodiments of the present invention.

[0305] Finally, it should be noted that the computer-readable storage medium (e.g., memory) herein may be a volatile memory or a non-volatile memory, or may include both a volatile memory and a non-volatile memory. As an example and not by way of limitation, a non-volatile memory may include a read-only memory (ROM), a programmable ROM (PROM), an electrically programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM), or a flash memory. A volatile memory may include a random access memory (RAM), which may act as an external cache memory. As an example and not by way of limitation, RAM may be obtained in a variety of forms, such as synchronous RAM (DRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct Rambus RAM (DRRAM). The storage devices of the disclosed aspects are intended to include, but are not limited to, these and other suitable types of memory.

[0306] The various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure herein may be implemented or executed using the following components designed to perform the functions herein: a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination of these components. A general-purpose processor may be a microprocessor, but alternatively, the processor may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, for example, a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP, and / or any other such configuration.

[0307] The above are exemplary embodiments disclosed in the present invention, but it should be noted that various changes and modifications may be made without departing from the scope disclosed in the embodiments of the present invention as defined in the claims. The functions, steps and / or actions of the method claims according to the disclosed embodiments described herein do not need to be performed in any particular order. In addition, although the elements disclosed in the embodiments of the present invention may be described or required in individual form, they may also be understood as multiple unless explicitly limited to the singular.

[0308] It should be understood that, as used herein, the singular form "a" or "an" is intended to include the plural form as well, unless the context clearly supports an exception. It should also be understood that, as used herein, "and / or" refers to any and all possible combinations of one or more of the items listed in association. The serial numbers of the embodiments disclosed in the above embodiments of the present invention are for description only and do not represent the advantages and disadvantages of the embodiments.

[0309] A person skilled in the art should understand that the discussion of any of the above embodiments is only exemplary and is not intended to imply that the scope of the disclosure of the embodiments of the present invention (including the claims) is limited to these examples; under the concept of the embodiments of the present invention, the technical features in the above embodiments or different embodiments can also be combined, and there are many other changes in different aspects of the above embodiments of the present invention, which are not provided in detail for the sake of simplicity. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present invention should be included in the protection scope of the embodiments of the present invention.

Claims

1. A DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis, characterized in that: The method comprises the following steps: The voltage signal data of the DC power supply is collected in real time through the sensor collection equipment, the instantaneous data collection frequency is set and the collection time is recorded to obtain the real-time change trend of the voltage signal data; After preprocessing the voltage signal data, multi-dimensional mutation detection is performed to analyze the time domain, frequency domain and time-frequency domain characteristics of the voltage signal, identify the mutation points and abnormal fluctuations in the voltage signal, and obtain the mutation detection results; Extract characteristic parameters of mutation amplitude, duration, frequency component and energy distribution from mutation detection results, input the extracted characteristic parameters into the pre-trained fault type recognition model to identify the fault type in the voltage signal; Analyze fault diagnosis results based on real-time change trends, identified fault types and characteristic parameters, including fault occurrence time, duration, fault location and fault cause; Establish a real-time monitoring system for uninterrupted monitoring of DC power supply, update the data and analysis results of voltage signals in real time, set up a multi-level alarm mechanism, and trigger alarms based on fault diagnosis results.

2. The DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis according to claim 1, characterized in that: When obtaining the real-time change trend of voltage signal data, the following steps are included: Setting a sampling frequency, wherein the sampling frequency is set to be at least twice the highest frequency component of the voltage signal; Use NTP server for time calibration and clock synchronization. Record the current timestamp in real time every time voltage signal data is collected. Use Apache Flink for data stream processing, obtain and process voltage signal data according to the current timestamp, and use sliding windows to perform trend analysis on the real-time collected voltage signal data, where the instantaneous change trend is calculated based on the data points in the sliding window; The obtained real-time voltage signal data is plotted into a graph to show the trend of voltage change.

3. The DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis according to claim 2, characterized in that: After preprocessing the voltage signal data, multi-dimensional mutation detection is performed, including the following steps: The Daubechies wavelet basis function is selected to perform wavelet decomposition on the preprocessed voltage signal data to obtain wavelet coefficients at different scales, wherein the wavelet coefficients include approximate coefficients and detail coefficients, the approximate coefficients are the low-frequency part, and the detail coefficients are the high-frequency part; Search for abnormal coefficient values ​​in the wavelet coefficients of each scale. The abnormal coefficient values ​​correspond to the mutation points of the voltage signal. The positions of the signal mutation points are located according to the set threshold and through the time series distribution. The inverse wavelet transform is performed using the detail coefficients and approximate coefficients that retain the mutation point information, the signal is reconstructed, the reconstructed signal is analyzed, and the mutation point is verified.

4. The DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis according to claim 1, characterized in that: Short-time Fourier transform is used to verify the mutation point. When verifying the mutation point, the following steps are included: Select the Hanning window, set the window overlap rate to 70%, and select the window length L to calculate the short-time Fourier transform; The STFT calculation is performed on the preprocessed voltage signal, and the calculation formula is: ； in, is a voltage signal; It is the function window; is the window length; the overlap rate is 70%, that is, the overlap length is ; Generate a time-frequency matrix based on the STFT calculation results, where each column corresponds to the spectrum information in a time window; Use Matplotlib to draw the time-frequency diagram of the voltage signal, where the horizontal axis is time and the vertical axis is frequency. The color / intensity represents the amplitude of the frequency component. The mutation points are identified in the time-frequency diagram. According to the mutation characteristics in the time-frequency diagram, the time position of the mutation point is located, and the position of the mutation point identified in the Daubechies wavelet basis function is compared and verified.

