A remote monitoring method and system for AC power supply operation status

Abstract

The present invention relates to the field of power supply monitoring and analysis, and in particular to a method and system for remote monitoring of the operating status of an AC power supply. The method comprises the following steps: obtaining a power supply historical operating status log and an AC power supply component structure diagram; performing a time-series load operation analysis on the power supply historical operating status log, and performing load demand prediction to generate a load demand prediction curve; obtaining multi-dimensional operating monitoring parameters, performing dynamic temperature distribution rendering on the operating temperature monitoring parameters, and constructing a three-dimensional temperature distribution field of the power supply; performing low-frequency filtering processing on multi-part acoustic wave signals, and performing power supply aging status prediction to generate power supply life status prediction data; performing dynamic evolution of the temperature distribution of the AC power supply component structure diagram based on the power supply life status prediction data and the three-dimensional temperature distribution field of the power supply, and constructing a dynamic twin evolution model. The present invention realizes accurate and efficient remote monitoring and analysis of AC power supplies.

Description

Technical Field

[0001] The present invention relates to the field of power supply monitoring and analysis, and in particular to a method and system for remotely monitoring the operating status of an alternating current power supply. Background Art

[0002] The power supply system is the basic support for modern industry and daily life, and its stable and reliable operation is crucial. As a key output device in the power system, the operating status of the AC power supply directly affects the safety and reliability of the entire system. However, with the continuous improvement of industrial automation, power supply equipment is facing new challenges. During long-term continuous operation, the AC power supply may be affected by various internal and external factors, such as component aging, load fluctuations, environmental changes, etc. These factors may cause abnormal deviations in key parameters such as the output power, voltage, and frequency of the power supply, thereby affecting the normal operation of the power supply equipment and even causing serious failures.

[0003] Traditional AC power supply operation status monitoring methods usually rely on manual inspections and manual testing. This method has many problems: first, manual observations are prone to overlook some hidden abnormal signs. Second, manual testing is inefficient and cannot achieve continuous monitoring of the power supply throughout its life cycle. Manual judgment has certain subjectivity and uncertainty, and it is difficult to accurately diagnose the specific cause of the power supply failure. Therefore, it is urgent to develop an intelligent and automated AC power supply remote monitoring method to meet the needs of modern power systems for reliable power supply. Summary of the invention

[0004] In order to solve the above technical problems, the present invention proposes a method and system for remotely monitoring the operating status of an AC power supply, so as to solve at least one of the above technical problems.

[0005] To achieve the above object, the present invention provides a method for remotely monitoring the operating status of an AC power supply, comprising the following steps:

[0006] Step S1: Obtain a power supply historical operation status log and an AC power supply component structure diagram; perform a time series load operation analysis on the power supply historical operation status log, and perform load demand prediction to generate a load demand prediction curve;

[0007] Step S2: acquiring multi-dimensional operation monitoring parameters, wherein the multi-dimensional operation monitoring parameters include operation temperature monitoring parameters and multi-position acoustic wave signals; performing dynamic temperature distribution rendering on the operation temperature monitoring parameters to construct a three-dimensional temperature distribution field of the power supply;

[0008] Step S3: performing low-frequency filtering on the acoustic wave signals of multiple parts and predicting the power supply aging state to generate power supply life state prediction data;

[0009] Step S4: dynamically evolve the temperature distribution of the AC power component structure diagram based on the power life state prediction data and the three-dimensional temperature distribution field of the power supply, and construct a dynamic twin evolution model;

[0010] Step S5: simulating the load demand evolution of the dynamic twin evolution model according to the load demand prediction curve, and performing abnormal fault prediction to generate abnormal fault prediction data;

[0011] Step S6: Make iterative fault diagnosis decisions based on abnormal fault prediction data, perform diagnosis visualization, and build a visualization view of the optimal diagnosis strategy.

[0012] By analyzing the historical operation status log of the power supply, the present invention can understand the usage of the load, which is helpful to optimize the power supply strategy. Generating a load demand prediction curve is helpful to predict the changes in the load in advance so as to make corresponding adjustments and arrangements. Constructing a three-dimensional temperature distribution field of the power supply is helpful to monitor and manage the temperature changes in the operation of the power supply and improve the understanding of the power supply status. Through the power supply life status prediction data, it is helpful to predict the aging of the power supply components in advance, so as to take corresponding maintenance and replacement measures. Constructing a dynamic twin evolution model can help monitor the temperature distribution evolution of the power supply structure in real time, improve the real-time and accuracy of the power supply operation status, and simulate the load demand evolution according to the load demand prediction curve to understand the future load situation and provide a basis for adjusting the power supply plan. Generating abnormal fault prediction data can discover potential fault problems in advance and reduce losses and maintenance time. Through iterative fault diagnosis decisions, power supply faults can be located quickly and accurately, and fault handling efficiency can be improved. Constructing an optimal diagnostic strategy visualization view helps to intuitively display the fault diagnosis results and provide visualization support for decision-making.

[0013] Preferably, step S1 comprises the following steps:

[0014] Step S11: Obtain the power supply historical operation status log and AC power supply component structure diagram;

[0015] Step S12: Fill missing values ​​in the historical operation status log of the power supply to obtain a standardized operation status log;

[0016] Step S13: performing sequential load operation analysis on the standardized operation status log to generate a sequential load operation status feature of the power supply;

[0017] Step S14: sampling the power supply timing load operation state characteristics at multiple time points to obtain the load operation states at multiple time points;

[0018] Step S15: Calculating the load demand for the load operating status at multiple time points, thereby obtaining multi-stage load demand data of the power supply;

[0019] Step S16: performing load demand prediction on the multi-stage load demand data of the power supply to generate a load demand prediction curve.

[0020] By filling in missing values, the present invention can ensure the integrity and accuracy of the data and provide a reliable data basis for subsequent analysis. Standardized data helps to unify the data format, simplify the subsequent analysis process, and improve the comparability and interpretability of the data. By analyzing the time-series load operation status, the load changes at different time points can be understood, which is helpful to discover potential problems and optimize the operation strategy. These characteristics help to reveal the laws and characteristics of power load operation and provide an important basis for subsequent analysis. Through multi-time point sampling, the changes in the power load operation status can be more comprehensively understood, providing more data support for subsequent load demand calculations. According to the load operation status at multiple time points, the load demand is calculated, which helps to accurately grasp the changes in the demand of the power system and provide basic data for subsequent analysis. These data reflect the demand of the power supply at different stages and provide an important basis for subsequent predictions. By predicting the load demand, the power supply strategy can be planned in advance, the equipment operation efficiency can be optimized, and the stability and reliability of the power system can be ensured.

[0021] Preferably, step S2 comprises the following steps:

[0022] Step S21: performing real-time operation monitoring of the target AC power source based on the sensor array to obtain multi-dimensional operation monitoring parameters, wherein the multi-dimensional operation monitoring parameters include operation temperature monitoring parameters and multi-position sound wave signals;

[0023] Step S22: performing temperature distribution analysis on the operating temperature monitoring parameters to identify power supply operating temperature distribution data;

[0024] Step S23: Calculating the temperature fluctuation trend of the operating temperature monitoring parameters to generate a power supply temperature fluctuation trend;

[0025] Step S24: fitting the temperature difference fluctuation between components to the power supply temperature fluctuation trend based on the AC power supply component structure diagram, and constructing a three-dimensional temperature difference change curve;

[0026] Step S25: dynamically rendering the three-dimensional temperature difference change curve based on the power supply operation temperature distribution data to construct a three-dimensional temperature distribution field of the power supply.

[0027] The present invention monitors the target AC power supply in real time through a sensor array, obtains multi-dimensional operation monitoring parameters, including operation temperature monitoring parameters and multi-part acoustic wave signals, which helps to fully understand the operation status of the power supply. Analyzing the operation temperature monitoring parameters and identifying the temperature distribution data of the power supply can help find abnormal temperature areas and evaluate the thermal effect of the power supply. By calculating the temperature fluctuation trend, the changing law of the power supply temperature can be revealed, which helps predict potential temperature problems and provides a basis for subsequent maintenance and management. Based on the AC power supply component structure diagram, the temperature fluctuation trend of the power supply is fitted with the temperature difference fluctuation between components, and a three-dimensional temperature difference change curve is constructed, which helps to analyze the temperature difference between different components. Based on the power supply operation temperature distribution data, the three-dimensional temperature difference change curve is dynamically rendered with temperature distribution, and a three-dimensional temperature distribution field of the power supply is constructed, which can intuitively display the internal temperature distribution of the power supply and help monitor and manage the thermal state of the power supply.

