Intelligent identification system for operation failure of electrical automation control system

Abstract

The application relates to the technical field of fault identification, in particular to an intelligent operation fault identification system of an electrical automation control system, which comprises a current characteristic extraction module, a spectrum self-adaptive adjustment module, a fault signal monitoring module, a filter optimization module and a fault mode identification module. According to the application, the skewness and kurtosis values of the current signal are calculated, the system can not only identify the symmetry and sharpness of the current waveform, but also can capture the slight change of the signal in detail, the identification ability of the nonlinear fault is greatly improved, the self-adaptive adjustment of the spectrum analysis can dynamically optimize the window size and the sliding step according to the characteristics of the real-time waveform, the signal fluctuation in the electrical environment is effectively coped with, the precision of the spectrum analysis is guaranteed, the error caused by the fixed parameter setting is avoided, the mutation and fluctuation intensification of the current waveform are monitored, the system is more sensitive in the capture and processing of the fault signal, and the potential fault can be identified in the early stage.

Description

Technical Field

[0001] This invention relates to the field of fault identification technology, and in particular to an intelligent fault identification system for electrical automation control systems. Background Technology

[0002] Fault identification technology mainly involves the detection, identification, and diagnosis of abnormal states of systems, equipment, or machines during operation. It is widely used in fields such as industrial automation, machine learning, and artificial intelligence. By monitoring the operating status of equipment in real time, it automatically detects potential faults and anomalies, and analyzes the causes of faults through intelligent algorithms or predetermined rules. Its application can help provide timely warnings, reduce equipment downtime, improve equipment reliability and safety, and optimize maintenance strategies.

[0003] Among them, the intelligent fault identification system for electrical automation control systems refers to an intelligent fault identification technology applied to electrical automation control systems. Its purpose is to automatically detect and identify faults that may occur in the electrical system during operation. By monitoring the operating status of the control equipment in real time and analyzing the collected data, the system can promptly detect system anomalies, predict fault occurrences, and help users perform maintenance in advance to avoid production interruptions or equipment damage.

[0004] Existing systems, based on fixed rules or low-order statistical methods, are ill-suited to handle complex nonlinear faults and cannot capture minute changes in current waveforms in a timely manner, leading to the neglect of potential faults. Furthermore, existing systems lack the ability to adapt to changes in the electrical environment, and their spectrum analysis and filtering effects are poor, failing to effectively remove noise interference. This results in low fault diagnosis accuracy in complex electrical environments, affecting maintenance and operation efficiency and increasing the risk of damage to electrical automation control systems. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and propose an intelligent fault identification system for electrical automation control systems.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an intelligent fault identification system for electrical automation control systems, the system comprising:

[0007] The current feature extraction module acquires real-time current waveform data, calculates its skewness and kurtosis based on the instantaneous value of the current signal, analyzes the nonlinear change trend of the current waveform, identifies potential fault signs in the current signal, and obtains the current nonlinear feature identifier.

[0008] The spectrum adaptive adjustment module acquires real-time current waveform data, statistically analyzes the time-domain characteristics of the current waveform, analyzes the size and sliding step of the spectrum window, dynamically adjusts the spectrum analysis window based on the amplitude and stability of the current waveform fluctuations, and then dynamically analyzes the frequency characteristics of the current waveform to obtain the dynamic spectrum characteristic output results.

[0009] Based on the current nonlinearity characteristic identifier, the fault signal monitoring module monitors the changes in the current waveform, analyzes the dynamic characteristics of the current waveform, determines whether the current waveform has undergone abnormal changes, marks abnormal signal events, and generates a fault signal change identifier.

[0010] The filter optimization module obtains the filter parameters of the intelligent fault identification system based on the fault signal change identifier, and adjusts the bandwidth and gain settings of the filter according to the sudden changes and abnormalities in the current waveform to eliminate noise interference in the current signal and obtain the optimized filter configuration.

[0011] The present invention is improved in that the current nonlinearity feature identifier includes waveform symmetry measurement, signal sharpness evaluation information, and waveform abrupt change characteristics; the dynamic spectrum feature output results include spectrum smoothness, spectrum resolution, and frequency change rate; the fault signal change identifier includes abrupt change time, fluctuation amplitude change, and abnormal event identifier; and the filter optimization configuration includes filter bandwidth range, gain adjustment coefficient, and noise filtering efficiency.

[0012] The present invention is improved in that the current feature extraction module includes:

[0013] The current signal acquisition submodule acquires real-time current waveform data, processes the current data to remove noise and frequency interference, verifies the integrity of the current data, and obtains the processed current data.

[0014] The skewness and kurtosis calculation submodule calculates the skewness and kurtosis of the current signal based on the processed current data and the instantaneous value of the current signal, analyzes the symmetry of the current waveform, and measures the sharpness of the signal waveform to obtain the characteristics of the current waveform.

[0015] The current waveform analysis submodule analyzes the nonlinear trend of the current waveform based on the current waveform characteristics, detects whether there are sudden changes or increased fluctuations in the current waveform, and identifies potential fault signs in the current signal using the following formula: ;

[0016] The current nonlinearity characteristic identifier (NLP) is obtained, where SK represents the skewness of the current waveform and KE represents the kurtosis of the current waveform. This represents the instantaneous quantity of current at time i. Represents the average value of the current waveform. This represents the total number of current data points.

