Intelligent fault diagnosis system of electrical control cabinet

Abstract

The invention relates to the technical field of intelligent diagnosis, in particular to an intelligent fault diagnosis system of an electrical control cabinet, which comprises a thermal trend identification module, a density abrupt change monitoring module, a contact evolution labeling module, a structure aggregation judgment module and a fault risk output module. According to the method, a synchronous mapping mechanism of a thermal hysteresis sequence is introduced into a sequential relationship between heat change and energy release, the thermal stability degradation state is judged through difference trend identification, and a sudden anomaly identification strategy of multi-cycle dynamic fluctuation is established by combining a current density sudden change amplitude and a space node ratio difference. And then the contact performance degradation risk is judged through the synchronous change of the micro-contact resistance value and the temperature trend, the evolution process monitoring of the electrical connection contact is realized, the structure risk concentration area is identified according to the space aggregation effect of the contact, and the careful perception of the internal operation state of the control cabinet and the intelligent identification of the multi-dimensional abnormal evolution process are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent diagnosis, in particular to an intelligent fault diagnosis system of an electrical control cabinet. BACKGROUND

[0002] The technical field of intelligent diagnosis involves monitoring the running state of equipment, data collection and analysis to identify fault states and abnormal behaviors. The core tasks include feature extraction based on data or real-time data, fault pattern recognition, knowledge rule modeling, and automated fault identification methods. The overall technical field combines data mining, pattern recognition, and logical reasoning techniques to achieve intelligent evaluation and diagnosis of the running health status of objects such as industrial systems, power systems, and transportation systems by building expert knowledge bases or sample data sets. Among them, the intelligent fault diagnosis system of the traditional electrical control cabinet refers to an application system that detects abnormalities and judges faults based on the running state of the internal components of the electrical control cabinet. The technical matter it targets is the identification of faults caused by component aging, loose connections, or external interference during the operation of the electrical control cabinet. Traditional electrical control cabinet fault identification uses threshold determination methods such as temperature, current, and voltage to solve the technical matter. This method compares the single signal collected by the sensor with the pre-set threshold to determine whether there is an abnormality or fault.

[0003] Since the traditional method relies on the comparison of a single sensing signal and a static threshold, it lacks responsiveness when dealing with dynamic characteristics such as component thermal lag, electrical density mutation, or connection contact degradation. When there is a path delay in heat energy release or the contact state is in a slow-changing stage, it is easy to miss faults. Especially in the context of component aging or environmental interference, it is difficult to accurately identify abnormal states in a timely manner, causing hidden risks to accumulate during equipment operation. The current fixed threshold judgment logic based on temperature, current, and other physical quantities cannot cover the multi-source coupled fault phenomena that evolve gradually in complex systems. If only the traditional method is used for state monitoring in an electrical control cabinet, it will lead to problems such as fault prediction lag, fuzzy positioning range, and coarse diagnosis granularity. SUMMARY

[0004] In order to solve the technical problems that the prior art has the problems of insufficient response capability when coping with the dynamic characteristics of thermal effects of components, such as lag, sudden change of electrical density or degradation of connection contact, and easy occurrence of fault and missed detection when there is path delay in thermal energy release or the contact state is in a slow change stage, especially under the background of component aging or environmental interference, it is difficult to accurately identify abnormal states in time, resulting in accumulation of hidden risks in equipment operation, the judgment logic of setting fixed critical values based on physical quantities such as temperature and current cannot cover the multi-source coupled fault phenomena gradually evolving in a complex system, and if only the mode is relied on for state monitoring in an electrical control cabinet, the problems of lag in fault prediction, fuzzy positioning range and coarse diagnosis granularity will occur, the embodiments of the present application provide an intelligent fault diagnosis system for an electrical control cabinet. The technical scheme is as follows:

[0005] In one aspect, an intelligent fault diagnosis system for an electrical control cabinet is provided, which comprises:

[0006] A thermal trend identification module obtains the initial temperature data and running time data of a heating component in the electrical control cabinet, performs adjacent point temperature rise rate calculation on the temperature change sequence of the heating component, synchronously maps the temperature rise rate sequence and the thermal lag sequence on the time axis, performs difference change trend judgment between the corresponding points, and generates a component thermal degradation state label;