[0028] Preferably, the specific steps of step S24 are:

[0029] Perform component visual identification on the AC power component structure diagram and mark multiple power component nodes;

[0030] Calculate the spatial position of multiple power component nodes and generate the spatial position coordinates of each component node;

[0031] Based on the spatial position coordinates of each component node, the operating temperature monitoring parameters are spatially correlated and matched to obtain the temperature monitoring value of each component node;

[0032] Based on the power supply temperature fluctuation trend, the temperature monitoring value of each component node is used to calculate the three-dimensional temperature difference between components, and generate three-dimensional temperature difference data between components;

[0033] Define the time interval for temperature fluctuations;

[0034] Perform temperature difference fluctuation analysis on the temperature difference data between three-dimensional components based on the temperature fluctuation time interval, and generate temperature difference fluctuation data of components at multiple intervals;

[0035] The temperature difference fluctuation data of components at multiple intervals are discretely fitted to construct a three-dimensional temperature difference change curve.

[0036] The present invention can accurately mark the node position of the power supply component through visual recognition, and provide an accurate data basis for subsequent spatial position calculation. Marking the nodes of the power supply component is helpful to understand the overall structure, and ensure that each component can be accurately identified and monitored during the monitoring process. The coordinates generated by the spatial position calculation can accurately describe the position of each component node, and provide an accurate spatial reference for subsequent temperature monitoring and analysis. The generation of spatial position coordinates is helpful to establish the spatial relationship between component nodes, and provide a basis for the correlation matching of temperature monitoring data. Spatial position correlation matching can ensure that the correct temperature monitoring parameters are associated with the corresponding component nodes, and realize temperature monitoring of each node. Obtaining the temperature monitoring value of each component node is helpful to realize real-time monitoring and remote monitoring of the operating status of the power supply component. Through the temperature difference calculation between three-dimensional components, the temperature difference between different component nodes can be analyzed. Temperature difference conditions help to discover abnormal changes and hot spots. The generated temperature difference data helps to optimize the configuration and operation of power components and improve the performance and stability of the overall system. Defining the temperature fluctuation time interval helps to determine the frequency of monitoring data collection and ensure timely response and monitoring of temperature fluctuations. The definition of time intervals can affect the efficiency of data collection and processing, ensuring the accuracy and real-time nature of data analysis. Temperature difference fluctuation analysis helps to monitor the changing trend of temperature differences between components and identify the laws and abnormalities of temperature fluctuations. The generated temperature difference fluctuation data can be used in the early warning system to help timely discover potential faults and problems. Discrete fitting and curve construction help to convert temperature difference fluctuation data into a visual form to more intuitively display the temperature change trend. Constructing a three-dimensional temperature difference change curve can help system administrators track temperature changes of power components and make timely adjustments and countermeasures.

[0037] Preferably, step S3 specifically comprises the following steps:

[0038] Step S31: performing low-frequency filtering on the sound wave signals of multiple parts to obtain filtered and noise-reduced sound wave signals;

[0039] Step S32: performing clock-synchronous digital processing on the filtered noise-reduced waveform signal to generate a noise-reduced waveform data sequence;

[0040] Step S33: performing sliding time segmentation processing on the noise reduction sound wave data sequence to obtain multiple sound wave sequences;

[0041] Step S34: performing power supply aging status prediction on the multiple sound wave sequences to generate power supply life status prediction data.

[0042] The present invention can effectively remove high-frequency noise through low-frequency filtering, improve the quality and accuracy of sound wave signals, the sound wave signals after filtering are clearer, which is helpful to accurately identify and analyze the signals sent by power supply equipment, clock synchronization processing is helpful to ensure the consistency and accuracy of data, and improve the reliability of data processing, and digitizing the signal can facilitate storage and processing, providing a better data basis for subsequent analysis, and sound wave sequences of different time periods can be obtained through segmentation processing, which is helpful to analyze the changes in sound wave characteristics of power supply equipment in different time periods, and segmentation processing is helpful to organize data, making the analysis of sound wave signals more organized and systematic, and the aging state of power supply equipment can be predicted through sound wave signal analysis, which helps to discover potential problems in advance and take corresponding measures, and the generated power supply life state prediction data can be used to evaluate the life status of power supply equipment and provide a basis for maintenance and replacement.

[0043] Preferably, the specific steps of step S34 are:

[0044] Decompose multiple sound wave sequences segment by segment in the frequency domain to obtain a frequency domain diagram of each sound wave sequence;

[0045] Perform vibration frequency characteristic analysis on the frequency domain graph of each sound wave sequence to extract the vibration frequency characteristics of each sound wave sequence;

[0046] Calculating the time series amplitude change of each sound wave vibration frequency characteristic to generate sound wave time series amplitude change data;

[0047] Extract the vibration frequency peak of each sound wave sequence based on the frequency domain diagram of each sound wave sequence;

[0048] Perform acoustic wave peak gradient fitting on the vibration frequency peak of each acoustic wave sequence to construct an acoustic wave peak gradient curve;

[0049] Calculating the amplitude decay rate of the acoustic wave peak gradient curve based on the acoustic wave time series amplitude change data to generate the acoustic wave amplitude decay rate;

[0050] Conduct in-depth mining of the attenuation change of the acoustic wave amplitude attenuation rate to generate the attenuation change law of the power supply acoustic wave;

[0051] The power supply aging status is predicted based on the power supply acoustic wave attenuation change law to generate power supply life status prediction data.

[0052] The present invention can help understand the characteristics of sound wave signals at different frequencies through frequency domain decomposition, and help to mine the hidden information in the sound wave signals. The frequency domain diagram obtained can intuitively display the distribution of sound waves in the frequency domain, providing a basis for subsequent frequency feature analysis. Extracting the vibration frequency feature can help determine the main vibration frequency components in each sound wave sequence, providing an important basis for subsequent analysis. The time series amplitude change data reflects the change of the sound wave vibration frequency characteristics over time, which helps to understand the dynamic change of the sound wave signal. Extracting the vibration frequency peak can accurately capture the most significant vibration frequency component in each sound wave sequence, providing important data for subsequent analysis. The gradient fitting can reveal the change trend of the sound wave vibration frequency peak, which helps to analyze the frequency feature evolution law of the sound wave signal. The amplitude decay rate reflects the decay speed of the sound wave signal amplitude over time, which is an important parameter for evaluating the change of the sound wave signal. Deeply mining the change of the decay rate can reveal the regularity of the power supply sound wave attenuation, which helps to understand the state change trend of the power supply equipment. The aging state of the power supply equipment can be predicted through the attenuation law of the sound wave signal, providing an important basis for maintenance and management.

[0053] Preferably, the specific steps of step S4 are:

[0054] Step S41: performing component topology association analysis on the AC power component structure diagram to generate component topology association data;

[0055] Step S42: Perform three-dimensional point cloud modeling on the component topology association data to construct a three-dimensional power supply topology structure model;

[0056] Step S43: performing life state mapping on the three-dimensional power supply topology structure model based on the power supply life state prediction data to generate a state mapping structure model;

[0057] Step S44: dynamically evolve the temperature distribution of the state mapping structure model based on the three-dimensional temperature distribution field of the power supply, and construct a dynamic twin evolution model.

[0058] By analyzing the topological relationship between components, the present invention can determine the connection mode and dependency relationship between components, provide basic data for subsequent modeling and analysis, and map the power supply life status prediction data to the three-dimensional power supply topological structure model, which can intuitively display the life status of each component and provide a basis for monitoring and analysis. By simulating the three-dimensional temperature distribution field of the power supply, the dynamic evolution of the temperature of the state mapping structure model can be realized, which helps to monitor the temperature changes of the power supply system. The construction of a dynamic twin evolution model can compare the evolution process of the actual system with the model, thereby more accurately evaluating the state changes of the power supply system.

