Intelligent monitoring and early warning system for safety state of electrical cabinet

Abstract

The invention discloses an intelligent monitoring and early warning system for the safety state of an electrical cabinet, and belongs to the technical field of electrical variable measurement, and the system comprises a data collection module which is used for obtaining multi-source sensing data of the electrical cabinet, and the multi-source sensing data comprises a current waveform parameter, an infrared temperature distribution parameter and a mechanical vibration spectrum parameter; the health feature evaluation module is used for generating a health state feature vector of the electrical cabinet by using the multi-source sensing data; the instruction generation module is used for outputting an early warning decision instruction according to a matching result of the health state feature vector and a preset fault propagation rule base; and the safety protection module is used for executing a safety protection action when the early warning decision instruction meets a preset execution condition. A multi-source sensing data fusion analysis technology is adopted to construct an electrothermal mechanical composite feature vector, a fault propagation rule base is coupled to realize cross-domain conduction path modeling, composite hidden dangers such as contact deterioration and the like can be accurately identified in a fault incubation period, and graded active safety protection measures are triggered.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical variable measurement, and in particular to an intelligent monitoring and early warning system for the safety status of electrical cabinets. Background Art

[0002] The safety monitoring of electrical equipment is a core link in the operation and maintenance of power systems. As a key power distribution node, hidden dangers such as overheating of internal contacts and insulation deterioration in electrical cabinets will trigger chain failures. Existing technologies usually use independent sensors to monitor single physical quantities. For example, current transformers are used to identify overcurrent signals, or infrared spot thermometers are used to detect local heating.

[0003] Current typical solutions mainly include setting a fixed threshold warning mechanism for the current harmonic distortion rate and an overlimit alarm device based on the absolute value of temperature. Some improved solutions introduce vibration sensors to detect mechanical looseness, but the judgment logics of the three types of parameters are independent of each other, and only isolated alarm signals can be triggered. At the level of fault prediction, existing methods rely on a linear regression model of historical temperature rise rate and load factor, and cannot characterize the dynamic coupling effect of electro-thermal-mechanical parameters.

[0004] However, it is difficult to distinguish normal load fluctuations from early anomalies with single-parameter threshold judgment; at the same time, the lack of a cross-domain correlation analysis model for electrical-thermal-mechanical parameters leads to a high misjudgment rate for compound faults such as contact deterioration; after an alarm, it is necessary to manually retrieve drawings to locate the fault point, and the remaining safe time cannot be quantified, resulting in a lag in protective measures. Summary of the Invention

[0005] To solve the above problems, the present invention provides an intelligent monitoring and early warning system for the safety status of electrical cabinets, which uses multi-source sensing data fusion analysis technology to construct an electro-thermal-mechanical composite feature vector, and couples a fault propagation rule library to realize cross-domain conduction path modeling, and can accurately identify compound hidden dangers such as contact deterioration during the fault latency period and trigger hierarchical active safety protection measures.

[0006] The above object can be achieved by the following solutions: An intelligent monitoring and early warning system for the safety status of electrical cabinets, including a data acquisition module for obtaining multi-source sensing data of the electrical cabinet, where the multi-source sensing data includes current waveform parameters, infrared temperature distribution parameters, and mechanical vibration spectrum parameters; a health feature evaluation module for generating a health status feature vector of the electrical cabinet using the multi-source sensing data; a command generation module for outputting a warning decision command according to the matching result of the health status feature vector and a preset fault propagation rule library; and a safety protection module for performing a safety protection action when the warning decision command meets a preset execution condition.

[0007] Optionally, the data acquisition module includes: a current parameter acquisition unit for acquiring current waveform parameters through a current transformer; a temperature parameter acquisition unit for scanning the surface of the cabinet body through an infrared thermal imaging array to locate the spatial coordinates of the temperature anomaly area and generate infrared temperature distribution parameters; and a vibration parameter acquisition unit for detecting vibration signals through a piezoelectric film sensor array, identifying the energy peak in the characteristic frequency band, and converting and generating mechanical vibration spectrum parameters.

[0008] Optionally, the health feature evaluation module includes: an anomaly judgment unit for using a preset baseline data set to perform anomaly judgment on the current waveform parameters, the infrared temperature distribution parameters, and the mechanical vibration spectrum parameters, and generating a health status feature vector when the judgment result is abnormal; a current feature extraction unit for extracting the harmonic distortion rate and transient fluctuation characteristics in the current waveform parameters; a temperature feature extraction unit for extracting the temperature rise rate corresponding to the spatial coordinates in the infrared temperature distribution parameters based on the harmonic distortion rate; an electro-thermal feature extraction unit for calculating the electro-thermal coupling factor according to the harmonic distortion rate and the temperature rise rate; a mechanical-thermal feature extraction unit for generating a mechanical-thermal stress index according to the temporal variation of the energy peak in the characteristic frequency band of the mechanical vibration spectrum parameters and the temperature rise rate; and a status feature extraction unit for performing feature fusion on the electro-thermal coupling factor and the mechanical-thermal stress index to construct a health status feature vector.

[0009] Optionally, the construction of the baseline data set includes: acquiring the current waveform parameters, the infrared temperature distribution parameters, and the mechanical vibration spectrum parameters under normal working conditions within a preset time period, establishing a reference parameter mapping table, where the reference parameter mapping table includes a harmonic distortion rate threshold, a temperature threshold, and a vibration energy threshold; obtaining the real-time load factor, and dynamically correcting the temperature threshold based on the real-time load factor; and combining the corrected temperature threshold, the harmonic distortion rate threshold, and the vibration energy threshold to obtain the baseline data set.

[0010] Optionally, the instruction generation module includes: a component warning unit for generating a contact component deterioration warning instruction when the electro-thermal coupling factor exceeds a preset coupling threshold and the mechanical-thermal stress index shows a monotonically increasing trend; an electrical warning unit for matching the health status feature vector with a preset fault propagation rule library and generating an electrical operation warning instruction according to the matching result; and a remaining safety time generation unit for extracting the change gradient of the health status feature vector to calculate the remaining safety time and attaching it to the warning instruction; where the warning decision instruction includes the contact component deterioration warning instruction and the electrical operation warning instruction.