[0017] The present invention is improved in that the spectrum adaptive adjustment module includes:

[0018] The current feature extraction submodule acquires real-time current waveform data, extracts the current peak value, cycle mean and waveform deviation in each cycle based on the time series, statistically analyzes the time domain amplitude fluctuation and stability in the current cycle, and generates current waveform amplitude evaluation results.

[0019] The fluctuation decision calculation submodule calculates the amplitude fluctuation rate and stability offset range based on the current waveform amplitude evaluation results, and determines whether the amplitude fluctuation exceeds the standard fluctuation range. If it does, the window length is increased and the sliding step size is decreased; if it does not exceed the standard range, the window length is decreased and the sliding step size is increased, using the following formula: ;

[0020] Obtain the current volatility decision degree The spectrum sliding control parameters are obtained, where, It is the average current amplitude of the current cycle. It is the average current amplitude of the previous cycle. This represents the degree of stability shift in the Z-th period. This represents the degree of current offset in the Zth cycle. Represents the center of the window frequency after the j-th adjustment. The mean of all frequency centers. For the number of cycles, To adjust the number of times;

[0021] The dynamic spectrum generation submodule calls the spectrum sliding control parameters to perform segmented windowing processing on the current current waveform sequence, calculates the dominant frequency and energy concentration of the frequency distribution in each segment, obtains the rate of change and the degree of change of the concentration trend of each segment spectrum, and obtains the dynamic spectrum feature output results.

[0022] The present invention is improved in that the fault signal monitoring module includes:

[0023] The current waveform monitoring submodule monitors the changes in the current waveform based on the current nonlinear characteristic identifier, acquires current time-domain change data, analyzes the nonlinear characteristics of the current waveform, extracts current abrupt change points, fluctuation amplitude and response rate, and constructs a current nonlinear characteristic vector.

[0024] The dynamic change recognition submodule calls the current nonlinear feature vector and, for the response range of the time interval, obtains the current change amplitude, the rate of change of adjacent points, and the peak spacing, using the formula: ;

[0025] Obtaining dynamic nonlinear volatility ,in, This represents the current response value at time i. This represents the current response value at time i+1. Represents the timestamp at time i. Represents the timestamp at time i+1. Represents the total number of time points;

[0026] The anomaly marker generation submodule determines whether the current waveform has undergone abnormal changes based on the dynamic nonlinear fluctuation degree. Combining the sudden changes and increased fluctuations in the current waveform, it marks abnormal signal events and generates fault signal change identifiers.

[0027] The present invention is improved in that the filter optimization module includes:

[0028] The fault signal acquisition submodule detects the trend of fault signal changes based on the fault signal change identifier and the sudden changes and abnormalities in the current waveform, thereby obtaining the change characteristics of the fault signal.

[0029] The filter parameter adjustment submodule adjusts the bandwidth and gain settings of the filter based on the variation characteristics of the fault signal and for waveform abrupt changes and abnormal situations, optimizes the signal filtering process, and obtains the adjusted filter parameters.

[0030] The filter optimization configuration submodule optimizes the filter based on the adjusted filter parameters to remove noise interference from the current signal, using the following formula: ;

[0031] Calculate the optimized adjustment value FV to obtain the optimal filter configuration, where, Represents the amplitude of the original current signal. BV represents the amplitude of the current signal after filter adjustment, GV represents the adjusted filter bandwidth, RV represents the filter gain adjustment parameter, KV represents the noise interference cancellation coefficient, and KV represents the weighting coefficient of the optimized configuration.

[0032] The present invention has an improvement, wherein the system further includes:

[0033] The fault mode identification module judges the fault characteristics in the current waveform based on the optimized configuration of the filter and the output results of the dynamic spectrum characteristics. Then, based on the changes in the current waveform and the known fault modes, it identifies potential faults, determines the fault type of the electrical automation control system, and obtains the fault diagnosis results.

[0034] The fault diagnosis results include fault type identification results, fault location accuracy, and fault impact assessment results.

[0035] The present invention is improved in that the fault mode identification module includes:

[0036] Based on the optimized filter configuration and dynamic spectrum feature output results, the fault identification submodule performs preliminary analysis on the current waveform, extracts the time-frequency features in the current waveform, obtains the amplitude and frequency information of the current waveform, and generates current waveform feature quantities.

[0037] The fault diagnosis submodule compares the known fault modes with the current waveform features based on the current waveform features, determines the matching degree between the known fault modes and the current waveform features, identifies potential faults, determines the fault type, and obtains the fault diagnosis result.