[0059] Preferably, the specific steps of step S5 are:

[0060] Step S51: simulating the load demand evolution of the dynamic twin evolution model according to the load demand prediction curve, and extracting power supply simulation response data;

[0061] Step S52: identifying the temperature field change of each component on the power supply simulation response data, and generating simulation data of the temperature field change of each component;

[0062] Step S53: performing component degradation accumulation calculation on the power supply simulation response data to obtain the degradation accumulation value of each component;

[0063] Step S54: calculating the component damage risk of each component based on the degradation accumulation value of each component according to the temperature field change simulation data of each component, thereby obtaining the damage risk data of each component;

[0064] Step S55: performing abnormal failure prediction on the damage risk data of each component to generate abnormal failure prediction data.

[0065] The present invention can predict the response of the power supply system under different loads by simulating the change of load demand, and provide a basis for system performance evaluation. The extracted power supply simulation response data can reflect the real-time response of the power supply system to the change of load demand, which is helpful to monitor the operation status of the system. Identifying the change of temperature field of each component can help evaluate the thermal condition of the power supply system under different loads and provide a reference for temperature management. The generated simulation data reflects the change of temperature field of each component, which is helpful to monitor the heat distribution of components in the system. By calculating the cumulative degradation value of the component, the health status of each component can be evaluated to provide a basis for preventive maintenance. Combining the temperature field change simulation data and the cumulative degradation value, the damage risk of each component is calculated, which is helpful to predict potential failures and take corresponding measures. By analyzing the damage risk data of the component, possible abnormal failure conditions can be predicted to provide a basis for early prevention and handling of abnormal conditions.

[0066] Preferably, the specific steps of step S6 are:

[0067] Step S61: performing fault location on the dynamic twin evolution model based on the abnormal fault prediction data to obtain the abnormal fault location point;

[0068] Step S62: Generate a remote abnormal warning signal according to the abnormal fault location point;

[0069] Step S63: performing iterative fault diagnosis decisions based on the remote abnormal warning signal to generate multiple fault diagnosis strategies;

[0070] Step S64: performing optimal collaborative analysis based on multiple fault diagnosis strategies to extract the optimal fault diagnosis strategy;

[0071] Step S65: Perform diagnostic visualization on the optimal fault diagnosis strategy and construct a visualization view of the optimal diagnosis strategy.

[0072] The present invention can use abnormal fault prediction data to locate faults in the dynamic twin evolution model and determine possible abnormal points in the system, which is helpful to quickly locate the specific location of potential problems. The generation of remote abnormal warning signals can promptly notify monitoring personnel or system administrators of abnormal situations and help them take corresponding measures to avoid further deterioration of the fault. Through iterative fault diagnosis decisions, multiple fault diagnosis strategies can be generated to improve the accuracy and comprehensiveness of diagnosis of abnormal situations. Through collaborative analysis of multiple fault diagnosis strategies, the optimal fault diagnosis strategy can be extracted to more effectively deal with abnormal situations. By constructing a visual view of the optimal diagnostic strategy, the results of fault diagnosis and the decision-making process can be intuitively presented, providing decision makers with intuitive reference information.

[0073] In this specification, a system for remotely monitoring the operating status of an AC power supply is provided, which is used to execute the method for remotely monitoring the operating status of an AC power supply as described above, including:

[0074] The load demand prediction module is used to obtain the power supply historical operation status log and the AC power supply component structure diagram; perform time-series load operation analysis on the power supply historical operation status log, and perform load demand prediction to generate a load demand prediction curve;

[0075] The temperature distribution module is used to obtain multi-dimensional operation monitoring parameters, which include operation temperature monitoring parameters and multi-position acoustic wave signals; dynamically render the operation temperature monitoring parameters to construct a three-dimensional temperature distribution field of the power supply;

[0076] The life prediction module is used to perform low-frequency filtering on the acoustic wave signals of multiple parts and predict the power supply aging status to generate power supply life status prediction data;

[0077] Dynamic twin module, which is used to dynamically evolve the temperature distribution of the AC power component structure diagram based on the power life status prediction data and the three-dimensional temperature distribution field of the power supply, and build a dynamic twin evolution model;

[0078] An abnormal fault prediction module is used to simulate the load demand evolution of the dynamic twin evolution model according to the load demand prediction curve, and to predict abnormal faults to generate abnormal fault prediction data;

[0079] The fault diagnosis decision module is used to make iterative fault diagnosis decisions based on abnormal fault prediction data, perform diagnosis visualization, and build a visualization view of the optimal diagnosis strategy.

[0080] By analyzing the historical operation status log of the power supply, the present invention can better understand the operation status of the system, and predict the load demand based on the historical data, which is helpful to optimize resource allocation and plan system capacity. Through multi-dimensional monitoring parameters, the power supply operation status, especially the temperature distribution, can be understood in real time. Dynamic temperature distribution rendering and three-dimensional temperature distribution field construction are helpful to early detection of potential overheating problems so that preventive maintenance measures can be taken. Through acoustic signal processing and life prediction, the aging trend of the power supply can be identified in advance, which is helpful to plan maintenance plans and preventive replacement of key components. Based on the life status prediction data and the temperature distribution field, a dynamic twin evolution model is constructed, which is helpful to monitor the state changes of the power supply components in real time and predict future situations. Through load demand evolution simulation and abnormal fault prediction, potential problems can be discovered in advance and the impact of system faults on operation can be reduced. Iterative fault diagnosis decisions based on abnormal fault prediction data are helpful to improve diagnostic accuracy. Diagnostic visualization and optimal strategy construction provide intuitive information to support decision makers to quickly formulate response plans. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 A schematic diagram of the steps of a method for remotely monitoring the operating status of an AC power supply according to the present invention;

[0082] Figure 2 Detailed implementation flow chart of step S1;

[0083] Figure 3 Detailed implementation flow chart of step S2;

[0084] Figure 4 Detailed implementation flow chart of step S3. DETAILED DESCRIPTION

[0085] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0086] The present application example provides a method and system for remote monitoring of the operating status of an AC power supply. The execution subject of the method and system for remote monitoring of the operating status of an AC power supply includes but is not limited to: mechanical equipment, data processing platform, cloud server node, network upload device, etc. equipped with the system can be regarded as a general computing node of the present application, and the data processing platform includes but is not limited to: at least one of an audio and image management system, an information management system, and a cloud data management system.

[0087] See also Figures 1 to 4 The present invention provides a method for remotely monitoring the operating status of an AC power supply, the method comprising the following steps:

[0088] Step S1: Obtain a power supply historical operation status log and an AC power supply component structure diagram; perform a time series load operation analysis on the power supply historical operation status log, and perform load demand prediction to generate a load demand prediction curve;

[0089] Step S2: acquiring multi-dimensional operation monitoring parameters, wherein the multi-dimensional operation monitoring parameters include operation temperature monitoring parameters and multi-position acoustic wave signals; performing dynamic temperature distribution rendering on the operation temperature monitoring parameters to construct a three-dimensional temperature distribution field of the power supply;

[0090] Step S3: performing low-frequency filtering on the acoustic wave signals of multiple parts and predicting the power supply aging state to generate power supply life state prediction data;

[0091] Step S4: dynamically evolve the temperature distribution of the AC power component structure diagram based on the power life state prediction data and the three-dimensional temperature distribution field of the power supply, and construct a dynamic twin evolution model;

[0092] Step S5: simulating the load demand evolution of the dynamic twin evolution model according to the load demand prediction curve, and performing abnormal fault prediction to generate abnormal fault prediction data;

[0093] Step S6: Make iterative fault diagnosis decisions based on abnormal fault prediction data, perform diagnosis visualization, and build a visualization view of the optimal diagnosis strategy.

[0094] By analyzing the historical operation status log of the power supply, the present invention can understand the usage of the load, which is helpful to optimize the power supply strategy. Generating a load demand prediction curve is helpful to predict the changes in the load in advance so as to make corresponding adjustments and arrangements. Constructing a three-dimensional temperature distribution field of the power supply is helpful to monitor and manage the temperature changes in the operation of the power supply and improve the understanding of the power supply status. Through the power supply life status prediction data, it is helpful to predict the aging of the power supply components in advance, so as to take corresponding maintenance and replacement measures. Constructing a dynamic twin evolution model can help monitor the temperature distribution evolution of the power supply structure in real time, improve the real-time and accuracy of the power supply operation status, and simulate the load demand evolution according to the load demand prediction curve to understand the future load situation and provide a basis for adjusting the power supply plan. Generating abnormal fault prediction data can discover potential fault problems in advance and reduce losses and maintenance time. Through iterative fault diagnosis decisions, power supply faults can be located quickly and accurately, and fault handling efficiency can be improved. Constructing an optimal diagnostic strategy visualization view helps to intuitively display the fault diagnosis results and provide visualization support for decision-making.