[0011] Optionally, the construction of the fault propagation rule library includes: obtaining historical partial discharge intensity, historical cabinet door vibration acceleration, and historical current harmonic distortion rate, and establishing a three-dimensional correlation matrix; obtaining the three-phase current unbalance degree and historical temperature rise rate in historical fault cases, analyzing the derivative relationship between the three-phase current unbalance degree and the historical temperature rise rate, and generating a fault path transfer function; constructing a fault propagation chain logic tree based on the three-dimensional correlation matrix and the fault path transfer function to obtain the fault propagation rule library.

[0012] Optionally, the safety protection module includes: a component protection unit, configured to drive a preset sound and light alarm to output positioning information and remaining safety time when the early warning decision instruction is a contact component deterioration early warning instruction; an electrical protection unit, configured to link corresponding electrical equipment to perform protection operations and generate fault location data to be sent to a preset maintenance terminal when the early warning decision instruction is an electrical operation early warning instruction; a work order generation unit, configured to generate a predictive maintenance work order based on the health status feature vector and associate it with a preset spare parts inventory database.

[0013] Optionally, the system further includes: a hidden danger tracing module; wherein, the hidden danger tracing module is configured to: retrieve the electrical connection diagram of the corresponding position according to the spatial coordinates of the abnormal area corresponding to the mechanical thermal stress index; combine the device parameters in the electrical connection diagram and the phase characteristics of the electro-thermal coupling factor to locate the faulty root device and generate a diagnostic report.

[0014] Optionally, generating the diagnostic report includes: obtaining the installation torque record and historical maintenance log of the faulty root device, and calculating the cumulative damage factor; evaluating the remaining life of the device based on the product result of the electro-thermal coupling factor and the cumulative damage factor to obtain a remaining life evaluation value; associating and storing the remaining life evaluation value with the priority level of the maintenance work order to generate a diagnostic report.

[0015] Based on the same inventive concept, the present invention also provides an intelligent monitoring and early warning method for the safety state of an electrical cabinet, the method including: obtaining multi-source sensing data of the electrical cabinet, the multi-source sensing data including current waveform parameters, infrared temperature distribution parameters, and mechanical vibration spectrum parameters; using the multi-source sensing data to generate a health status feature vector of the electrical cabinet; outputting an early warning decision instruction according to the matching result between the health status feature vector and a preset fault propagation rule library; and performing a safety protection action when the early warning decision instruction meets a preset execution condition.

[0016] Compared with the prior art, the present invention has the following advantages: 1. The present invention realizes the multi-dimensional accurate assessment of the safety status of electrical cabinets; by integrating the real-time monitoring data of three key parameters, namely current waveform, infrared temperature, and mechanical vibration, the system constructs composite eigenvectors such as electro-thermal coupling factors and mechanical thermal stress indices, overcomes the one-sidedness of single-parameter monitoring, and significantly improves the accuracy of early fault identification; 2. The present invention establishes a dynamically evolving fault warning mechanism; based on the fault propagation rule library constructed by the three-dimensional correlation matrix and the fault path transfer function, the system can identify the conduction law of faults among multi-physical field parameters, realize the quantitative prediction from local anomalies to global risks, and effectively compress the emergency response time window; 3. Optimize the active closed-loop control of safety protection; automatically trigger audible and visual positioning alarms or device linkage operations in combination with the type of warning instructions, and at the same time generate predictive maintenance work orders including spare part inventory status, forming an automated closed-loop management from anomaly identification to protection execution, and reducing the risk of fault escalation caused by manual response delays; 4. The present invention enhances the cross-domain traceability ability of the fault root cause; uses the spatial coordinates of the abnormal area of mechanical vibration thermal stress to reverse the electrical connection map, locates the physical position of the faulty device in combination with the phase characteristics of the electro-thermal coupling factor, and evaluates the remaining life through the cumulative damage factor, providing multi-dimensional evidence support for maintenance decisions.

[0017] Other features and advantages of the present invention will be described in the following specification, and, in part, will be obvious from the specification, or will be understood by implementing the present invention. The objectives and other advantages of the present invention can be realized and obtained by the structures pointed out in the specification, claims, and drawings. Description of the Drawings

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 It is a framework diagram of an intelligent monitoring and warning system for the safety status of electrical cabinets according to an embodiment of the present invention.

[0020] Figure 2 It is a structural schematic diagram of an intelligent monitoring and warning system for the safety status of electrical cabinets according to an embodiment of the present invention.[[ID=2,3]]

[0021] Figure 3 It is a spatial schematic diagram of the fault propagation path according to an embodiment of the present invention.

[0022] Figure 4It is a schematic flow diagram of an intelligent monitoring and early warning method for the safety status of an electrical cabinet according to an embodiment of the present invention. Detailed implementation manners

[0023] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0024] Referring to Figure 1 , an embodiment of the present invention proposes an intelligent monitoring and early warning system for the safety status of an electrical cabinet, which uses multi-source sensing data fusion analysis technology to construct an electro-thermal-mechanical composite feature vector, and couples a fault propagation rule base to realize cross-domain conduction path modeling, and can accurately identify composite hidden dangers such as contact degradation during the fault latency period and trigger hierarchical active safety protection measures.

[0025] The system of this embodiment specifically includes: A data acquisition module, configured to acquire multi-source sensing data of the electrical cabinet, where the multi-source sensing data includes current waveform parameters, infrared temperature distribution parameters, and mechanical vibration spectrum parameters; A health feature evaluation module, configured to generate a health status feature vector of the electrical cabinet by using the multi-source sensing data; An instruction generation module, configured to output a warning decision instruction according to the matching result between the health status feature vector and a preset fault propagation rule base; A safety protection module, configured to perform a safety protection action when the warning decision instruction meets a preset execution condition.