[0038] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0039] In this invention, by calculating the skewness and kurtosis of the current signal, the system can not only identify the symmetry and sharpness of the current waveform, but also capture minute changes in the signal, greatly improving the ability to identify nonlinear faults. The adaptive adjustment of the spectrum analysis can dynamically optimize the window size and sliding step size according to the characteristics of the real-time waveform, effectively dealing with signal fluctuations in the electrical environment, ensuring the accuracy of the spectrum analysis, and avoiding errors caused by fixed parameter settings. Combined with the monitoring of abrupt changes and increased fluctuations in the current waveform, the system is more sensitive in capturing and processing fault signals, and can identify potential faults in the early stages, thereby providing timely warnings. The optimization of the filter further enhances the system's anti-interference ability, ensuring that the accurate identification of fault signals can still be maintained in high-noise environments, thus enhancing the safety and operating efficiency of the electrical automation control system. Attached Figure Description

[0040] Figure 1 This is a system flowchart of the present invention;

[0041] Figure 2 This is a flowchart of the current feature extraction module in this invention;

[0042] Figure 3 This is a flowchart of the spectrum adaptive adjustment module in this invention;

[0043] Figure 4 This is a flowchart of the fault signal monitoring module in this invention;

[0044] Figure 5 This is a flowchart of the filter optimization module in this invention;

[0045] Figure 6 This is a flowchart of the fault mode identification module in this invention. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0047] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Example

[0048] Please see Figure 1 This invention provides a technical solution: an intelligent fault identification system for electrical automation control systems includes:

[0049] The current feature extraction module acquires real-time current waveform data from the electrical automation control system. Based on the instantaneous value of the current signal, it calculates its skewness and kurtosis. The symmetry of the current waveform is judged based on the skewness value, and the sharpness of the current waveform is judged based on the kurtosis value. The nonlinear change trend of the current waveform is analyzed, and it is detected whether the current waveform has abrupt changes or increased fluctuations. Potential fault signs in the current signal are identified, and the nonlinear characteristic identifier of the current is obtained.

[0050] The spectrum adaptive adjustment module acquires real-time current waveform data from the electrical automation control system, statistically analyzes the time-domain characteristics of the current waveform, analyzes the size of the spectrum window and the sliding step size, and dynamically adjusts the spectrum analysis window based on the amplitude and stability of the current waveform fluctuation. If the waveform fluctuation is too large, the window length is increased and the sliding step size is reduced; if the waveform is stable, the window length is reduced and the sliding step size is increased. Then, the frequency characteristics of the current waveform are dynamically analyzed to obtain the dynamic spectrum characteristic output results.

[0051] The fault signal monitoring module monitors changes in the current waveform based on the nonlinear characteristics of the current, analyzes the dynamic characteristics of the current waveform, determines whether the current waveform has undergone abnormal changes, and, in conjunction with the sudden changes and increased fluctuations in the current waveform, marks abnormal signal events and generates fault signal change indicators.

[0052] The filter optimization module obtains the filter parameters of the intelligent fault identification system based on the fault signal change identifier, and adjusts the filter bandwidth and gain settings according to the sudden changes and abnormalities in the current waveform to eliminate noise interference in the current signal and obtain the optimized filter configuration.

[0053] The fault mode identification module judges the fault characteristics in the current waveform based on the filter optimization configuration and dynamic spectrum characteristic output results. Then, based on the changes in the current waveform and the known fault modes, it identifies potential faults, determines the fault type of the electrical automation control system, and obtains the fault diagnosis results.

[0054] The current nonlinearity characteristic identifiers include waveform symmetry measurement, signal sharpness assessment information, and waveform abrupt change characteristics. The dynamic spectrum characteristic output results include spectrum smoothness, spectrum resolution, and frequency change rate. The fault signal variation identifiers include abrupt change time, fluctuation amplitude change, and abnormal event identifiers. The filter optimization configuration includes filter bandwidth range, gain adjustment coefficient, and noise filtering efficiency. The fault diagnosis results include fault type identification results, fault location accuracy, and fault impact assessment results.

[0055] Please see Figure 2 The current feature extraction module includes:

[0056] The current signal acquisition submodule acquires real-time current waveform data, processes the current data to remove noise and frequency interference, verifies the integrity of the current data, and obtains the processed current data.

[0057] Real-time current data is collected by sensors. The data contains detailed information about current fluctuations in the power system. To remove noise, low-pass filters or Kalman filters are typically used to remove high-frequency interference. In addition, frequency analysis tools (such as Fourier transform) are used to identify and remove noise in specific frequency bands. At this point, the current data has been processed and its integrity is verified (such as comparing the data with a preset time window or current waveform shape) to confirm its completeness. In this process, the current signal data is collected and processed, making the current waveform dataset the basis for subsequent analysis. For example, if the instantaneous current value fluctuates greatly through the collected data waveform graph, the fluctuation will be smoothed after filtering, leaving a more realistic current waveform and ensuring that the data integrity is not lost due to external interference. Finally, after this series of processing, the current data becomes a reliable data source for subsequent feature extraction.

[0058] The skewness and kurtosis calculation submodule calculates the skewness and kurtosis of the current signal based on the processed current data and the instantaneous value of the current signal, analyzes the symmetry of the current waveform, and measures the sharpness of the signal waveform to obtain the characteristics of the current waveform.