5. The DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis according to claim 4, characterized in that: Perform multi-dimensional mutation detection on voltage signals, analyze the time domain, frequency domain and time-frequency domain characteristics of voltage signals, and identify mutation points and abnormal fluctuations, including: 1) Time domain analysis: Use the sliding window to calculate the standard deviation of the signal within the sliding window and detect the mutation point; At each identified mutation point, the voltage difference before and after the mutation is calculated to obtain the mutation amplitude: Mutation amplitude = |voltage value (after mutation) - voltage value (before mutation)|; 2) Frequency domain analysis: Convert the signal from the time domain to the frequency domain through Fourier transform and analyze the frequency components; Calculate the power spectral density of the signal and identify the frequency components and their intensities; Analyze the changes in frequency domain characteristics to determine whether there are sudden changes or abnormal fluctuations in frequency components; 3) Time-frequency domain analysis: Compute the time-frequency plot of the signal using the short-time Fourier transform; Perform wavelet transform on the signal to analyze the changing characteristics at different scales; Combine the time domain, frequency domain and time-frequency domain features to identify mutation points and abnormal fluctuation locations.

6. The DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis according to claim 5, characterized in that: Extracting characteristic parameters of mutation amplitude, duration, frequency component and energy distribution from the mutation detection results includes the following steps: 1) Mutation point detection and mutation amplitude calculation: Use the sliding window to calculate the standard deviation of the signal within the sliding window and identify the mutation point of the voltage signal; At each identified mutation point, the difference in voltage value before and after the mutation is calculated to obtain the mutation amplitude: mutation amplitude = |voltage value (after mutation) - voltage value (before mutation)|; 2) Extraction duration: Determine the timestamps before and after the mutation point based on time domain analysis; Identify the time period when the voltage signal is stable before and after the mutation and obtain the duration: Duration = mutation end time - mutation start time; 3) Frequency component extraction: Through frequency domain and time-frequency domain analysis, a voltage signal data before and after the mutation is subjected to Fourier transform to obtain spectrum information; Identify the main frequency components through the spectrum diagram; Perform STFT on the entire voltage signal to generate a time-frequency matrix, extract the spectrum information at the mutation point from the time-frequency matrix, identify the frequency components, and extract the frequency components; 4) Energy distribution extraction: The power spectrum density of the voltage signal before and after the mutation point is calculated and the energy distribution is extracted.

7. The DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis according to claim 6, characterized in that: The fault diagnosis results include the fault occurrence time, duration, fault location and fault cause, including: When determining the time of fault occurrence, determine the timestamp of the mutation point as the time of fault occurrence. If there are multiple mutation points, record the timestamp of each mutation point. When determining the fault duration, the time period during which the voltage signal is stable before and after each mutation point is calculated to determine the fault duration; When locating a fault, the fault location is located by combining the grid topology and characteristic parameters of the fault voltage signal; When inferring the cause of a fault, the cause of the fault is inferred based on the fault type, spectrum characteristics, and energy distribution.

8. The DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis according to claim 7, characterized in that: When establishing a real-time monitoring system, the real-time monitoring system includes: Sensor layer: Install sensors for measuring voltage and current parameters at key nodes of the DC power supply and set the data collection frequency; Data acquisition layer: Use data acquisition equipment to obtain voltage signal data from sensors; Data transmission layer: used to transmit the voltage signal data acquired by the data acquisition layer to the monitoring center; Data storage and processing layer: Build servers to perform data storage, processing and analysis, perform time domain, frequency domain and time-frequency domain feature analysis and fault diagnosis on data; locate the fault location based on the grid topology and fault feature parameters, and analyze to obtain fault diagnosis results; User interface layer: Design a human-machine interface to display monitoring data and alarm information in real time.

9. A DC power supply over-voltage and under-voltage diagnosis system based on mutation data analysis, characterized in that: The system is used to perform the DC power supply over-voltage and under-voltage diagnosis method based on mutation data analysis as described in any one of claims 1 to 8, comprising: Sensor acquisition equipment: used to collect voltage signal data of DC power supply in real time, set the instantaneous data acquisition frequency and record the acquisition time, and obtain the real-time change trend of voltage signal data; Multidimensional mutation detection module: performs multidimensional mutation detection on voltage signal data after preprocessing, analyzes the time domain, frequency domain and time-frequency domain characteristics of the voltage signal, and identifies mutation points and abnormal fluctuations in the voltage signal; Feature extraction module: used to extract feature parameters from mutation detection results, including mutation amplitude, duration, frequency component and energy distribution; Fault type identification module: used to input the extracted characteristic parameters into the pre-trained fault type identification model to identify the fault type in the voltage signal; Fault diagnosis module: Based on real-time change trends, identified fault types and characteristic parameters, analyze fault diagnosis results, including fault occurrence time, duration, fault location and fault cause; Real-time monitoring module: used to establish a real-time monitoring system for uninterrupted monitoring of DC power supply, and update the data and analysis results of voltage signals in real time; Alarm mechanism module: Set up a multi-level alarm mechanism and trigger the corresponding level of alarm according to the fault diagnosis results.

10. The DC power supply over-voltage and under-voltage diagnosis system based on mutation data analysis according to claim 9, characterized in that: The system also includes: Data storage module: used to store all collected voltage signal data, processing results and alarm records; User interface module: provides a user interface for real-time display of voltage signal data, analysis results and alarm information.

Images