[0026] Specifically, three types of multi-source sensing data, namely current waveform, temperature distribution, and vibration spectrum, are synchronously acquired through a current transformer, an infrared thermal imaging array, and a piezoelectric film sensor; then, after abnormal judgment using a preset baseline data set, characteristic parameters such as current harmonic distortion rate and temperature rise rate are extracted, an electro-thermal coupling factor is constructed based on the electro-thermal conduction effect, and a mechanical thermal stress index is generated by combining the energy peak value of the vibration characteristic frequency band and the time sequence evolution of the temperature rise; finally, two types of indexes are fused to form a health status feature vector, which drives subsequent fault matching and protection actions. For the first time, three-dimensional collaborative monitoring of the electrical characteristics, thermodynamic state, and mechanical structure of the electrical cabinet is realized, and a composite health index is constructed through cross-domain feature fusion of electro-thermal-mechanical parameters, solving the false alarm blind area problem of single-sensor monitoring; the construction of the feature vector takes into account both transient fluctuations and long-term degradation trends, improving the recognition sensitivity of composite faults such as contact aging, providing a standardized input interface for fault propagation early warning, and laying a decision-making foundation for active protection.

[0027] Optionally, as Figure 2 shown, the data acquisition module includes: a current parameter acquisition unit for acquiring current waveform parameters through a current transformer; Specifically, the current parameter acquisition unit is connected to a specified wire or a key node in the electrical cabinet circuit through a current transformer, measures the real-time alternating current signal and converts it into digital waveform data; extracts waveform characteristics such as amplitude, frequency and harmonic components through signal processing, and outputs them as current waveform parameters.

[0028] a temperature parameter acquisition unit for scanning the surface of the cabinet body through an infrared thermal imaging array, positioning the spatial coordinates of the temperature abnormal area, and generating infrared temperature distribution parameters; Specifically, the temperature parameter acquisition unit performs a grid scan on the entire surface of the cabinet body through an infrared thermal imaging array, the array is arranged at multiple fixed positions around the cabinet body, acquires infrared radiation data and converts it into a temperature distribution image; then uses the built-in coordinate mapping algorithm to process the image, identifies the area exceeding the preset reference temperature as the temperature abnormal area; calculates and outputs its exact position coordinates in the cabinet coordinate system through spatial transformation, and generates infrared temperature distribution parameters.

[0029] a vibration parameter acquisition unit for detecting vibration signals through a piezoelectric film sensor array, identifying the energy peak value in the characteristic frequency band, and converting and generating mechanical vibration spectrum parameters.

[0030] Specifically, the vibration parameter acquisition unit is attached to the surface of a specified mechanical component of the cabinet body through a piezoelectric film sensor array, and the array detects the mechanical vibration fluctuation and converts it into an analog electrical signal for sampling; after being digitized through analog-to-digital conversion, the fast Fourier transform is applied to process and calculate the spectral density. For the vibration energy density at the discrete frequency , there is: , where is the complex amplitude value at the kth frequency point after the fast Fourier transform, corresponding to the vibration acceleration signal output by the piezoelectric sensor; k is the frequency point index, ranging from 0 to K - 1; K is the total number of frequency points equal to the number of sampling points; is the frequency domain resolution, which is calculated through the sampling interval and the sampling window size ; then, identify the maximum value within a specified frequency range such as the high-frequency resonance band as the energy peak value of the characteristic frequency band, and output it as mechanical vibration spectrum parameters.

[0031] Specifically, by integrating electrical, thermodynamic, and mechanical sensing technologies, multi-modal data synchronization acquisition and feature extraction are achieved, improving the comprehensiveness and accuracy of electrical cabinet monitoring. By early identifying abnormal points, equipment failures can be effectively prevented, the system's timely warning response is supported, and maintenance costs and downtime risks are reduced.

[0032] Optionally, the health feature evaluation module includes: An abnormality judgment unit, which is used to perform abnormality judgment on the current waveform parameters, the infrared temperature distribution parameters, and the mechanical vibration spectrum parameters by using a preset baseline data set. When the judgment result is abnormal, a health status feature vector is generated. Specifically, during operation, real-time multi-source sensing data is received, the current waveform parameters are parsed to calculate the real-time harmonic distortion rate value, the infrared temperature distribution parameters are parsed to obtain the maximum temperature value on the cabinet surface, and the mechanical vibration spectrum parameters are parsed to obtain the energy peak value of the characteristic frequency band. Then, the real-time harmonic distortion rate is compared with the harmonic distortion rate threshold of the baseline data set, the real-time maximum temperature value is compared with the corrected temperature threshold, and the real-time vibration energy peak value is compared with the vibration energy threshold. If any parameter exceeds the corresponding baseline data threshold, the unit determines an abnormal trigger signal. When an abnormality is determined, an initial signal of the health status feature vector is generated, and the subsequent units of the health feature evaluation module are called for feature extraction and fusion processing.

[0033] A current feature extraction unit, which is used to extract the harmonic distortion rate and transient fluctuation characteristics in the current waveform parameters. Specifically, the harmonic distortion rate is extracted from the received current waveform parameters. The harmonic distortion rate is defined as the percentage of the effective value of the total harmonic current to the effective value of the fundamental current, and is calculated by performing spectral analysis on the current data through fast Fourier transform. At the same time, the transient fluctuation characteristics are extracted. The transient fluctuation characteristics use a time series sliding window to calculate the standard deviation of the current waveform as the fluctuation amplitude index.

[0034] A temperature feature extraction unit, which is used to extract the temperature rise rate corresponding to the spatial coordinates in the infrared temperature distribution parameters based on the harmonic distortion rate. Specifically, the operation of the temperature feature extraction unit depends on the above harmonic distortion rate value. When the harmonic distortion rate exceeds the preset threshold, a specific spatial analysis of the infrared temperature distribution parameters is triggered. The spatial coordinates of the abnormal area are obtained from the infrared temperature distribution parameters, and this coordinate is obtained through the temperature distribution image recognition process. Based on the spatial coordinates, a temperature data sequence corresponding to the point is extracted, and the temperature rise rate variable is calculated as the temperature difference between adjacent time intervals t and t + 1 divided by its time step dt, and the absolute value is taken to obtain a positive temperature rise change without negative values.