[0059] Based on the processed current data, its instantaneous values ​​are analyzed, and then the skewness and kurtosis of the current signal are calculated. Skewness reflects the symmetry of the waveform distribution, while kurtosis reflects the sharpness of the waveform peaks. First, by extracting the instantaneous values ​​of the processed current waveform, i.e., the current value at each specific moment, skewness and kurtosis are calculated for the data values. The calculation of skewness requires comparing the deviation between the mean and the instantaneous value of the current signal. By performing mathematical calculations (such as the calculation of the third moment) on all data points, its asymmetry is quantified. For example, if the current at a certain time period... The current waveform exhibits relatively sharp peaks, and the left side of the waveform is steeper than the right side. The skewness value will be positive. Kurtosis is determined by calculating the sharpness of the current waveform, usually using the fourth moment. For example, when analyzing a current signal, assuming its instantaneous value is [2.5, 3.0, 3.5] and the mean is 3.0, a high kurtosis indicates that the peaks of the current waveform are sharp, indicating the instability of the current signal. It can quantitatively analyze the symmetry and sharpness characteristics of the current waveform, providing key features for subsequent waveform analysis.

[0060] The current waveform analysis submodule analyzes the nonlinear trend of the current waveform based on its characteristics, detects whether there are sudden changes or increased fluctuations in the current waveform, and identifies potential fault signs in the current signal using the following formula: ;

[0061] The current nonlinearity characteristic identifier (NLP) is obtained, where SK represents the skewness of the current waveform, which is an index describing the asymmetry of the current waveform distribution and reflects the degree of deviation of the waveform from the left and right sides; KE represents the kurtosis of the current waveform, which is an index measuring the sharpness of the current waveform peaks. This represents the instantaneous current at time i, that is, the current data measured at a certain moment. Represents the average value of the current waveform. Represents the total number of current data points;

[0062] By calculating the skewness and kurtosis of the current waveform, we can analyze whether there are abrupt changes or increased fluctuations in the current waveform, thereby detecting potential fault signs. To calculate the skewness and kurtosis, we first need to collect current signal data at multiple time points and calculate their average values. Assuming the instantaneous values ​​of the current signal over a certain time interval are {2.5, 3.0, 3.5}, then The value is 3.0. Next, we calculate the skewness and kurtosis. Assuming the skewness Sk is 0.5 and the kurtosis KE is 2.0, we can obtain the following results using the formula: For each current value, the deviation is obtained by subtracting the mean, for example: , , ;

[0063] Calculate the difference between each instantaneous value and the mean:

[0064] ;

[0065] ;

[0066] ;

[0067] Then, summing the current value deviations yields the total deviation:

[0068] ;

[0069] therefore, The calculation is as follows: ;

[0070] The results indicate that although there are certain deviations and sharp peaks (characteristics provided by skewness and kurtosis) in this current waveform, its nonlinear characteristics are not significant, and the current waveform does not show any signs of abrupt changes or increased fluctuations.

[0071] Please see Figure 3 The spectrum adaptive adjustment module includes:

[0072] The current feature extraction submodule acquires real-time current waveform data, extracts the current peak value, cycle mean and waveform deviation in each cycle based on the time series, statistically analyzes the time domain amplitude fluctuation and stability in the current cycle, and generates current waveform amplitude evaluation results.

[0073] The current signal for each cycle is analyzed using a time series analysis method to extract the current peak value, cycle mean, and waveform deviation. First, the current signal is sampled, and a data point is recorded at regular time intervals to form a continuous current waveform data sequence. From this data, the system can calculate the current peak value for each cycle. This process involves finding the maximum value of the current waveform from the original data sequence and recording it as the peak value for that cycle. The cycle mean is obtained by calculating the average of all data points within each cycle. The waveform deviation is assessed by calculating the difference between the current value within the cycle and the mean value. Next... The system analyzes the time-domain amplitude fluctuation and stability of the current signal within a given cycle by statistically analyzing the values. Specifically, it calculates the fluctuation range of the current amplitude within each cycle, assesses its variation within that cycle, and evaluates the stability of the waveform. By comparing the mean and the fluctuation amplitude, it obtains the amplitude evaluation result of the current waveform. For example, if the peak current is 5A, the cycle mean is 3A, and the waveform deviation is 1A within a certain cycle, this indicates that the current waveform in that cycle has a certain fluctuation amplitude, and the fluctuation amplitude is relatively large, thus affecting the stability of the system. Therefore, the system outputs the current waveform amplitude evaluation result to indicate the stability level of the current waveform.

[0074] The fluctuation decision calculation submodule calculates the amplitude fluctuation rate of change and the stability offset range based on the current waveform amplitude evaluation results. It determines whether the amplitude fluctuation exceeds the standard fluctuation range. If it does, the window length is increased and the sliding step size is decreased; if it does not exceed the range, the window length is decreased and the sliding step size is increased. The formula used is: ; Obtain the current fluctuation decision degree WD, and obtain the spectrum sliding control parameters, where, It is the average current amplitude of the current cycle. It is the average current amplitude of the previous cycle. This represents the degree of stability shift in the Z-th period. This represents the degree of current offset in the Zth cycle. This represents the frequency center of the window after the j-th adjustment, indicating the frequency center value of the window, used to specify the frequency range of the window during spectrum analysis. The mean of all frequency centers. For the number of cycles, To adjust the number of times;

[0075] Set the number of cycles and number of adjustments The average amplitude of the current in the current cycle The current is 5A, and the average amplitude of the current in the previous cycle is... Given a value of 3A, based on this data, first calculate the rate of change of amplitude fluctuation: The amplitude fluctuation rate is 2A. Let:

[0076] Stability deviation in the first cycle Current offset degree ;