[0095] In the embodiment of the present invention, refer to Figure 1 , is a flowchart of a method for remotely monitoring the operating status of an AC power supply according to the present invention. In this example, the method for remotely monitoring the operating status of an AC power supply includes the following steps:

[0096] Step S1: Obtain a power supply historical operation status log and an AC power supply component structure diagram; perform a time series load operation analysis on the power supply historical operation status log, and perform load demand prediction to generate a load demand prediction curve;

[0097] In this embodiment, historical operation status logs are extracted from the power management system or monitoring system. These logs usually include timestamps, load current, voltage, power factor, temperature and other data. The collected log data is organized into a unified format, such as CSV or a database table, for subsequent analysis. It is ensured that each record contains time, various parameters and units. Missing values, abnormal values ​​and duplicate records in the log data are checked and processed accordingly to ensure the accuracy and completeness of the data. The component structure diagram of the AC power supply is obtained. It can usually be extracted from the design documents or engineering drawings. Ensure that the drawings are clear and readable, including all important components and their connection relationships. If the structure diagram is a paper version, use a scanner or image processing software to digitize it and convert it into an editable electronic file format (such as PDF or CAD format). The historical operation status log is cleaned to remove missing values ​​and abnormal values ​​to ensure data quality. Pandas and other libraries are used for data cleaning and processing. Time series analysis methods (such as moving average, trend analysis, etc.) are used to analyze load data to identify patterns and periodic fluctuations in load changes. Data analysis tools (such as Python's StatsModels library or R language) are used for time series analysis to generate visualization charts of load operation (such as line charts or bar charts). According to the characteristics of time series load data, appropriate load demand forecasting models are selected, such as ARIMA model, SARIMA model or machine learning model (such as linear regression, random forest, etc.). Historical load data is divided into training sets and test sets. The selected model is trained with the training set to ensure that the model can effectively fit the data. The trained model is used to predict future load demand and generate a load demand forecasting curve. Different forecasting time ranges (such as day, week, month) can be selected according to specific needs.

[0098] Step S2: acquiring multi-dimensional operation monitoring parameters, wherein the multi-dimensional operation monitoring parameters include operation temperature monitoring parameters and multi-position acoustic wave signals; performing dynamic temperature distribution rendering on the operation temperature monitoring parameters to construct a three-dimensional temperature distribution field of the power supply;

[0099] In this embodiment, the sources of the operating temperature monitoring parameters and the acoustic wave signals of multiple parts are confirmed, including monitoring equipment, sensors or data acquisition systems, and the operating temperature monitoring parameters and acoustic wave signals are extracted from the monitoring system using data extraction tools (such as SQL queries, API calls or data export tools). The extracted data formats are ensured to be consistent, including timestamps, temperature values, acoustic wave signal amplitudes, etc. Common formats include CSV, JSON or database tables. The operating temperature monitoring parameters are cleaned, missing values ​​and outliers are removed, and data quality is ensured. Pandas and other libraries can be used for data cleaning and processing, and the temperature monitoring data is organized into a format suitable for three-dimensional rendering, including the coordinates of each measurement point (x, y, z) and the corresponding temperature values, use 3D modeling tools (such as MATLAB, Python's Matplotlib, Mayavi or ParaView) to construct a 3D temperature distribution field, apply interpolation algorithms (such as Kriging interpolation or cubic spline interpolation) to interpolate the scattered temperature data to generate a continuous temperature distribution field, perform dynamic rendering through heat maps or color maps, use color coding to represent different temperature areas, ensure that the rendering can show the change of temperature over time, form a dynamic temperature distribution field, visualize the constructed 3D temperature distribution field, and save it in a suitable file format (such as 3D model files or dynamic images) for subsequent analysis and display.

[0100] Step S3: performing low-frequency filtering on the acoustic wave signals of multiple parts and predicting the power supply aging state to generate power supply life state prediction data;

[0101] In this embodiment, acoustic signal data is collected from multiple sensors to ensure that the data acquisition frequency and time of each sensor are synchronized for subsequent analysis. The collected acoustic signal data is organized into a unified format (such as a CSV file) to ensure that key information such as timestamps and acoustic amplitudes are included. Determine the type of low-frequency filter, such as a Butterworth filter or a Chebyshev filter, and set a suitable cutoff frequency to remove high-frequency noise. The choice of cutoff frequency usually depends on the signal characteristics (for example, less than 100 Hz). Apply digital signal processing tools (such as the SciPy library in Python) to perform low-frequency filtering on the original acoustic signal to remove high-frequency noise. In this step, the generated filtered signal should maintain the main features of the original signal while eliminating unnecessary noise. Compare the signal waveforms before and after filtering to ensure that the filtering process is effective. Time domain and frequency domain diagrams can be used for visualization to check the spectral changes of the signal. Extract key features from the filtered acoustic signal, which may include: time domain features: mean, standard deviation, peak, root mean square (RMS), etc. Frequency domain features: spectrum analysis, main frequency components, etc. Train the selected model using historical data (including acoustic signal features and known aging states). Ensure the division of training and test sets to evaluate the performance of the model. Apply the trained model to the extracted acoustic signal features to generate power supply aging state prediction data. These data will reflect the current and future health status of the power supply. Verify the accuracy of the prediction results and compare the prediction results with the actual aging state to evaluate the prediction ability of the model. Confusion matrix, accuracy, recall and other indicators can be used for evaluation. Store the generated power supply life state prediction data in a structured manner, such as a database or CSV file, for subsequent analysis and visualization.

[0102] Step S4: dynamically evolve the temperature distribution of the AC power component structure diagram based on the power life state prediction data and the three-dimensional temperature distribution field of the power supply, and construct a dynamic twin evolution model;

[0103] In this embodiment, the power supply life status prediction data is collected, including the remaining useful life (RUL), aging degree and related confidence of each component. The prediction data format is ensured to be consistent, including fields such as timestamp, component ID, RUL, etc., usually in CSV, JSON or database table format. The previously constructed three-dimensional temperature distribution field data of the power supply is obtained to ensure that the data contains the temperature value, position coordinates (x, y, z) and timestamp of each component. The temperature distribution data is organized into a format suitable for dynamic evolution to ensure that the temperature data at each time point can be associated with the corresponding component. A suitable dynamic twin modeling framework is selected, such as MATLAB. Simulink, Python's SimPy library or other 3D modeling software (such as Blender) can be used to combine the power component structure diagram with the temperature distribution data to ensure that each component has a corresponding temperature value and aging status in the model. Dynamic evolution rules can be set in the modeling software to simulate the change of temperature distribution over time. The heat conduction equation or finite element analysis (FEA) method can be used to calculate the temperature change of each component. The power life state prediction data is used as input to affect the temperature evolution process in the model. For example, components with a higher degree of aging may exhibit higher operating temperatures. Historical data (such as actual temperature monitoring data) is used to compare with the model output to verify the accuracy of the dynamic twin evolution model. The model parameters or evolution rules are adjusted according to the verification results to ensure that the model can accurately reflect the dynamic behavior of the power system.

[0104] Step S5: simulating the load demand evolution of the dynamic twin evolution model according to the load demand prediction curve, and performing abnormal fault prediction to generate abnormal fault prediction data;

[0105] In this embodiment, the load demand prediction curve is input into the dynamic twin evolution model to ensure that the model can receive real-time load data and make corresponding adjustments. The established dynamic twin model is used to simulate the evolution of the load demand. The dynamic load simulation method can be applied to reflect the response of the power supply system under different load conditions, and the operating status changes of each component under the simulation process are recorded, including indicators such as temperature, power consumption, and efficiency, so as to facilitate subsequent analysis and select a suitable abnormal fault prediction model, such as a machine learning-based model (such as a decision tree, random forest, support vector machine) or a statistical model (such as time series analysis). The operating status data obtained from the load demand simulation and the historical fault data are integrated to provide a basis for model training. These data should include temperature and power consumption. The selected abnormal fault prediction model is trained using the training set based on the characteristics of load, speed, and running time to ensure that the model can learn the relationship between load changes and faults. The simulated state data is input into the trained model to predict abnormal faults and generate abnormal fault prediction data, including the failure probability and possible failure types of each component. The generated abnormal fault prediction data is stored in a data frame or database and visualized to facilitate subsequent analysis and decision-making. The accuracy of the abnormal fault prediction model is evaluated using the test set, and the prediction results are compared with the actual fault records. The accuracy rate, recall rate and other indicators are calculated. According to the abnormal fault prediction data, further analysis and decision support are carried out to identify high-risk components and formulate corresponding maintenance strategies.