[0035] An electro-thermal feature extraction unit, which is used to calculate the electro-thermal coupling factor according to the harmonic distortion rate and the temperature rise rate. Specifically, the operation of the electro-thermal feature extraction unit inputs the harmonic distortion rate and the temperature rise rate. The harmonic distortion rate is a percentage variable, and the electro-thermal coupling factor is calculated. : , where is the normalization factor of the temperature rise rate, is the input variable of the harmonic distortion rate, is the adjustment coefficient, which is set by regression analysis of the historical data of the load condition, such as 1.2. is the input variable of the temperature rise rate, is the constant of the reference temperature change rate, which is obtained from the baseline data set.

[0036] The mechanical-thermal feature extraction unit is used to generate a mechanical-thermal stress index according to the characteristic frequency band energy peak of the mechanical vibration spectrum parameter and the time series change of the temperature rise rate. Specifically, the mechanical-thermal feature extraction unit receives the characteristic frequency band energy peak of the mechanical vibration spectrum parameter and the time series change of the temperature rise rate. The unit of the characteristic frequency band energy peak is millivolt and is directly used. The time series change of the temperature rise rate takes the derivative form of the average value of its past 10 time points; the mechanical-thermal stress index is calculated. : , where is the output variable of the mechanical-thermal stress index, dimensionless, is the input variable of the characteristic frequency band energy peak, is the time series change amount of the temperature rise rate, which is obtained by calculating the average change rate within a preset time window through difference and normalizing it after comparing with the reference value.

[0037] The state feature extraction unit is used to perform feature fusion on the electro-thermal coupling factor and the mechanical-thermal stress index to construct a health state feature vector.

[0038] Specifically, the operation of the state feature extraction unit performs feature fusion on the above electro-thermal coupling factor and mechanical-thermal stress index. The fusion method adopts standardization processing. For example, the electro-thermal coupling factor and the mechanical-thermal stress index are normalized to the 0-1 interval and then concatenated in series to form a two-dimensional vector, which is the output of the health state feature vector. This operation is run by a data processing chip in an embedded system to ensure real-time performance.

[0039] Exemplarily, the system implements and reproduces the following operating steps in the industrial electrical cabinet monitoring scenario: First, the current feature extraction unit calculates the harmonic distortion rate to be 8.5 for the real-time current waveform parameter data and extracts the standard deviation of the transient fluctuation feature to be 1.2; then the temperature feature extraction unit triggers an analysis action based on the harmonic distortion rate, locks the temperature data sequence [25°C, 27°C, 30°C] at the spatial coordinates (1, 2) in the upper right corner area of the cabinet, and calculates the temperature rise rate to be 0.083°C / s at a 1-minute interval dt; then the electro-thermal feature extraction unit uses the input harmonic distortion rate of 8.5 and the temperature rise rate of 0.083 to substitute into the formula to calculate the electro-thermal coupling factor, sets 0.01, to be 1.2, to be 5°C / s, and obtains the electro-thermal coupling factor to be 0.00017; subsequently, the mechanical-thermal feature extraction unit obtains the energy peak value of the feature frequency band to be 50 mV and the time-series change of the temperature rise rate at the same position calculates to be 0.01, and calculates the mechanical-thermal stress index to be 0.5; finally, after the state feature extraction unit normalizes the electro-thermal coupling factor and the mechanical-thermal stress index to 0.17 and 0.55, it constructs a health state feature vector [0.17, 0.55], and this vector generates a warning instruction after matching the fault rule base. Through the seamless fusion of multi-source features, accurate fault identification is achieved, and the system's early warning avoids potential electro-thermal-mechanical chain failures, significantly improves the operating reliability of the electrical cabinet, optimizes the configuration efficiency of maintenance resources, and reduces the equipment downtime risk.

[0040] The principle of the whole method lies in using the time-series correlation of electrical, thermal, and mechanical features to construct health indicators for multi-dimensional state assessment to ensure early fault warning. The technical effect of this method is reflected in improving the comprehensiveness and response timeliness of electrical cabinet safety monitoring, reducing misjudgment and extending the equipment service life through efficient feature fusion, while enhancing the initiative of maintenance and reducing the comprehensive operation and maintenance cost; in the verification example, the system successfully prevented the evolution process of equipment overheating accidents based on the combined warning mechanism of temperature rise value and vibration features in the distribution center application.

[0041] Optionally, the construction of the baseline dataset includes: Collect the current waveform parameters, infrared temperature distribution parameters, and mechanical vibration spectrum parameters under normal working conditions within a preset time period, and establish a reference parameter mapping table, where the reference parameter mapping table includes a harmonic distortion rate threshold, a temperature threshold, and a vibration energy threshold; Obtain the real-time load factor, and dynamically correct the temperature threshold based on the real-time load factor; Combine the corrected temperature threshold, the harmonic distortion rate threshold, and the vibration energy threshold to obtain the baseline dataset.

[0042] Specifically, the embedded controller is operated to preset a normal condition data acquisition window for a period of time, such as 7 days, to collect multi-source sensing data of the electrical cabinet, including current waveform parameters, infrared temperature distribution parameters, and mechanical vibration spectrum parameters; after using the data storage module to record these parameters, the reference indicators are calculated. Among them, the harmonic distortion rate threshold is determined by the upper limit of the normal value of the harmonic distortion rate within the statistical period, which is the 95th percentile. The harmonic distortion rate is defined as the ratio of the total harmonic distortion to the fundamental wave amplitude to evaluate the current quality; the temperature threshold is calculated based on the statistical analysis of all the infrared temperature point data on the cabinet surface, and the temperature reference value under normal conditions is taken as the average value plus twice the standard deviation as the initial reference temperature threshold; the vibration energy threshold selects the maximum value of the energy peak in the mechanical vibration spectrum characteristic frequency band as the reference setting value; the real-time load factor is obtained by extracting the maximum current amplitude from the current waveform parameters and dividing it by the rated current value to get the load factor; based on the real-time load factor and the initial reference temperature threshold, the temperature threshold is corrected, and for the corrected temperature threshold , there is: , where is the initial reference temperature threshold obtained from the baseline dataset; is the temperature rise compensation coefficient set as an empirical value through the regression analysis of historical load-temperature data; is the real-time load factor calculated through the current waveform amplitude; is the preset standard load factor set to 1.0; the corrected temperature threshold, the harmonic distortion rate threshold, and the vibration energy threshold are combined and stored in the embedded database to form the baseline dataset.