[0077] Stability shift in the second cycle Current offset degree ;

[0078] Stability shift in the third cycle Current offset degree ;

[0079] Calculate based on the value. :

[0080] First cycle calculation: ;

[0081] Second cycle calculation: ;

[0082] Third cycle calculation: ;

[0083] Sum the calculation results for all cycles: ;

[0084] Number of adjustments After each adjustment, the frequency center The changes are as follows:

[0085] The center of the window frequency after the first adjustment is ;

[0086] The center of the window frequency after the second adjustment is ;

[0087] The center of the window frequency after the third adjustment is ;

[0088] mean of frequency center The calculation is as follows:

[0089] ;

[0090] Next, calculate the absolute value of the difference between the frequency center and the mean after each adjustment, and sum them:

[0091] ;

[0092] ;

[0093] ;

[0094] After summing:

[0095] ;

[0096] Substitute all the calculation results into the formula for the volatility decision degree (WD) for calculation:

[0097] ;

[0098] The results indicate that the current fluctuation decision factor WD=1.35, which means that the amplitude fluctuation of the current waveform is relatively large. It is necessary to make corresponding adjustments to the spectrum sliding control parameters, such as increasing the window length and reducing the sliding step size, in order to smooth the fluctuation.

[0099] The dynamic spectrum generation submodule calls the spectrum sliding control parameters to perform segmented windowing processing on the current current waveform sequence, calculates the dominant frequency and energy concentration of the frequency distribution in each segment, obtains the rate of change and the degree of change of the concentration trend of each segment's spectrum, and obtains the dynamic spectrum feature output results.

[0100] Based on the previously calculated spectrum sliding control parameters (such as the adjusted window length and step size), the current waveform is divided into multiple spectrum segments. Then, the dominant frequency and energy concentration within each spectrum segment are calculated. The dominant frequency represents the peak frequency of the current signal in that segment, while the energy concentration is assessed by calculating the energy distribution of frequency components near the peak frequency to determine whether the energy is concentrated near the dominant frequency. Next, the rate of change of each spectrum segment is calculated, i.e., the frequency characteristics and energy concentration of two consecutive spectrum segments are compared to obtain the degree of change. Through the data, the concentration trend of the spectrum is further analyzed to determine whether the signal has changed significantly. Based on the trend of change, the current signal processing strategy is adjusted to obtain the dynamic spectrum characteristic output result. In this process, the purpose of spectrum analysis is to accurately capture the frequency fluctuations and changes in the current signal to provide data support for subsequent optimization decisions. For example, if the dominant frequency of a certain spectrum segment changes significantly, a more refined frequency adjustment strategy will be adopted for that signal segment to ensure the stable operation of the entire system.

[0101] Please see Figure 4 The fault signal monitoring module includes:

[0102] The current waveform monitoring submodule monitors changes in the current waveform based on the current nonlinearity feature identifier, acquires current time-domain change data, analyzes the nonlinear characteristics of the current waveform, extracts current abrupt change points, fluctuation amplitude and response rate, and constructs a current nonlinearity feature vector.

[0103] Monitoring changes in current waveforms typically relies on real-time data streams from sensors, such as current sensors or acquisition devices, to collect changes in the current signal, obtain time-domain data of the current, and perform further analysis. During monitoring, the first step is to record the time-domain changes of the current signal, i.e., to collect the current response values ​​at various time points. The current response value is analyzed based on the waveform characteristics of the current to extract its nonlinear features. Taking a current waveform at a certain instant as an example, if the current signal shows a sudden change point, that is, the current changes significantly in a short period of time, then this point will be identified as a current sudden change point, and the fluctuation amplitude of the current (such as jumping from 10A to 30A) can be calculated. Next, the response rate of the current waveform is calculated, usually based on the rate of change of the current within a time interval. For example, if the current increases from 10A to 20A in one second, the response rate is 10A / s. Through the monitoring process, a current nonlinear feature vector of the current waveform is constructed. This feature vector comprehensively reflects the changing law of the current waveform and can ultimately provide data support for further subsequent analysis.

[0104] The dynamic change recognition submodule calls the current nonlinearity feature vector to obtain the current change amplitude, the rate of change between adjacent points, and the peak spacing for the response range over a time interval, using the following formula: ; Obtain dynamic nonlinear volatility ,in, This represents the current response value at time i, that is, the value of the current at the i-th time point. This represents the current response value at time i+1. This represents the timestamp at time i, that is, the time corresponding to the current value i. Represents the timestamp at time i+1. Represents the total number of time points;

[0105] First, extract the response range of the current waveform in different time intervals and calculate the amplitude of current change. To identify dynamic fluctuation characteristics, a certain time interval needs to be selected (e.g., every 0.1 seconds is one time interval), and the change of current value needs to be monitored within this time interval. For example, if the current at time i is... The current at time i+1 is The amplitude of the current change is Next, the rate of change between adjacent points is calculated, i.e., the rate of change of current is calculated. This often depends on the difference in current over time periods divided by the time difference, expressed as: For example, if the value at time i is 5A and the value at time i+1 is 15A, and the time difference is 0.5 seconds, then the rate of change is... Furthermore, calculating the peak spacing, i.e., the time difference between two adjacent peaks, can help identify periodic changes in the current waveform when it is periodic. For formula calculations, if the total time is 2, then... Then consider the current data at time 1 and time 2: , Seconds and Seconds, substitute into the formula to calculate: ;

[0106] Therefore, the calculated dynamic nonlinear fluctuation degree is This helps to assess the intensity of dynamic changes in the current waveform and thus infer whether a fault signal exists.