[0106] Step S6: Make iterative fault diagnosis decisions based on abnormal fault prediction data, perform diagnosis visualization, and build a visualization view of the optimal diagnosis strategy.

[0107] In this embodiment, an iterative fault diagnosis decision framework is designed to clarify the decision-making steps and standards, including fault identification, impact assessment, processing strategy generation and feedback mechanism, determine the standards for evaluating faults, such as fault probability, impact degree, repair cost, etc., to ensure that the decision-making process is scientific and reasonable, identify possible fault types and their corresponding processing strategies based on abnormal fault prediction data, such as inspection and repair, component replacement, monitoring and recording, evaluate each strategy, and consider factors including implementation difficulty, required time, cost and effectiveness, record the generated fault diagnosis strategy in a data frame, attach the advantages and disadvantages of each strategy and applicable conditions, select a suitable optimization method (such as linear programming, genetic algorithm or multi-objective optimization) to evaluate and extract the optimal diagnosis strategy, use the selected optimization method to evaluate all strategies, extract the optimal strategy that can achieve the best balance between cost, time and effect, record the optimal diagnosis strategy in a data frame, ensure that detailed implementation steps and expected effects are included, select a suitable visualization tool (such as Tableau, Power BI, D3.js, or Python's Matplotlib) to build visualizations. Design visualizations that display various aspects of the optimal diagnostic strategy, including: Evaluation metrics for each strategy (such as cost, time, and effectiveness) Association between fault prediction data and decision results Use the selected visualization tools to create interactive visualizations that allow users to intuitively understand the decision process and results.

[0108] In this embodiment, refer to Figure 2 , is a flowchart of detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include:

[0109] Step S11: Obtain the power supply historical operation status log and AC power supply component structure diagram;

[0110] Step S12: Fill missing values ​​in the historical operation status log of the power supply to obtain a standardized operation status log;

[0111] Step S13: performing sequential load operation analysis on the standardized operation status log to generate a sequential load operation status feature of the power supply;

[0112] Step S14: sampling the power supply timing load operation state characteristics at multiple time points to obtain the load operation states at multiple time points;

[0113] Step S15: Calculating the load demand for the load operating status at multiple time points, thereby obtaining multi-stage load demand data of the power supply;

[0114] Step S16: performing load demand prediction on the multi-stage load demand data of the power supply to generate a load demand prediction curve.

[0115] In this embodiment, historical power operation status logs are extracted from the power management system or monitoring system. These logs should include timestamps, operating parameters (such as voltage, current, power, etc.) and fault records. A structural diagram of the AC power component is obtained to ensure a clear understanding of the composition of the power system, which usually includes the identification and layout of components such as generators, transformers, and distribution boards. The obtained historical operation status logs and structural diagrams are stored in a database or document management system for subsequent analysis. Data analysis tools (such as Pandas) are used to detect missing values ​​in the historical operation status logs, and the location and number of missing values ​​are recorded. A suitable missing value filling method is selected, such as mean filling, linear interpolation, or forward filling. A suitable method is selected according to the nature of the data. The filled operation status logs are standardized (such as Z-score standardization) to ensure that the data are compared and analyzed on the same scale. A suitable time series analysis method is selected, such as an autoregressive model (AR), a moving average (MA), or seasonal decomposition. Changes in the load operation status are analyzed, and key features of the load operation status are calculated. Such as peak value, valley value, average load and load volatility, etc., to form the characteristic data of power supply time series load operation status, determine the strategy of load status sampling at multiple time points, ensure coverage of the entire time series data (such as specific time points every hour and every day), extract the load status at the specified time point from the power supply time series load operation status characteristics, record the characteristic data of these time points, store the sampled load operation status at multiple time points in an array or data frame for subsequent analysis, select a suitable load demand calculation model (such as linear regression, time series regression, etc.), calculate the load demand according to the time series load characteristics, use the sampled load status data at multiple time points to calculate the load demand, generate power supply multi-stage load demand data, select a suitable load demand prediction model according to the data characteristics, such as ARIMA (autoregressive integrated moving average model), LSTM (long short-term memory network), etc., use historical load demand data to train the selected prediction model, and perform cross-validation to ensure the accuracy of the model, predict future load demand based on the trained model, and generate a load demand prediction curve.

[0116] In this embodiment, refer to Figure 3 , is a flowchart of detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include:

[0117] Step S21: performing real-time operation monitoring of the target AC power source based on the sensor array to obtain multi-dimensional operation monitoring parameters, wherein the multi-dimensional operation monitoring parameters include operation temperature monitoring parameters and multi-position sound wave signals;

[0118] Step S22: performing temperature distribution analysis on the operating temperature monitoring parameters to identify power supply operating temperature distribution data;

[0119] Step S23: Calculating the temperature fluctuation trend of the operating temperature monitoring parameters to generate a power supply temperature fluctuation trend;

[0120] Step S24: fitting the temperature difference fluctuation between components to the power supply temperature fluctuation trend based on the AC power supply component structure diagram, and constructing a three-dimensional temperature difference change curve;

[0121] Step S25: dynamically rendering the three-dimensional temperature difference change curve based on the power supply operation temperature distribution data to construct a three-dimensional temperature distribution field of the power supply.

[0122] In this embodiment, a sensor array is deployed to ensure coverage of multiple operating parts of the target AC power supply to obtain comprehensive monitoring data, collect operating temperature monitoring parameters (such as the temperature of heating elements, transformers, etc.) and multi-part acoustic wave signals (such as using acoustic sensors to monitor noise and vibration) in real time, store the acquired multi-dimensional operating monitoring parameters in a database, ensure that the data is organized by timestamp for subsequent analysis, use statistical methods (such as histograms or kernel density estimation) to analyze temperature data, identify temperature distribution characteristics, calculate average temperature, maximum value, minimum value, etc., record temperature distribution data in a data frame for subsequent analysis and visualization, organize temperature monitoring parameters in chronological order to ensure data continuity, calculate temperature fluctuation trends, and use Use sliding average or exponential smoothing methods to smooth data, identify the rising and falling trends of temperature, identify the temperature monitoring points and corresponding positions of each component according to the power component structure diagram, calculate the temperature difference between different components, select a suitable fitting method (such as polynomial fitting or spline fitting) to model the temperature difference fluctuation, generate a three-dimensional temperature difference change curve, integrate the power supply operation temperature distribution data with the three-dimensional temperature difference change curve to obtain the necessary parameters for dynamic temperature distribution, select a suitable rendering tool (such as Matplotlib's 3D drawing or professional visualization software) to construct a three-dimensional temperature distribution field, use the rendering tool to generate a dynamic temperature distribution field, display the temperature conditions and changing trends of different parts, and ensure that the visualization effect can reflect real-time data.

[0123] In this embodiment, the specific steps of step S24 are:

[0124] Perform component visual identification on the AC power component structure diagram and mark multiple power component nodes;

[0125] Calculate the spatial position of multiple power component nodes and generate the spatial position coordinates of each component node;

[0126] Based on the spatial position coordinates of each component node, the operating temperature monitoring parameters are spatially correlated and matched to obtain the temperature monitoring value of each component node;

[0127] Based on the power supply temperature fluctuation trend, the temperature monitoring value of each component node is used to calculate the three-dimensional temperature difference between components, and generate three-dimensional temperature difference data between components;

[0128] Define the time interval for temperature fluctuations;

[0129] Perform temperature difference fluctuation analysis on the temperature difference data between three-dimensional components based on the temperature fluctuation time interval, and generate temperature difference fluctuation data of components at multiple intervals;

[0130] The temperature difference fluctuation data of components at multiple intervals are discretely fitted to construct a three-dimensional temperature difference change curve.