[0043] Optionally, the instruction generation module includes: The component warning unit is used to generate a contact component deterioration warning instruction when the electro-thermal coupling factor exceeds the preset coupling threshold and the mechanical thermal stress index shows a monotonically increasing trend; Specifically, the component warning unit receives the electro-thermal coupling factor and the mechanical thermal stress index. The electro-thermal coupling factor is output by the health feature evaluation module, and the mechanical thermal stress index is generated by the state feature extraction unit. When the electro-thermal coupling factor exceeds the preset coupling threshold (for example, the typical value is 0.002) and the mechanical thermal stress index shows an upward trend in 3 consecutive sampling periods, a contact component deterioration warning instruction is generated.

[0044] The electrical warning unit is used to match the health state feature vector with the preset fault propagation rule library and generate an electrical operation warning instruction according to the matching result; Specifically, the electrical warning unit calls the fault propagation rule library, which is a pre-built logical tree database, and the health state feature vector (including the normalized electro-thermal coupling factor and the normalized mechanical thermal stress index ), perform matching calculations with the coordinates of the rule base nodes to obtain the Euclidean distance of the vector : , where 、 are the coordinate values of the rule base nodes; when is less than the matching radius (preset 0.1), it is determined that the matching is successful, and an electrical operation warning instruction corresponding to the fault type is generated.

[0045] The safety time generation unit is used to extract the change gradient of the health state feature vector to calculate the remaining safety time and attach it to the warning instruction; wherein, the warning decision instruction includes the contact component deterioration warning instruction and the electrical operation warning instruction.

[0046] Specifically, the safety time generation unit extracts the time series data of the health state feature vector in the recent 5 minutes, and calculates the change rate of the feature vector norm , that is, the change amount of per unit time. The formula for the remaining safety time is where is the time coefficient (obtained through regression analysis of historical fault cases), and the calculation result is attached to the warning instruction.

[0047] Exemplarily, the reproduction instruction generation process of a power distribution cabinet monitoring system: the component warning unit detects that the electrothermal coupling factor is 0.0025, exceeding the coupling threshold of 0.002, and the time series values of the mechanical thermal stress index are [0.51, 0.53, 0.55], showing a monotonic increase, and immediately generates a contact component deterioration warning instruction. At the same time, the electrical warning unit obtains the current health state feature vector [0.25, 0.60], and matches it with the coordinate cluster 2 (105 pC, 0.28 m / s) of the fault propagation rule base 2It is converted into a normalized vector [0.43, 0.57]. The Euclidean distance is calculated to be 0.08, which is less than the matching radius of 0.1. The matching is successful, and an electrical operation warning instruction for overheating of the busbar connection is triggered. The safety time generation unit calculates that the modulus change rate of the feature vector in the last 5 minutes is 0.021 / s. Taking k = 180 (calibrated based on the evolution data of similar faults), the calculated remaining safety time is 14 minutes. It realizes the quantification of fault classification warning and emergency disposal window, accurately identifies the early signs of contact degradation through the electro-thermal and mechanical dual-condition trigger mechanism, combines with the rule base matching to clarify that the fault type is abnormal busbar connection, and at the same time the remaining time calculation provides a clear disposal time limit for on-site personnel; when the system outputs a warning instruction, the maintenance personnel immediately tighten the bolts of the C-phase busbar, eliminate the risk of out-of-control temperature rise caused by the increase of contact resistance within the remaining safety time, avoid causing an inter-phase short circuit accident, effectively improve the fault disposal efficiency and ensure the power supply continuity to the greatest extent.

[0048] Optionally, the construction of the fault propagation rule base includes: Obtain the historical partial discharge intensity, historical cabinet door vibration acceleration, and historical current harmonic distortion rate, and establish a three-dimensional correlation matrix; Obtain the three-phase current unbalance degree and historical temperature rise rate in historical fault cases, analyze the derivative relationship between the three-phase current unbalance degree and the historical temperature rise rate, and generate a fault path transfer function; Based on the three-dimensional correlation matrix and the fault path transfer function, construct a fault propagation chain logic tree to obtain the fault propagation rule base.

[0049] Specifically, the fault propagation path is as Figure 3As shown, the point clouds from left to right represent normal operating conditions, early deterioration, and high-risk faults respectively. Three types of time series data, namely historical partial discharge intensity, historical cabinet door vibration acceleration, and historical current harmonic distortion rate, are obtained within a preset historical period. The historical partial discharge intensity is obtained by measuring the peak current at the key cable connection point using a high-frequency current transformer and converting it into picocoulomb value. The historical cabinet door vibration acceleration is obtained by sampling the root mean square value at the cabinet door hinge using a three-axis piezoelectric accelerometer. The historical current harmonic distortion rate is obtained by collecting the total harmonic distortion percentage of the current using a power quality analyzer. When establishing the three-dimensional correlation matrix, the three parameters are aligned in time series and the data segments are divided with a preset time window (for example, every 10 minutes as a window). For each time window, characteristic quantities are calculated: the maximum value within the partial discharge intensity window, the average value within the cabinet door vibration acceleration window, and the standard deviation within the current harmonic distortion rate window. These three characteristic quantities are used as the X-axis, Y-axis, and Z-axis coordinate values of the matrix respectively to form the three-dimensional space point cloud of the three-dimensional correlation matrix, and each point represents a combination of key state characteristics of a time window. Secondly, the fault records in the historical fault case library are analyzed, and the time series data of the three-phase current unbalance degree within 1 hour before the fault and the historical temperature rise rate data of the abnormal area in the same period are extracted. For each group of fault samples, the first derivative of the three-phase current unbalance degree with respect to time is calculated, and at the same time, the numerical value of the historical temperature rise rate at the corresponding moment is calculated to construct a derivative-rate paired data set. The coefficients a and b of the fault path transfer function are determined by fitting the linear regression model between the two parameters by the least squares method , where is the derivative value of the three-phase current unbalance degree, is the measured value of the historical temperature rise rate, and this function characterizes the physical correlation relationship between the change of current unbalance and the aggravation of temperature rise. Finally, based on the point cloud distribution of the three-dimensional correlation matrix, a spatial distribution foundation is constructed, and the high-density clustering area is identified as the initial fault node. The output of the fault path transfer function is used as the state transfer condition to establish a directed connection edge between adjacent nodes in the three-dimensional space. When exceeds the preset critical value (for example, 0.5 °C / s), the node jump is activated to form a fault propagation chain logic tree with multiple conduction paths. The logic tree is stored as a queryable database structure to form the fault propagation rule library