[0007] A density mutation monitoring module locates the current density monitoring nodes at the busbar and the wiring end in the connection path of the corresponding component based on the component thermal degradation state label, performs difference processing on the real-time and last current density values and records the mutation amplitude, and generates a path mutation abnormal record table;

[0008] A contact evolution labeling module collects the periodic voltage and current pairs of the connection contact according to the path mutation abnormal record table, performs voltage and current ratio operation to form a micro-contact resistance value sequence, identifies the continuous change trend of the micro-contact resistance value sequence with the period, and generates a contact point failure trend label;

[0009] A structure aggregation determination module, based on the contact point failure trend label, counts the spatial distribution position of the connection contact inside the electrical control cabinet, divides the position of the failure trend contact point, and generates a fault concentrated distribution mapping diagram.

[0010] As a further scheme of the present application, the component thermal degradation state label comprises thermal lag sequence change characteristics, temperature rise rate synchronization state and thermal stability decline signal, the path mutation abnormal record table comprises current density mutation amplitude, current density mutation ratio and mutation time period, the contact point failure trend label comprises micro-contact resistance value change trend, temperature synchronization rising period number and failure evolution risk state, and the fault concentrated distribution mapping diagram comprises failure trend contact point spatial distribution, structure focusing area position and area failure contact point number.

[0011] As a further scheme of the present application, the heat trend identification module comprises:

[0012] The temperature sequence acquisition submodule acquires the running initial temperature data and running time data of the heat generating elements in the electrical control cabinet, performs ratio calculation between the temperature difference and the time difference of adjacent temperature sampling points of any heat generating element, records the temperature rising rate per unit time represented by the ratio in the time period, and generates a temperature rising rate change sequence;

[0013] The thermal hysteresis sequence construction submodule calls the running time interval corresponding to the sampling points in the temperature rising rate change sequence, acquires the temperature release delay time value in the interval at the nodes in the energy release path, and sequentially accumulates and records the temperature release delay time value on the time axis to generate a thermal hysteresis change sequence;

[0014] The thermal stability judgment submodule filters the interval in which the temperature release delay time value continuously presents a monotonous decreasing trend based on the thermal hysteresis change sequence, calls the temperature rising rate change sequence value in the corresponding time period, judges whether it is continuously in the rising state, and if the above two conditions are met, records the thermal stability state identification value of the heat generating element, and acquires the element thermal degradation state label.

[0015] As a further scheme of the present application, the density mutation monitoring module comprises:

[0016] The monitoring node positioning submodule positions the busbar and the wiring end region in the current connection path where the heat generating element is located based on the heat generating element number information corresponding to the thermal degradation state label, acquires the current density acquisition node number in the monitoring period in the region, and obtains a density monitoring node list;

[0017] The mutation amplitude calculation submodule calls the density monitoring node list, calls the real-time current density value and the current density value of the last monitoring period of the node, performs difference processing on the current density change of the same node in two periods and records the result, extracts the adjacent node number of the node, compares the mutation amplitude value according to the spatial position relationship between the nodes, and obtains a node mutation amplitude ratio sequence;

[0018] The risk point identification submodule filters the record in which the amplitude ratio of the node is greater than three times of the corresponding adjacent node according to the node mutation amplitude ratio sequence, counts the number of marks in the continuous period, and if the mutation condition is met in two or more continuous periods, marks the node as a risk mutation point, and generates a path mutation abnormal record table by summarizing the risk mutation point number and the corresponding monitoring time.

[0019] As a further scheme of the present application, the contact evolution labeling module comprises:

[0020] The micro-resistance sequence submodule collects the voltage and current measurements of the connected contact points within the monitoring period according to the risk mutation points marked in the path mutation anomaly record table, performs a step-by-step ratio operation on the voltage and current values ​​within the period, uses the ratio as the periodic micro-contact resistance value of the connected contact point, arranges the ratio in chronological order, and generates a micro-contact resistance value sequence for the numbered contact points.

[0021] The failure trend identification submodule calls the continuous periodic values ​​of the measurement points in the micro-contact resistance value sequence of the numbered contact, identifies the direction of change within five or more periods, determines whether it shows a monotonically increasing trend, calls the corresponding periodic connection contact temperature change value sequence, and performs a monotonically increasing trend identification operation on the connection contact temperature change value sequence within the same period range. If both sets of sequences continue to rise for more than five periods, the connection contact is marked as a failure evolution risk node, and a contact failure trend mark is generated.