[0107] The anomaly marker generation submodule determines whether the current waveform has undergone abnormal changes based on the dynamic nonlinear fluctuation degree. Combining the sudden changes and increased fluctuations in the current waveform, it marks abnormal signal events and generates fault signal change identifiers.

[0108] The system monitors sudden changes and fluctuations in the current waveform to determine if abnormal changes have occurred. For the current waveform, if the rate of change at a certain point in time exceeds a preset threshold for sudden change judgment (e.g., 5A / s), it is marked as a sudden change point. Next, the dynamic nonlinear fluctuation of the current waveform is combined. If the fluctuation exceeds a set threshold (e.g., set to 0.1), it is marked as an abnormal event. This process is judged by comparing the dynamic fluctuation and the rate of change with the preset threshold. Finally, the system judges whether the current waveform has changed abnormally based on the information. For example, if the current sudden change exceeds 5A / s and the fluctuation exceeds 0.1 at a certain moment, the signal at that moment will be marked as an abnormal event, thereby generating a fault signal change indicator and providing corresponding early warning or fault handling for the system.

[0109] Please see Figure 5 The filter optimization module includes:

[0110] The fault signal acquisition submodule detects the trend of fault signal changes based on the fault signal change identifier and the sudden changes and abnormalities in the current waveform, and obtains the change characteristics of the fault signal.

[0111] The system monitors current signals using sensors, collects current data in real time, and preprocesses it. During this process, the sensors record information such as the amplitude, frequency, and phase of the current. If the system detects a sudden change in the current signal, such as a sudden increase in the current amplitude or a waveform interruption, it triggers the generation of a fault signal change indicator. For example, when the system detects a jump in current amplitude from 5A to 10A, it automatically identifies this as a sudden change and marks this location as a fault signal. By further tracking the waveform change trend, the system classifies the waveform change according to a preset threshold to determine whether it belongs to normal fluctuation or abnormal fault. For example, if the change amplitude of the current signal exceeds 5A, this amplitude can be set as the fault signal change indicator. Through this indicator, the system can further track and identify the specific characteristics of the current fault and finally obtain the change trend of the fault signal. This process is based on specific threshold settings to ensure accurate identification of fault signals. For example, suppose in an industrial application, the current fluctuates between 1-2A when working normally, but a sudden change is detected, and the current rises rapidly to 6A. At this time, the system will automatically generate a fault signal indicator.

[0112] The filter parameter adjustment submodule adjusts the bandwidth and gain settings of the filter based on the variation characteristics of the fault signal and for waveform abrupt changes and abnormal situations, optimizes the signal filtering process, and obtains the adjusted filter parameters.

[0113] The system analyzes the current signal based on the change indicators of the fault signal to confirm the specific location and abnormal amplitude of the sudden change. Subsequently, the filtering process is optimized by adjusting the filter bandwidth and gain in real time. For example, if the current signal waveform changes drastically, the filter bandwidth is increased to capture the sudden change signal more broadly; if the abnormal amplitude is small, the bandwidth can be appropriately reduced to avoid noise interference. In addition, the adjustment of the filter gain also plays a crucial role. The system automatically adjusts the gain coefficient according to the changing characteristics of the current signal to amplify the useful signal and suppress noise. Suppose that in a specific application, a current signal exhibits a 20% fluctuation amplitude. The filter will adjust the bandwidth based on this data to effectively filter out high-frequency noise in the signal. At the same time, by adjusting the gain, it ensures that the effective components in the signal are appropriately amplified. Through the above process, the system can adjust the filter parameters to optimize the current signal processing and obtain the adjusted filter parameters. For example, the bandwidth is adjusted from 50Hz to 200Hz, and the gain value is adjusted from 1.0 to 1.5 to ensure more accurate signal transmission.

[0114] The filter optimization configuration submodule optimizes the filter based on the adjusted filter parameters to remove noise interference from the current signal, using the following formula: ;

[0115] Calculate the optimization adjustment value The optimized filter configuration is obtained, where, Represents the amplitude of the original current signal. This represents the amplitude of the current signal after filter adjustment; it refers to the amplitude of the current signal after the filter parameters have been adjusted. This represents the adjusted filter bandwidth, indicating the bandwidth setting selected during the filter optimization process. This represents the filter gain adjustment parameter, indicating the parameter for adjusting the filter gain. The gain determines the degree of signal amplification. Setting the gain too high will amplify noise, while setting it too low will prevent the signal from being processed effectively. Represents the noise interference cancellation coefficient. The weighting coefficients represent the optimized configuration.