[0050] Exemplarily, the fault propagation rule library of an electrical cabinet in a substation collects the daily monitoring data within three months to obtain the historical partial discharge intensity (measured by the sensor model HFCT001), the historical cabinet door vibration acceleration (the root mean square value measured by the sensor model PZT-A3, with an average of 0.15 m / s 2), historical current harmonic distortion rate (recorded average 3.2%); construct a point cloud containing 4320 groups of data by statistically analyzing characteristic quantities within a 10-minute time window. Analyze 6 cases of contact welding failures in the historical archives, extract data for 1 hour before the failure, and fit to obtain the failure path transfer function as (The goodness of fit is 0.89). Identify three high-density clusters in the three-dimensional correlation matrix. The coordinates of cluster 1 are (32 pC, 0.12 m / s 2 , 3.8%), cluster 2 (105 pC, 0.28 m / s 2 , 7.1%), and cluster 3 (240 pC, 0.41 m / s 2 , 1.9%). When the failure path transfer function outputs , establish a directed link from cluster 1 to cluster 2 (corresponding to the event of a temperature rise rate of 0.53 °C / s); when the state of cluster 2 persists for 15 minutes and , activate the transfer to cluster 3, and finally generate a fault propagation chain logic tree containing two-level conduction paths. Break through the limitations of traditional single-parameter threshold alarms, accurately capture the cross-domain coupling effect of current-temperature rise-vibration parameters through a physical association model, and identify the conduction path pattern in the early stage of equipment performance degradation; when the actual monitoring data matches the cluster 2 node, the system can early warn of the risk of overheating of the contact and locate it to the C-phase busbar joint, enabling maintenance personnel to complete intervention before insulation degradation and avoid triggering discharge faults; this method significantly improves the fault tracing efficiency and compresses the emergency response time window, while reducing the false alarm rate through path prediction.

[0051] Optionally, the safety protection module includes: A component protection unit, which is used to drive a preset sound and light alarm to output positioning information and remaining safety time when the early warning decision instruction is a contact component degradation early warning instruction; An electrical protection unit, which is used to link corresponding electrical equipment to perform protection operations and generate fault location data to be sent to a preset maintenance terminal when the early warning decision instruction is an electrical operation early warning instruction; A work order generation unit, which is used to generate a predictive maintenance work order based on the health status feature vector and associate it with a preset spare parts inventory database.

[0052] Specifically, when the component protection unit receives the deterioration warning instruction of the contact component, it parses the abnormal area spatial coordinates and the remaining safety time data items contained in the instruction; it sends a pulse signal sequence to the preset audible and visual alarm array through the hardware drive circuit. The first section of the pulse coding format identifies the spatial coordinates (such as the cabinet position coding XYZ), and the second section identifies the remaining safety time (in minutes). It triggers the alarm at the specified position to output a visual alarm at a specific flashing frequency (for example, 5Hz for the coordinate alarm section and 2Hz for the time alarm section) and color (for example, red represents a contact fault), and at the same time broadcasts a voice prompt of the coordinate position number and the time value. When the electrical protection unit receives the electrical operation warning instruction, it identifies the fault type coding (such as the overcurrent code OC) by instruction parsing, queries the preset operation mapping table to obtain the corresponding protection operation items (such as cutting off the corresponding feeder circuit breaker and starting the standby power supply). The operation mapping table is stored in the embedded memory and burned during the assembly stage; after the operation is executed, a fault location data packet is generated, which includes the health status feature vector, the matching rule library node number, and the affected circuit topology identifier, and is transmitted to the maintenance terminal IP address through the RS485 communication interface according to the Modbus protocol. The work order generation unit calls the health status feature vector, calculates the Euclidean distance between it and the case feature vector stored in the historical fault case database, and uses the fault repair plan of the case with the minimum distance as the predictive maintenance work order template : , where is the case index set, represents the th case and the th dimensional eigenvalue; is the data in the health status feature vector , taking 1 or 2; after the work order is generated, it accesses the SQL interface of the spare parts inventory database in real time, and associates the replacement component model with the inventory quantity field to generate a complete work order with a spare parts status identifier.

[0053] Exemplarily, a practical reproduction of a data center power distribution system: The component protection unit issues a warning instruction for component degradation of the contact component (including the coordinate area B23 and the remaining time of 15 minutes), driving the audible and visual alarm installed on the top of the cabinet No. 23 in Column B. The red light flashes at 5 Hz and synchronously broadcasts "The bolt in B23 is loose, and the disposal time limit is 14 minutes and 30 seconds". At the same time, the electrical protection unit receives an electrical operation warning instruction for bus overheating (type code BH01), looks up the table and executes to cut off the bus tie switch in Column B and switch to the standby bus, generates a fault location data packet and sends it to the central monitoring room terminal (including the eigenvector [0.33, 0.68] and the fault point topology number CT-1021). The work order generation unit calculates the minimum distance of the current eigenvector from the case library as 0.11 (matching case ID 203 for replacing the busbar connector), calls the inventory system to confirm that the inventory balance of the spare part model MLA-300 is 3 pieces, and generates a predictive maintenance work order marked with "Replace the busbar joint in Column B, 1 piece of spare part MLA-300 is in sufficient stock". The multi-level protection measures are accurately and collaboratively executed. The audible and visual alarm realizes rapid and accurate fault point location to guide on-site disposal by personnel. The electrical operation automatically isolates the fault to prevent the spread of the accident and ensure the safety of upstream equipment. The intelligent work order system is seamlessly connected to the material management to avoid maintenance delays, thereby forming a complete closed-loop from emergency response to planned maintenance and greatly improving the system reliability.