[0131] In this embodiment, a computer vision tool (such as OpenCV) is used to process the AC power component structure diagram, identify each component, and mark the identified power component nodes, including generators, transformers, distribution boards, etc., to ensure that each component node has a unique identifier. The marked component node information is stored in a database, including the component type and the corresponding identifier, for subsequent processing. A suitable three-dimensional coordinate system (such as a Cartesian coordinate system) is defined to represent the spatial position of the component, and the spatial position of each component node is calculated to obtain its coordinates in the three-dimensional space (x, y, z), which can be calculated by the scale and known reference points in the image, organize the operating temperature monitoring parameters, ensure that the data is complete and sorted by timestamp, use a spatial matching algorithm (such as the nearest neighbor algorithm) to associate and match the temperature monitoring parameters of each component with its corresponding spatial position coordinates, extract the corresponding temperature monitoring value of each component node according to the matching result, store the temperature monitoring value of each component node together with its spatial coordinates in a data frame, calculate the temperature difference for each pair of component nodes, create a three-dimensional array or data frame, record the temperature difference data between each pair of components, define the time interval for temperature fluctuation analysis according to actual needs, such as every minute, every hour or a custom time period, select a suitable fluctuation analysis method (such as variance analysis, moving average) to analyze the temperature difference data between three-dimensional components, calculate the temperature difference fluctuation within each time interval, identify the temperature difference change trend, select a suitable fitting method (such as spline fitting, polynomial fitting) to process the temperature difference fluctuation data of components with multiple interval points, apply the fitting algorithm to generate a three-dimensional temperature difference change curve, ensure that the fitting curve can accurately reflect the temperature difference change trend, store the fitting results, and use visualization tools (such as Matplotlib) to draw the three-dimensional temperature difference change curve for easy analysis and display.

[0132] In this embodiment, refer to Figure 4 , is a flowchart of detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include:

[0133] Step S31: performing low-frequency filtering on the sound wave signals of multiple parts to obtain filtered and noise-reduced sound wave signals;

[0134] Step S32: performing clock-synchronous digital processing on the filtered noise-reduced waveform signal to generate a noise-reduced waveform data sequence;

[0135] Step S33: performing sliding time segmentation processing on the noise reduction sound wave data sequence to obtain multiple sound wave sequences;

[0136] Step S34: performing power supply aging status prediction on the multiple sound wave sequences to generate power supply life status prediction data.

[0137] In this embodiment, a low-frequency filter (such as a Butterworth filter or a Chebyshev filter) is designed, and a suitable cutoff frequency is selected to remove high-frequency noise. A digital signal processing library (such as signal.butter and signal.filtfilt in SciPy) is used to filter the acoustic wave signals of multiple parts to obtain a filtered noise-reduced acoustic wave signal, and it is ensured that each acoustic wave signal data point has a corresponding timestamp to facilitate subsequent synchronization and analysis. The filtered noise-reduced acoustic wave signal is digitized, and usually the analog signal is converted into a digital signal. A quantization method (such as uniform quantization) is used to generate a noise-reduced acoustic wave data sequence, and it is ensured that the data format is unified (such as a Numpy array) for subsequent analysis and processing. The size of the sliding window (such as every 10 seconds or every 30 seconds) and the sliding step size (such as every 5 seconds) are determined, and the noise-reduced acoustic wave signal is digitized using the sliding window technology. The data sequence can be segmented. The slicing operation in Python can be used to extract the data of each window. Each segmented sound wave sequence is stored in a list or data frame for subsequent analysis. The start and end time of each sound wave sequence is recorded to facilitate subsequent tracking and analysis. A suitable prediction model (such as machine learning model: random forest, support vector machine, or deep learning model: LSTM) is selected to analyze and predict the aging status of the power supply. Features (such as frequency domain features, time domain features) are extracted from multiple sound wave sequences. For example, the spectrum features are analyzed using fast Fourier transform (FFT). The selected model is trained using the labeled aging status data to ensure that the model can accurately identify the relationship between the aging status and the sound wave signal. The aging status of each sound wave sequence is predicted to generate power supply life status prediction data, and the prediction results and corresponding confidence levels are recorded.

[0138] In this embodiment, the specific steps of step S34 are:

[0139] Decompose multiple sound wave sequences segment by segment in the frequency domain to obtain a frequency domain diagram of each sound wave sequence;

[0140] Perform vibration frequency characteristic analysis on the frequency domain graph of each sound wave sequence to extract the vibration frequency characteristics of each sound wave sequence;

[0141] Calculating the time series amplitude change of each sound wave vibration frequency characteristic to generate sound wave time series amplitude change data;

[0142] Extract the vibration frequency peak of each sound wave sequence based on the frequency domain diagram of each sound wave sequence;

[0143] Perform acoustic wave peak gradient fitting on the vibration frequency peak of each acoustic wave sequence to construct an acoustic wave peak gradient curve;

[0144] Calculating the amplitude decay rate of the acoustic wave peak gradient curve based on the acoustic wave time series amplitude change data to generate the acoustic wave amplitude decay rate;

[0145] Conduct in-depth mining of the attenuation change of the acoustic wave amplitude attenuation rate to generate the attenuation change law of the power supply acoustic wave;

[0146] The power supply aging status is predicted based on the power supply acoustic wave attenuation change law to generate power supply life status prediction data.

[0147] In this embodiment, a fast Fourier transform (FFT) is applied to each sound wave sequence to convert the time domain signal into a frequency domain signal, generate a corresponding frequency domain graph, calculate the frequency component and amplitude of each signal, and obtain a frequency domain graph, which is usually displayed in the form of frequency as the X-axis and amplitude as the Y-axis. The frequency domain graph of each sound wave sequence is analyzed to identify the main vibration frequency characteristics. A peak detection algorithm (such as find_peaks of SciPy) can be used to identify significant peaks in the frequency domain graph, extract the vibration frequency characteristics of each sound wave (such as frequency value, amplitude value, etc.), perform time series analysis on all extracted sound wave vibration frequency characteristics, and calculate the vibration at each time point. Amplitude change, which can be obtained by comparing the amplitude of the vibration frequency, generating sound wave time series amplitude change data, recording the amplitude value change at each time point, analyzing the frequency domain diagram of each sound wave sequence, extracting the peak value of the vibration frequency, representing the main vibration frequency, and storing the vibration frequency peak value of each sound wave sequence in a data frame for subsequent analysis, calculating the change rate of the vibration frequency peak value of each sound wave sequence, using linear regression or other fitting methods (such as polynomial fitting) to process the peak data, and generating a sound wave peak gradient curve to reflect the change trend of the vibration frequency peak value over time. According to the sound wave time series amplitude change data, calculate the attenuation rate, attenuation rate = , conduct in-depth analysis on the acoustic wave amplitude attenuation rate to find the attenuation pattern and law. Time series analysis methods (such as ARIMA model) can be used to identify the change trend and generate the power supply acoustic wave attenuation change law. According to the extracted power supply acoustic wave attenuation change law, select a suitable prediction model (such as machine learning model, regression model, etc.) to predict the power supply aging status, use historical data to train the model, ensure that the model can accurately identify the relationship between the acoustic wave characteristics and the power supply aging status, and predict the aging status of the power supply based on the trained model to generate power supply life status prediction data.

[0148] In this embodiment, the specific steps of step S4 are:

[0149] Step S41: performing component topology association analysis on the AC power component structure diagram to generate component topology association data;

[0150] Step S42: Perform three-dimensional point cloud modeling on the component topology association data to construct a three-dimensional power supply topology structure model;

[0151] Step S43: performing life state mapping on the three-dimensional power supply topology structure model based on the power supply life state prediction data to generate a state mapping structure model;

[0152] Step S44: dynamically evolve the temperature distribution of the state mapping structure model based on the three-dimensional temperature distribution field of the power supply, and construct a dynamic twin evolution model.