[0022] As a further aspect of the present invention, the process of identifying the direction of change within five or more cycles is as follows:

[0023] Collect the micro-contact resistance values ​​of any six consecutive cycles in the micro-contact resistance value sequence of the numbered contact, and determine whether the micro-contact resistance values ​​of the six cycles sequentially satisfy the condition that the difference between any adjacent cycles is greater than 0.2 milliohms.

[0024] If at least five of the six cycles meet the conditions, it is considered to show a monotonically increasing trend;

[0025] The process of calling the corresponding period's connection contact temperature change value sequence and performing monotonic trend identification operation on the connection contact temperature change value sequence within the same period range is as follows:

[0026] Collect the temperature change sequence of the contact points corresponding to the six cycles, and determine whether the temperature difference between any adjacent cycles is greater than 1.5 degrees Celsius. If at least five of the six cycles meet the condition, it is considered that the temperature value shows a monotonically increasing trend.

[0027] As a further aspect of the present invention, the structure aggregation determination module includes:

[0028] The location clustering submodule, based on the contact point failure trend marker, counts the spatial positioning coordinates corresponding to the contact point numbers, calculates the spatial offset feature value, and performs location classification operations on the connection contacts with failure trends according to the preset area grid division in the electrical control cabinet structure diagram, generating an area failure contact distribution table.

[0029] The focus area identification submodule calls the number of failure trend contact points in the region from the regional failure contact point distribution table, performs a cumulative operation on the number of regional contact points, filters the region numbers whose number of failure trend contact points exceeds the preset structural focus threshold, marks them as structural focus areas in spatial distribution, and integrates the focus area location information with the number to generate a fault concentration distribution mapping map.

[0030] As a further aspect of the present invention, in the regional failure contact distribution table, the corresponding failure trend connection contact number is extracted for each regional number, and the total number of the number is counted.

[0031] The process of filtering the region numbers whose number of failure trend contact points exceeds the preset structural focus threshold is as follows: compare the number of failure trend contact points corresponding to the region number with the set structural focus threshold 20, and only retain the region numbers whose number of failure trend contact points is greater than 20.

[0032] The process of mapping and integrating the focus area location information with the number is as follows:

[0033] Extract the spatial coordinate range of each of the filtered area numbers in the electrical control cabinet structure diagram, integrate them one-to-one according to the area number, and form the integrated result into the fault concentration distribution mapping map.

[0034] As a further aspect of the present invention, the system also includes a fault risk output module:

[0035] The fault risk output module records and counts the number of risk elements in the area based on the fault concentration distribution mapping map and the component thermal degradation status label and contact point failure trend mark of the corresponding components in the structural focus area. When the number in the area meets the set conditions at the same time, the diagnostic output logic is triggered to generate a set of regional fault diagnosis conclusions.

[0036] The regional fault diagnosis conclusion set includes the structural focal area identification results, component degradation risk level, and contact point failure trend indicators.

[0037] As a further aspect of the present invention, the fault risk output module includes:

[0038] The risk element aggregation submodule calls the component number and contact point number information included in the region according to the area number of the structural focus area in the fault concentration distribution map, extracts the thermal degradation status label of the corresponding component and the contact point failure trend mark of the corresponding contact, counts the number of risk elements by region dimension, and generates a risk element statistics table for the focus area.

[0039] The diagnostic trigger judgment submodule calls the risk element statistics table of the focus area, and compares the number of risk elements with the set thermal degradation element judgment value, failure trend mark judgment value and combined threshold standard in turn. If all three conditions are met, the area number is recorded as the area with diagnostic output conditions, and the diagnostic output trigger area number set is obtained.

[0040] The conclusion construction submodule integrates the component number, contact point number, associated thermal degradation status label, and contact point failure trend mark content in the corresponding area according to the diagnostic output trigger area number set, constructs structured record entries, summarizes and outputs diagnostic entries according to area number, and generates a set of regional fault diagnosis conclusions.