[0054] Optionally, the system further includes: a hidden danger tracing module; wherein, the hidden danger tracing module is used for: According to the spatial coordinates of the abnormal area corresponding to the mechanical thermal stress index, retrieve the electrical connection diagram of the corresponding position; Combined with the device parameters in the electrical connection diagram and the phase characteristics of the electro-thermal coupling factor, locate the root cause device of the fault and generate a diagnostic report.

[0055] Optionally, the generation of the diagnostic report includes: Obtain the installation torque record and historical maintenance log of the root cause device of the fault, and calculate the cumulative damage factor; Based on the product result of the electro-thermal coupling factor and the cumulative damage factor, evaluate the remaining life of the device to obtain a remaining life evaluation value; Associate and store the remaining life evaluation value with the priority of the maintenance work order to generate a diagnostic report.

[0056] Specifically, query the preset digital twin model database of the electrical cabinet according to the spatial coordinates associated with the generation of the mechanical thermal stress index (such as the cabinet grid coordinates [3,5]), and retrieve the connection diagram data of all electrical devices in this coordinate area, including device identifiers, connection relationship topologies, and rated current parameters. Then extract the phase characteristics of the electrothermal coupling factor, calculate the phase offset angle through the time delay of synchronously collecting the current phase and temperature change, and its physical meaning is the electrothermal response hysteresis. Combine the ratio of the operating current to the rated current of the device in the electrical connection diagram for fault location: calculate the fault probability weight of the kth device : , where is the set of devices in the area, which increases the weight of devices with high load and phase matching, is the ratio of the operating current to the rated current of the kth device in the electrical connection diagram, is the maximum value of the ratio of the real-time operating current to the rated current of the device in the electrical connection diagram, is the phase offset angle calculated from the time delay of the current phase and temperature change. Select the device corresponding to the maximum fault probability weight as the fault root device. When generating a diagnostic report, obtain the installation torque record of this device and the cumulative on-off times in the historical maintenance log from the equipment management system, and calculate the cumulative damage factor : , In the formula is the installation torque record, is the standard installation torque, such as 30 Nm, is the cumulative on-off times in the historical maintenance log, , The material coefficients are obtained by looking up the table according to the device model. Evaluate the remaining life based on the product of the electrothermal coupling factor and the cumulative damage factor. For the remaining life evaluation value , there is: , In the formula, is the acceleration aging coefficient determined through accelerated life tests. Take the numerical value of the remaining life evaluation value to obtain the final remaining life. Finally, compare the remaining life evaluation value with the preset life threshold to divide the priority levels. For example is urgent, is important, and others are routine, and associate and store to generate a diagnostic report

[0057] Exemplarily, when dealing with an abnormality of a certain switch cabinet, the hidden danger tracing module retrieves the electrical connection diagram according to the alarm coordinates [4,2], and the area includes the circuit breaker QF201 ( ), the contactor KM17 ( ). Phase shift of electrothermal coupling factor The failure probability weight calculation of the circuit breaker is 0.85, which is higher than that of the contactor. Query the installation torque record of QF201, which is 45 Nm, and the historical maintenance log (cumulative operation: 2,000 times), and take , , and calculate the cumulative damage factor to be 0.67. The current electrothermal coupling factor is 0.0018, and take , and calculate the remaining life , and generate a diagnostic report marked as "The remaining life of QF201 is 689 hours, and the priority is important". By fusing multiple evidences of electrical topology and thermal response, accurately locate the root cause of the aging of the circuit breaker contacts, and avoid misjudging the heating as a contactor failure; by quantitatively evaluating the remaining life and linking the maintenance priority with the physical entity of the material, enable the maintenance team to accurately formulate the circuit breaker replacement plan and prepare spare parts in advance, effectively eliminate the resource waste or sudden failure risk caused by traditional regular maintenance, and ensure zero unplanned downtime of the power supply system in continuous production scenarios.

[0058] Based on the same inventive concept, as Figure 4 shown, the present invention also provides an intelligent monitoring and early warning method for the safety state of an electrical cabinet. The method includes: Obtain multi-source sensing data of the electrical cabinet, where the multi-source sensing data includes current waveform parameters, infrared temperature distribution parameters, and mechanical vibration spectrum parameters; is used to generate a health state feature vector of the electrical cabinet by using the multi-source sensing data; According to the matching result between the health state feature vector and a preset fault propagation rule library, output an early warning decision instruction; When the early warning decision instruction meets the preset execution conditions, perform safety protection actions.

[0059] It should be noted that for the formulas mentioned above, through the principle of dimensional consistency and mathematical standardization means (such as normalization processing, dimensionless parameter conversion, or unit system unification), physical quantities with different attributes can be translated into dimensionless standard values or superimposable parameters with the same dimension, so as to eliminate the interference of different dimensions on the operation logic, and make the formulas have mathematical operation rationality and objective law adaptability while retaining the original data distribution characteristics. This is a conventional technical means and will not be elaborated here. The electrical connections between the above-mentioned various units do not necessarily represent direct connections of the lines. Indirect connection methods can be applied to the embodiments of the present invention as long as the purpose of the present invention is achieved. The above are only exemplary embodiments of the present invention, and the scope of the present invention cannot be limited thereby.

[0060] That is, any equivalent changes and modifications made in accordance with the teachings of the present invention still fall within the scope covered by the present invention. Those skilled in the art will readily conceive of other embodiments of the present invention after considering the specification and the disclosure of the practical truth. This application is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of the present invention and include the common general knowledge or conventional technical means in the technical field not recorded in the present invention.

Claims

1. An intelligent monitoring and early warning system for the safety status of an electrical cabinet, characterized in that, The system includes: A data acquisition module for obtaining multi-source sensing data of the electrical cabinet, where the multi-source sensing data includes current waveform parameters, infrared temperature distribution parameters, and mechanical vibration spectrum parameters; A health feature evaluation module for generating a health state feature vector of the electrical cabinet by using the multi-source sensing data; An instruction generation module for outputting a warning decision instruction according to the matching result between the health state feature vector and a preset fault propagation rule library; A safety protection module for performing safety protection actions when the warning decision instruction meets the preset execution conditions.