[0153] In this embodiment, a graph theory method (such as an adjacency matrix or an edge list) is used to perform a topological analysis on the components, the connection relationship and dependency relationship between the components are determined, the connection information of each component is recorded, and the component topology association data is generated, which usually includes the component ID, connection type and connection weight, etc. A three-dimensional point cloud is generated based on the component topology association data, and a three-dimensional modeling software (such as Blender, MeshLab) is used to visualize the position and connection relationship of each component as a point cloud. A point cloud processing algorithm (such as a RANSAC algorithm) is applied to remove noise and optimize the point cloud data to ensure the accuracy of the model. A three-dimensional power supply topology structure model is constructed using the point cloud data to ensure that each component and its connection relationship are reflected in the model, and the power supply life status prediction data is obtained, including the life status indicators of each component (such as the remaining life, aging degree, etc.), and an appropriate A combined mapping method (such as interpolation or heat map generation) is used to map the life status data to the three-dimensional topological structure model, and the life status indicator of each component is mapped to the corresponding three-dimensional model to generate a state mapping structural model. Color coding is usually used to represent different states. The three-dimensional temperature distribution field data of the power supply is collected to ensure that the data can reflect the temperature changes of each component during operation. Based on the state mapping structural model, a dynamic simulation algorithm (such as finite element analysis or fluid dynamics simulation) is applied to dynamically evolve the temperature distribution, dynamically update the temperature distribution in the three-dimensional model, and display the temperature changes of the power supply components under different operating conditions in real time to form a dynamic twin evolution model. The dynamic twin evolution model is demonstrated through visualization tools (such as Unity, Blender) to facilitate users to understand the operating status and temperature changes of the power supply components.

[0154] In this embodiment, the specific steps of step S5 are:

[0155] Step S51: simulating the load demand evolution of the dynamic twin evolution model according to the load demand prediction curve, and extracting power supply simulation response data;

[0156] Step S52: identifying the temperature field change of each component on the power supply simulation response data, and generating simulation data of the temperature field change of each component;

[0157] Step S53: performing component degradation accumulation calculation on the power supply simulation response data to obtain the degradation accumulation value of each component;

[0158] Step S54: calculating the component damage risk of each component based on the degradation accumulation value of each component according to the temperature field change simulation data of each component, thereby obtaining the damage risk data of each component;

[0159] Step S55: performing abnormal failure prediction on the damage risk data of each component to generate abnormal failure prediction data.

[0160] In this embodiment, the load demand prediction curve is input into the dynamic twin evolution model to simulate the response of the power supply under different load conditions, and the simulated response data of the power supply is recorded, including current, voltage, power and their changes over time, to ensure the timing and accuracy of the data. The power supply simulation response data is processed using a data analysis library (such as Pandas or NumPy), and the temperature field changes of each component are identified. Based on the simulated response data, the temperature changes of each component under different load conditions are calculated, and the simulation data of the temperature field changes of each component are generated. A suitable degradation model (such as a linear cumulative model or an exponential decay model) is selected to calculate the degradation cumulative value of each component. According to the temperature field change simulation data and the known degradation rate, the degradation cumulative value of each component is calculated, and the degradation cumulative value = ∑ (temperature change × degradation rate) A suitable risk assessment model (such as a probability model or a threshold-based model) is selected to calculate the damage risk of each component. The temperature field change simulation data and the degradation cumulative value are used to calculate the damage risk of each component. The damage risk = f (degradation cumulative value × temperature change) Wherein, f It is a risk assessment function. According to the damage risk data, a suitable abnormal failure prediction model (such as a classification model, regression model or time series model) is selected, and the selected prediction model is trained using historical failure data to ensure that the model can identify the relationship between damage risk and failure. Based on the damage risk data, abnormal failures that may occur in the future are predicted to generate abnormal failure prediction data.

[0161] In this embodiment, the specific steps of step S6 are:

[0162] Step S61: performing fault location on the dynamic twin evolution model based on the abnormal fault prediction data to obtain the abnormal fault location point;

[0163] Step S62: Generate a remote abnormal warning signal according to the abnormal fault location point;

[0164] Step S63: performing iterative fault diagnosis decisions based on the remote abnormal warning signal to generate multiple fault diagnosis strategies;

[0165] Step S64: performing optimal collaborative analysis based on multiple fault diagnosis strategies to extract the optimal fault diagnosis strategy;

[0166] Step S65: Perform diagnostic visualization on the optimal fault diagnosis strategy and construct a visualization view of the optimal diagnosis strategy.

[0167] In this embodiment, abnormal fault prediction data is collected, including damage risk data and historical fault records of each component. The abnormal fault prediction data is input into the dynamic twin evolution model to analyze the performance of each component in the simulated operation. The data is analyzed using methods such as fault tree analysis (FTA) or Bayesian network to identify possible abnormal fault location points. The abnormal fault location point is recorded in the data frame to ensure that it includes information such as component ID, fault type and confidence. According to the abnormal fault location point, a remote abnormal warning signal is generated. This can be achieved by setting a threshold or using logical rules. If the damage risk > threshold, the generated warning signal will be sent to the monitoring system or relevant personnel to ensure timely response. An iterative fault diagnosis decision framework is established to clarify the steps and criteria for decision-making. According to the current abnormal warning signal analysis, multiple fault diagnosis strategies are generated, including: preliminary inspection strategy, detailed inspection strategy, component replacement strategy, and a suitable multi-objective optimization method (such as Pareto optimization or weighted scoring method) is selected to analyze multiple fault diagnosis strategies. According to indicators such as the effectiveness, cost and implementation time of the strategy, the optimal fault diagnosis strategy is extracted to ensure that it can provide the best results in actual operation. Record the optimal fault diagnosis strategy in a data frame, making sure it includes a description of the strategy and implementation steps. Choose an appropriate visualization tool (such as Tableau, Power BI, or D3.js) to build a visualization. Present the optimal fault diagnosis strategy in a visual way, including a flowchart, decision tree, or dashboard, so that users can understand and operate it. Ensure that the visualization can be updated in real time to reflect the current fault status and diagnosis strategy, and provide intuitive decision support.

[0168] In this embodiment, a system for remotely monitoring the operating status of an AC power supply is provided, which is used to execute the method for remotely monitoring the operating status of an AC power supply as described above, including:

[0169] The load demand prediction module is used to obtain the power supply historical operation status log and the AC power supply component structure diagram; perform time-series load operation analysis on the power supply historical operation status log, and perform load demand prediction to generate a load demand prediction curve;

[0170] The temperature distribution module is used to obtain multi-dimensional operation monitoring parameters, which include operation temperature monitoring parameters and multi-position acoustic wave signals; dynamically render the operation temperature monitoring parameters to construct a three-dimensional temperature distribution field of the power supply;

[0171] The life prediction module is used to perform low-frequency filtering on the acoustic wave signals of multiple parts and predict the power supply aging status to generate power supply life status prediction data;

[0172] Dynamic twin module, which is used to dynamically evolve the temperature distribution of the AC power component structure diagram based on the power life status prediction data and the three-dimensional temperature distribution field of the power supply, and build a dynamic twin evolution model;

[0173] An abnormal fault prediction module is used to simulate the load demand evolution of the dynamic twin evolution model according to the load demand prediction curve, and to predict abnormal faults to generate abnormal fault prediction data;

[0174] The fault diagnosis decision module is used to make iterative fault diagnosis decisions based on abnormal fault prediction data, perform diagnosis visualization, and build a visualization view of the optimal diagnosis strategy.

[0175] By analyzing the historical operation status log of the power supply, the present invention can better understand the operation status of the system, and predict the load demand based on the historical data, which is helpful to optimize resource allocation and plan system capacity. Through multi-dimensional monitoring parameters, the power supply operation status, especially the temperature distribution, can be understood in real time. Dynamic temperature distribution rendering and three-dimensional temperature distribution field construction are helpful to early detection of potential overheating problems so that preventive maintenance measures can be taken. Through acoustic signal processing and life prediction, the aging trend of the power supply can be identified in advance, which is helpful to plan maintenance plans and preventive replacement of key components. Based on the life status prediction data and the temperature distribution field, a dynamic twin evolution model is constructed, which is helpful to monitor the state changes of the power supply components in real time and predict future situations. Through load demand evolution simulation and abnormal fault prediction, potential problems can be discovered in advance and the impact of system faults on operation can be reduced. Iterative fault diagnosis decisions based on abnormal fault prediction data are helpful to improve diagnostic accuracy. Diagnostic visualization and optimal strategy construction provide intuitive information to support decision makers to quickly formulate response plans.