[0116] After obtaining the adjusted filter bandwidth and gain values, the system performs comprehensive parameter processing and optimization. For example, if the bandwidth is too wide and noise is not effectively filtered out, the system will reduce the bandwidth to enhance the noise suppression effect of the signal; conversely, if the gain is too high and the effective signal is over-amplified, the system will correspondingly lower the gain value. During this process, the filter optimization settings are calculated using formulas to adjust the settings. To obtain the filter optimization configuration process, if the following data is available:

[0117] Original current signal amplitude (Original values), after normalization (Number after normalization);

[0118] Adjusted current signal amplitude (Original values), after normalization (Number after normalization);

[0119] Adjusted filter bandwidth (Original values), after normalization (Number after normalization);

[0120] Filter gain (Original values), after normalization (Number after normalization);

[0121] Noise interference cancellation coefficient (Original values), after normalization (Number after normalization);

[0122] Weighting coefficients for optimized configuration (Original values), after normalization (Number after normalization);

[0123] Substitute into the formula to calculate: ;

[0124] Therefore, the optimized adjustment value for the filter optimization configuration process is 1.8333 (normalized value), which represents the result after filter optimization configuration and is used to eliminate noise interference in the current signal.

[0125] Please see Figure 6 The fault mode identification module includes:

[0126] The fault identification submodule performs preliminary analysis of the current waveform based on the filter optimization configuration and dynamic spectrum feature output results, extracts the time-frequency features of the current waveform, obtains the amplitude and frequency information of the current waveform, and generates current waveform feature quantities.

[0127] First, the current waveform is analyzed through filter optimization and dynamic spectrum feature analysis. In practical applications, assuming the current waveform of a motor is being monitored, the current waveform data of the motor is first acquired through sensors. Then, based on the filter optimization configuration, noise in the current waveform is filtered out to ensure that the acquired data is clearer and can reflect the true working state of the motor. Next, the time-frequency characteristics of the current waveform are analyzed through dynamic spectrum feature analysis to analyze the amplitude and frequency information of the current and obtain the time-frequency characteristic value of the current. Taking motor current data as an example, assuming the current amplitude is 20A and the frequency is 50Hz during the acquisition process, a specific reference value can be set according to the working characteristics of the motor. If the deviation between the actual amplitude and frequency exceeds a certain threshold, it indicates that there is a problem with the current waveform. At this time, by comparing with the set threshold, the current waveform characteristic quantity is generated, providing data support for subsequent fault diagnosis.

[0128] The fault diagnosis submodule is based on current waveform characteristics. It compares known fault modes with current waveform characteristics, determines the matching degree between known fault modes and current waveform characteristics, identifies potential faults, determines the fault type, and obtains fault diagnosis results.

[0129] Based on the obtained current waveform features, combined with a known fault mode library, further analysis is performed. In actual operation, assuming that the motor exhibits a current waveform different from the standard model during operation, the spectral feature values ​​of the current are first obtained by extracting waveform features. If a certain mode in the known fault mode library has a high degree of matching with the current current waveform features, the system will determine that the motor has a certain type of fault based on this mode. For example, if the matching degree of a certain fault mode in the library is above 80% and has a high similarity to the spectral features of the current waveform, it can be identified as a potential fault. Suppose that by comparison, the matching degree of the current waveform with the "overload fault" mode is 85%, which indicates that the motor has an overload fault. After this series of comparisons and judgments, the fault diagnosis result is finally generated, and the fault type of the motor is determined.

[0130] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. An intelligent fault identification system for electrical automation control systems, characterized in that, The system includes: The current feature extraction module acquires real-time current waveform data, calculates its skewness and kurtosis based on the instantaneous value of the current signal, analyzes the nonlinear change trend of the current waveform, identifies potential fault signs in the current signal, and obtains the current nonlinear feature identifier. The spectrum adaptive adjustment module acquires real-time current waveform data, statistically analyzes the time-domain characteristics of the current waveform, analyzes the size and sliding step of the spectrum window, dynamically adjusts the spectrum analysis window based on the amplitude and stability of the current waveform fluctuations, and then dynamically analyzes the frequency characteristics of the current waveform to obtain the dynamic spectrum characteristic output results. Based on the current nonlinearity characteristic identifier, the fault signal monitoring module monitors the changes in the current waveform, analyzes the dynamic characteristics of the current waveform, determines whether the current waveform has undergone abnormal changes, marks abnormal signal events, and generates a fault signal change identifier. The filter optimization module obtains the filter parameters of the intelligent fault identification system based on the fault signal change identifier, and adjusts the bandwidth and gain settings of the filter according to the sudden changes and abnormalities in the current waveform to eliminate noise interference in the current signal and obtain the optimized filter configuration.

2. The intelligent fault identification system for electrical automation control systems according to claim 1, characterized in that, The current nonlinearity feature identifier includes waveform symmetry measurement, signal sharpness evaluation information, and waveform abrupt change characteristics. The dynamic spectrum feature output results include spectrum smoothness, spectrum resolution, and frequency change rate. The fault signal variation identifier includes abrupt change time, fluctuation amplitude change, and abnormal event identifier. The filter optimization configuration includes filter bandwidth range, gain adjustment coefficient, and noise filtering efficiency.