2. The intelligent monitoring and early warning system for the safety status of an electrical cabinet according to claim 1, wherein The data acquisition module includes: A current parameter acquisition unit for acquiring current waveform parameters through a current transformer; A temperature parameter acquisition unit for scanning the surface of the cabinet body through an infrared thermal imaging array, positioning the spatial coordinates of the temperature anomaly area, and generating infrared temperature distribution parameters; A vibration parameter acquisition unit for detecting vibration signals through a piezoelectric film sensor array, identifying the energy peak value in the characteristic frequency band, and converting and generating mechanical vibration spectrum parameters.

3. The intelligent monitoring and early warning system for the safety status of an electrical cabinet according to claim 1 is characterized in that: The health feature evaluation module includes: An anomaly judgment unit for using a preset baseline data set to perform anomaly judgment on the current waveform parameters, the infrared temperature distribution parameters, and the mechanical vibration spectrum parameters, and generating a health state feature vector when the judgment result is abnormal; A current feature extraction unit for extracting the harmonic distortion rate and transient fluctuation characteristics in the current waveform parameters; A temperature feature extraction unit for extracting the temperature rise rate corresponding to the spatial coordinates in the infrared temperature distribution parameters based on the harmonic distortion rate; An electro-thermal feature extraction unit for calculating the electro-thermal coupling factor according to the harmonic distortion rate and the temperature rise rate; A mechanical-thermal feature extraction unit for generating a mechanical-thermal stress index according to the sequential change of the energy peak value in the characteristic frequency band of the mechanical vibration spectrum parameters and the temperature rise rate; A state feature extraction unit for performing feature fusion on the electro-thermal coupling factor and the mechanical-thermal stress index to construct a health state feature vector.

4. The intelligent monitoring and early warning system for the safety status of an electrical cabinet according to claim 3, characterized in that, The construction of the baseline data set includes: Collecting the current waveform parameters, infrared temperature distribution parameters, and mechanical vibration spectrum parameters under normal working conditions within a preset time period, and establishing a reference parameter mapping table, where the reference parameter mapping table includes a harmonic distortion rate threshold, a temperature threshold, and a vibration energy threshold; Obtaining the real-time load factor and dynamically correcting the temperature threshold based on the real-time load factor; Combining the corrected temperature threshold, the harmonic distortion rate threshold, and the vibration energy threshold to obtain a baseline data set.

5. The intelligent monitoring and early warning system for the safety status of an electrical cabinet according to claim 4, characterized in that, The instruction generation module includes: A component warning unit for generating a contact component degradation warning instruction when the electro-thermal coupling factor exceeds a preset coupling threshold and the mechanical-thermal stress index shows a monotonically increasing trend; An electrical warning unit for matching the health state feature vector with a preset fault propagation rule library and generating an electrical operation warning instruction according to the matching result; A safety time generation unit for extracting the change gradient of the health state feature vector to calculate the remaining safety time and attaching it to the warning instruction; Among them, the early warning decision instruction includes the deterioration early warning instruction of the contact component and the electrical operation early warning instruction.

6. The intelligent monitoring and early warning system for the safety status of an electrical cabinet according to claim 5, characterized in that, The construction of the fault propagation rule base includes: Obtaining the historical partial discharge intensity, historical cabinet door vibration acceleration, and historical current harmonic distortion rate, and establishing a three-dimensional correlation matrix; Obtaining the three-phase current unbalance degree and historical temperature rise rate in historical fault cases, analyzing the derivative relationship between the three-phase current unbalance degree and the historical temperature rise rate, and generating a fault path transfer function; Based on the three-dimensional correlation matrix and the fault path transfer function, constructing a fault propagation chain logic tree to obtain a fault propagation rule base.

7. An intelligent monitoring and early warning system for the safety status of an electrical cabinet according to claim 5, characterized in that, The safety protection module includes: A component protection unit, which is used to drive a preset sound and light alarm to output positioning information and remaining safety time when the early warning decision instruction is the deterioration early warning instruction of the contact component; An electrical protection unit, which is used to link corresponding electrical equipment to perform protection operations when the early warning decision instruction is the electrical operation early warning instruction, and generate fault location data and send it to a preset maintenance terminal; A work order generation unit, which is used to generate a predictive maintenance work order based on the health status feature vector and associate it with a preset spare parts inventory database.

8. An intelligent monitoring and early warning system for the safety status of an electrical cabinet according to claim 5, characterized in that, The system further includes: a hidden danger tracing module; among them, the hidden danger tracing module is used for: According to the spatial coordinates of the abnormal area corresponding to the mechanical thermal stress index, retrieving the electrical connection diagram of the corresponding position; Combining the device parameters in the electrical connection diagram and the phase characteristics of the electro-thermal coupling factor, positioning the faulty root device and generating a diagnostic report.

9. The intelligent monitoring and early warning system for the safety status of an electrical cabinet according to claim 8, characterized in that: The generation of the diagnostic report includes: Obtaining the installation torque record and historical maintenance log of the faulty root device, and calculating the cumulative damage factor; Based on the product result of the electro-thermal coupling factor and the cumulative damage factor, evaluating the remaining life of the device to obtain a remaining life evaluation value; Associating and storing the remaining life evaluation value with the priority of the maintenance work order to generate a diagnostic report.

10. An intelligent monitoring and early warning method for the safety status of an electrical cabinet, which applies the intelligent monitoring and early warning system for the safety status of an electrical cabinet described in any one of claims 1-9, is characterized in that The method includes: Obtaining multi-source sensing data of the electrical cabinet, where the multi-source sensing data includes current waveform parameters, infrared temperature distribution parameters, and mechanical vibration spectrum parameters; Using the multi-source sensing data to generate a health status feature vector of the electrical cabinet; According to the matching result between the health status feature vector and a preset fault propagation rule base, outputting an early warning decision instruction; When the early warning decision instruction meets the preset execution conditions, performing safety protection actions.

Images