[0176] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0177] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for remotely monitoring the operating status of an AC power supply, characterized in that: The following steps are involved: Step S1: Obtain the power supply historical operation status log and AC power supply component structure diagram; Conduct time-series load operation analysis on the power supply historical operation status log and perform load demand forecasting to generate a load demand forecasting curve; Step S2: acquiring multi-dimensional operation monitoring parameters, wherein the multi-dimensional operation monitoring parameters include operation temperature monitoring parameters and multi-position acoustic wave signals; performing dynamic temperature distribution rendering on the operation temperature monitoring parameters to construct a three-dimensional temperature distribution field of the power supply; Step S3: performing low-frequency filtering on the acoustic wave signals of multiple parts and predicting the power supply aging state to generate power supply life state prediction data; Step S4: dynamically evolve the temperature distribution of the AC power component structure diagram based on the power life state prediction data and the three-dimensional temperature distribution field of the power supply, and construct a dynamic twin evolution model; Step S5: simulating the load demand evolution of the dynamic twin evolution model according to the load demand prediction curve, and performing abnormal fault prediction to generate abnormal fault prediction data; Step S6: Make iterative fault diagnosis decisions based on abnormal fault prediction data, perform diagnosis visualization, and build a visualization view of the optimal diagnosis strategy.

2. The method according to claim 1, characterized in that The specific steps of step S1 are: Step S11: Obtain the power supply historical operation status log and AC power supply component structure diagram; Step S12: Fill missing values ​​in the historical operation status log of the power supply to obtain a standardized operation status log; Step S13: performing sequential load operation analysis on the standardized operation status log to generate a sequential load operation status feature of the power supply; Step S14: sampling the power supply timing load operation state characteristics at multiple time points to obtain the load operation states at multiple time points; Step S15: Calculating the load demand for the load operating status at multiple time points, thereby obtaining multi-stage load demand data of the power supply; Step S16: performing load demand prediction on the multi-stage load demand data of the power supply to generate a load demand prediction curve.

3. The method for remotely monitoring the operation status of an AC power supply according to claim 1, characterized in that: The specific steps of step S2 are: Step S21: performing real-time operation monitoring of the target AC power source based on the sensor array to obtain multi-dimensional operation monitoring parameters, wherein the multi-dimensional operation monitoring parameters include operation temperature monitoring parameters and multi-position sound wave signals; Step S22: performing temperature distribution analysis on the operating temperature monitoring parameters to identify power supply operating temperature distribution data; Step S23: Calculating the temperature fluctuation trend of the operating temperature monitoring parameters to generate a power supply temperature fluctuation trend; Step S24: fitting the temperature difference fluctuation between components to the power supply temperature fluctuation trend based on the AC power supply component structure diagram, and constructing a three-dimensional temperature difference change curve; Step S25: dynamically rendering the three-dimensional temperature difference change curve based on the power supply operation temperature distribution data to construct a three-dimensional temperature distribution field of the power supply.

4. The method for remotely monitoring the operation status of an AC power supply according to claim 3, characterized in that: The specific steps of step S24 are: Perform component visual identification on the AC power component structure diagram and mark multiple power component nodes; Calculate the spatial position of multiple power component nodes and generate the spatial position coordinates of each component node; Based on the spatial position coordinates of each component node, the operating temperature monitoring parameters are spatially correlated and matched to obtain the temperature monitoring value of each component node; Based on the power supply temperature fluctuation trend, the temperature monitoring value of each component node is used to calculate the three-dimensional temperature difference between components, and generate three-dimensional temperature difference data between components; Define the time interval for temperature fluctuations; Perform temperature difference fluctuation analysis on the temperature difference data between three-dimensional components based on the temperature fluctuation time interval, and generate temperature difference fluctuation data of components at multiple intervals; The temperature difference fluctuation data of components at multiple intervals are discretely fitted to construct a three-dimensional temperature difference change curve.

5. The method for remotely monitoring the operation status of an AC power supply according to claim 1, characterized in that: The specific steps of step S3 are: Step S31: performing low-frequency filtering on the sound wave signals of multiple parts to obtain filtered and noise-reduced sound wave signals; Step S32: performing clock-synchronous digital processing on the filtered noise-reduced waveform signal to generate a noise-reduced waveform data sequence; Step S33: performing sliding time segmentation processing on the noise reduction sound wave data sequence to obtain multiple sound wave sequences; Step S34: performing power supply aging status prediction on the multiple sound wave sequences to generate power supply life status prediction data.

6. The method for remotely monitoring the operation status of an AC power supply according to claim 5, characterized in that: The specific steps of step S34 are: Decompose multiple sound wave sequences segment by segment in the frequency domain to obtain a frequency domain diagram of each sound wave sequence; Perform vibration frequency characteristic analysis on the frequency domain graph of each sound wave sequence to extract the vibration frequency characteristics of each sound wave sequence; Calculating the time series amplitude change of each sound wave vibration frequency characteristic to generate sound wave time series amplitude change data; Extract the vibration frequency peak of each sound wave sequence based on the frequency domain diagram of each sound wave sequence; Perform acoustic wave peak gradient fitting on the vibration frequency peak of each acoustic wave sequence to construct an acoustic wave peak gradient curve; Calculating the amplitude decay rate of the acoustic wave peak gradient curve based on the acoustic wave time series amplitude change data to generate the acoustic wave amplitude decay rate; Conduct in-depth mining of the attenuation change of the acoustic wave amplitude attenuation rate to generate the attenuation change law of the power supply acoustic wave; The power supply aging status is predicted based on the power supply acoustic wave attenuation change law to generate power supply life status prediction data.

7. The method for remotely monitoring the operation status of an AC power supply according to claim 1, characterized in that: The specific steps of step S4 are: Step S41: performing component topology association analysis on the AC power component structure diagram to generate component topology association data; Step S42: Perform three-dimensional point cloud modeling on the component topology association data to construct a three-dimensional power supply topology structure model; Step S43: performing life state mapping on the three-dimensional power supply topology structure model based on the power supply life state prediction data to generate a state mapping structure model; Step S44: dynamically evolve the temperature distribution of the state mapping structure model based on the three-dimensional temperature distribution field of the power supply, and construct a dynamic twin evolution model.

8. The method for remotely monitoring the operation status of an AC power supply according to claim 1, characterized in that: The specific steps of step S5 are: Step S51: simulating the load demand evolution of the dynamic twin evolution model according to the load demand prediction curve, and extracting power supply simulation response data; Step S52: identifying the temperature field change of each component on the power supply simulation response data, and generating simulation data of the temperature field change of each component; Step S53: performing component degradation accumulation calculation on the power supply simulation response data to obtain the degradation accumulation value of each component; Step S54: calculating the component damage risk of each component based on the degradation accumulation value of each component according to the temperature field change simulation data of each component, thereby obtaining the damage risk data of each component; Step S55: performing abnormal failure prediction on the damage risk data of each component to generate abnormal failure prediction data.

9. The method for remotely monitoring the operation status of an AC power supply according to claim 1, characterized in that: The specific steps of step S6 are: Step S61: performing fault location on the dynamic twin evolution model based on the abnormal fault prediction data to obtain the abnormal fault location point; Step S62: Generate a remote abnormal warning signal according to the abnormal fault location point; Step S63: performing iterative fault diagnosis decisions based on the remote abnormal warning signal to generate multiple fault diagnosis strategies; Step S64: performing optimal collaborative analysis based on multiple fault diagnosis strategies to extract the optimal fault diagnosis strategy; Step S65: Perform diagnostic visualization on the optimal fault diagnosis strategy and construct a visualization view of the optimal diagnosis strategy.

10. A remote monitoring system for AC power supply operation status, characterized in that: The method for remotely monitoring the operation status of an AC power supply according to claim 1 comprises: The load demand prediction module is used to obtain the power supply historical operation status log and the AC power supply component structure diagram; perform time-series load operation analysis on the power supply historical operation status log, and perform load demand prediction to generate a load demand prediction curve; The temperature distribution module is used to obtain multi-dimensional operation monitoring parameters, which include operation temperature monitoring parameters and multi-position acoustic wave signals; dynamically render the operation temperature monitoring parameters to construct a three-dimensional temperature distribution field of the power supply; The life prediction module is used to perform low-frequency filtering on the acoustic wave signals of multiple parts and predict the power supply aging status to generate power supply life status prediction data; Dynamic twin module, which is used to dynamically evolve the temperature distribution of the AC power component structure diagram based on the power life status prediction data and the three-dimensional temperature distribution field of the power supply, and build a dynamic twin evolution model; An abnormal fault prediction module is used to simulate the load demand evolution of the dynamic twin evolution model according to the load demand prediction curve, and to predict abnormal faults to generate abnormal fault prediction data; The fault diagnosis decision module is used to make iterative fault diagnosis decisions based on abnormal fault prediction data, perform diagnosis visualization, and build a visualization view of the optimal diagnosis strategy.

Images