[0041] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0042] By introducing a synchronous mapping mechanism of thermal hysteresis sequence into the temporal relationship between heat change and energy release, the judgment of thermal stability degradation state is realized through differential trend identification. Combining the difference in current density mutation amplitude and spatial node ratio, a sudden anomaly identification strategy for multi-cycle dynamic fluctuation is established. Then, the risk of contact performance degradation is judged by the synchronous change of micro-contact resistance and temperature trend, realizing the monitoring of the evolution process of electrical connection contact points. Based on the spatial aggregation effect of contact points, the structural risk concentration area is identified. In the regional statistical logic, three types of label indicators, namely thermal degradation, density mutation and contact trend, are integrated to construct an association triggering mechanism under multi-source diagnostic conditions. The technical path integrates multi-dimensional trend data of thermal dynamic response, electric density change and contact impedance evolution to achieve detailed perception of the internal operating status of the control cabinet and intelligent identification of multi-dimensional anomaly evolution process, thereby improving the accuracy of fault identification and the precision of fault location. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a schematic diagram of the system provided by the present invention;

[0045] Figure 2 This is a schematic diagram of the system framework of the present invention;

[0046] Figure 3 This is a flowchart of the heat trend recognition module in this invention;

[0047] Figure 4 This is a flowchart of the density mutation monitoring module in this invention;

[0048] Figure 5 This is a flowchart of the contact evolution annotation module in this invention;

[0049] Figure 6 This is a flowchart of the structure aggregation determination module in this invention;

[0050] Figure 7 This is a flowchart of the fault risk output module in this invention. Detailed Implementation

[0051] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0052] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0053] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0054] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0055] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0056] This invention provides an intelligent fault diagnosis system for electrical control cabinets, such as... Figures 1-2 The diagram shown illustrates an intelligent fault diagnosis system for an electrical control cabinet. This system includes:

[0057] The thermal trend recognition module acquires the initial operating temperature data and operating time data of the heating element in the electrical control cabinet, performs adjacent point temperature rise rate calculation on the temperature change sequence of the heating element, records the corresponding time period, obtains the energy release delay value in the heat dissipation path, performs accumulation to form a thermal hysteresis sequence within the path node, synchronously maps the temperature rise rate sequence and the thermal hysteresis sequence on the time axis, performs difference change trend judgment between corresponding points, when multiple consecutive segments in the thermal hysteresis sequence show a downward trend and the temperature rise rate remains in an upward state, records the heating element state as thermal stability degradation, and generates a component thermal degradation state label;

[0058] The density mutation monitoring module locates the current density monitoring nodes at the busbar and terminals in the corresponding component connection path based on the component thermal degradation status label. It performs difference processing on the real-time and previous current density values ​​and records the mutation amplitude. It extracts the mutation amplitude of adjacent positions of the node and compares the amplitude ratio. When the mutation amplitude exceeds three times that of the adjacent node and lasts for more than two cycles, it is marked as a risk mutation point and generates a path mutation anomaly record table.

[0059] The contact evolution labeling module collects periodic voltage and current pairs at the connecting contact points based on the path mutation anomaly record table, performs voltage and current ratio operation to form a micro-contact resistance sequence, identifies the continuous change trend of the micro-contact resistance sequence with the period, and combines the temperature change value of the connecting contact point to perform a synchronization judgment of the two sets of trend data. If both continue to rise for more than five periods, the connecting contact point is marked as a failure evolution risk node, and a contact point failure trend mark is generated.

[0060] The structural clustering determination module is based on the contact point failure trend markers, statistically analyzes the spatial distribution of the contact points inside the electrical control cabinet, divides the location of the contact points with failure trends, performs cumulative judgment on the number of failure trends in the area, records the dense area as the structural focus area, and generates a fault concentration distribution mapping map.

[0061] The fault risk output module records and counts the number of risk elements in the area based on the fault distribution map and the component thermal degradation status label and contact point failure trend mark of the corresponding components in the structural focus area. When the number of risk elements in the area meets the set conditions at the same time, the diagnostic output logic is triggered to generate a set of regional fault diagnosis conclusions.

[0062] The component thermal degradation status label includes thermal hysteresis sequence change characteristics, temperature rise rate synchronization status, and thermal stability decline signal. The path mutation anomaly record table includes current density mutation amplitude, current density mutation ratio, and mutation time period. The contact point failure trend marker includes micro-contact resistance change trend, number of synchronous temperature rise cycles, and failure evolution risk status. The fault concentration distribution mapping map includes the spatial distribution of failure trend contact points, the location of structural focal area, and the number of regional failure contacts. The regional fault diagnosis conclusion set includes the structural focal area identification results, component degradation risk level, and contact point failure trend index.