3. The intelligent fault identification system for electrical automation control systems according to claim 1, characterized in that, The current feature extraction module includes: The current signal acquisition submodule acquires real-time current waveform data, processes the current data to remove noise and frequency interference, verifies the integrity of the current data, and obtains the processed current data. The skewness and kurtosis calculation submodule calculates the skewness and kurtosis of the current signal based on the processed current data and the instantaneous value of the current signal, analyzes the symmetry of the current waveform, and measures the sharpness of the signal waveform to obtain the characteristics of the current waveform. The current waveform analysis submodule analyzes the nonlinear trend of the current waveform based on the current waveform characteristics, detects whether there are sudden changes or increased fluctuations in the current waveform, and identifies potential fault signs in the current signal using the following formula: ; Obtain the nonlinear characteristic identifier of the current. ,in, Represents the skewness of the current waveform. Represents the kurtosis of the current waveform. Representing the The instantaneous quantity of current at a given moment. Represents the average value of the current waveform. This represents the total number of current data points.

4. The intelligent fault identification system for electrical automation control systems according to claim 1, characterized in that, The spectrum adaptive adjustment module includes: The current feature extraction submodule acquires real-time current waveform data, extracts the current peak value, cycle mean and waveform deviation in each cycle based on the time series, statistically analyzes the time domain amplitude fluctuation and stability in the current cycle, and generates current waveform amplitude evaluation results. The fluctuation decision calculation submodule calculates the amplitude fluctuation rate and stability offset range based on the current waveform amplitude evaluation results, and determines whether the amplitude fluctuation exceeds the standard fluctuation range. If it does, the window length is increased and the sliding step size is decreased; if it does not exceed the standard range, the window length is decreased and the sliding step size is increased, using the following formula: ; Obtain the current volatility decision degree The spectrum sliding control parameters are obtained, where, It is the average current amplitude of the current cycle. It is the average current amplitude of the previous cycle. Representing the The degree of stability deviation of the cycle, Representing the The degree of current deviation in the cycle, Representing the The adjusted window frequency center The mean of all frequency centers. For the number of cycles, To adjust the number of times; The dynamic spectrum generation submodule calls the spectrum sliding control parameters to perform segmented windowing processing on the current current waveform sequence, calculates the dominant frequency and energy concentration of the frequency distribution in each segment, obtains the rate of change and the degree of change of the concentration trend of each segment spectrum, and obtains the dynamic spectrum feature output results.

5. The intelligent fault identification system for electrical automation control systems according to claim 1, characterized in that, The fault signal monitoring module includes: The current waveform monitoring submodule monitors the changes in the current waveform based on the current nonlinear characteristic identifier, acquires current time-domain change data, analyzes the nonlinear characteristics of the current waveform, extracts current abrupt change points, fluctuation amplitude and response rate, and constructs a current nonlinear characteristic vector. The dynamic change recognition submodule calls the current nonlinear feature vector and, for the response range of the time interval, obtains the current change amplitude, the rate of change of adjacent points, and the peak spacing, using the formula: ; Obtaining dynamic nonlinear volatility ,in, Representing the Current response value at time, Representing the Current response value at time, Representing the The timestamp of the moment Representing the The timestamp of the moment Represents the total number of time points; The anomaly marker generation submodule determines whether the current waveform has undergone abnormal changes based on the dynamic nonlinear fluctuation degree. Combining the sudden changes and increased fluctuations in the current waveform, it marks abnormal signal events and generates fault signal change identifiers.

6. The intelligent fault identification system for electrical automation control system according to claim 1, characterized in that, The filter optimization module includes: The fault signal acquisition submodule detects the trend of fault signal changes based on the fault signal change identifier and the sudden changes and abnormalities in the current waveform, thereby obtaining the change characteristics of the fault signal. The filter parameter adjustment submodule adjusts the bandwidth and gain settings of the filter based on the variation characteristics of the fault signal and for waveform abrupt changes and abnormal situations, optimizes the signal filtering process, and obtains the adjusted filter parameters. The filter optimization configuration submodule optimizes the filter based on the adjusted filter parameters to remove noise interference from the current signal, using the following formula: ; Calculate the optimization adjustment value The optimized filter configuration is obtained, where, Represents the amplitude of the original current signal. This represents the amplitude of the current signal after it has been filtered and adjusted. This represents the adjusted filter bandwidth. This represents the filter gain adjustment parameter. Represents the noise interference cancellation coefficient. This represents the weighting coefficient for optimized configuration.

7. The intelligent fault identification system for electrical automation control systems according to claim 1, characterized in that, The system also includes: The fault mode identification module judges the fault characteristics in the current waveform based on the optimized configuration of the filter and the output results of the dynamic spectrum characteristics. Then, based on the changes in the current waveform and the known fault modes, it identifies potential faults, determines the fault type of the electrical automation control system, and obtains the fault diagnosis results. The fault diagnosis results include fault type identification results, fault location accuracy, and fault impact assessment results.

8. The intelligent fault identification system for electrical automation control system according to claim 7, characterized in that, The fault mode identification module includes: Based on the optimized filter configuration and dynamic spectrum feature output results, the fault identification submodule performs preliminary analysis on the current waveform, extracts the time-frequency features in the current waveform, obtains the amplitude and frequency information of the current waveform, and generates current waveform feature quantities. The fault diagnosis submodule compares the known fault modes with the current waveform features based on the current waveform features, determines the matching degree between the known fault modes and the current waveform features, identifies potential faults, determines the fault type, and obtains the fault diagnosis result.

Images