[0063] Specifically, such as Figure 2 , 3 As shown, the heat trend recognition module includes:

[0064] The temperature sequence acquisition submodule acquires the initial operating temperature data and operating time data of the heating elements in the electrical control cabinet. It calculates the ratio between the temperature difference and the time difference for adjacent temperature sampling points of any heating element, records the temperature rise rate per unit time represented by the ratio within the time period, and generates a temperature rise rate change sequence.

[0065] Multiple temperature acquisition devices need to be installed, such as thermocouples or infrared temperature sensors deployed at contactors, frequency converters, reactors, etc., to ensure a constant acquisition frequency, such as recording data every 10 seconds. Each data entry includes the component number, sampling timestamp, and sampling temperature. Two consecutive data points are processed adjacently, and the temperature change value is directly divided by the time interval value to obtain the temperature rise rate. The rates are then arranged in chronological order to form a temperature rise rate change sequence. For example, if the temperature of component number A is 35℃ at a certain time point and 36.5℃ 10 seconds later, the temperature rise rate between these two points is 0.15℃ per second. During continuous acquisition, if a data point changes abruptly from the previous value beyond the set standard, the system will automatically determine it as invalid data and discard it to ensure the stability of the temperature rise rate sequence. For different types of heating elements, the effective temperature rise rates are uniformly organized into structured data to generate a temperature rise rate change sequence.

[0066] The thermal hysteresis sequence construction submodule calls the running time interval corresponding to the sampling point in the temperature rise rate change sequence, obtains the temperature release delay time value within the interval at the node in the energy release path, and accumulates and records the temperature release delay time value sequentially on the time axis to generate the thermal hysteresis change sequence.

[0067] The target range is determined based on the temperature rise rate change sequence. For example, a time period with a significant rate change is selected as the analysis object. Simultaneously, the system automatically retrieves data from sensor points installed on the heat transfer path outside the component and compares it synchronously with the temperature curve of the target component. If a certain segment of the rate curve shows a continuous downward trend, the system uses this time period as a benchmark to search for a corresponding temperature response signal delay in the heat transfer path and extracts this time difference as the thermal hysteresis time. The time differences are recorded sequentially to form a hysteresis time list. Throughout the process, the acquisition system must ensure the consistency of sensor timestamps and set the range boundaries of the hysteresis time to avoid invalid extreme value data. In practical applications, if the temperature of the inverter's core components changes, the system will determine whether the hysteresis time of heat transfer to the aluminum heat sink is within the normal range. If it is not within the reasonable range, the data is marked as invalid, and the valid delay time values ​​are cascaded and added to the change sequence in chronological order to generate a thermal hysteresis change sequence.

[0068] The thermal stability determination submodule is based on the thermal hysteresis change sequence. It filters the interval where the temperature release delay time value shows a monotonically decreasing trend for multiple consecutive segments, calls the temperature rise rate change sequence value within the corresponding time period, and determines whether it is continuously in an upward state. If the above two conditions are met, the thermal stability status identifier value of the heating element is recorded, and the element thermal degradation status label is obtained.

[0069] By using the established judgment rules, the system searches for multiple consecutive decreasing time periods in the sequence. For example, if the first of three consecutive data points is greater than the next, it can be considered a segment that meets the condition. The filtering process uses a sliding window strategy to gradually search for time periods that satisfy the decreasing relationship. After finding them, the system calls the data in the temperature rise rate sequence within the corresponding time period and judges whether the temperature rise rate shows a continuous upward trend. If the rate continues to increase within the time period, the system records the thermal stability status indicator for that time. Multiple indicator records are accumulated one by one. According to the established rules, if the stability condition is not met three times in a row, the component is considered to have a thermal degradation trend. The judgment process is applied to high-frequency working components in the control cabinet, such as motor drive modules and rectifiers. By analyzing the correspondence between hysteresis changes and rate changes in real time, the system identifies unstable signs that occur during operation and obtains the component thermal degradation status label.

[0070] Specifically, such as Figure 2 , 4 As shown, the density mutation monitoring module includes:

[0071] The monitoring node positioning submodule locates the busbar and terminal area in the current connection path where the heating element is located based on the heating element number information corresponding to the thermal degradation status label, obtains the current density acquisition node number corresponding to the monitoring cycle in the area, and obtains a density monitoring node list.

[0072] Extract the unique identification information of the components. Based on the control cabinet design drawings and circuit logic structure, unfold the upstream and downstream connections of the heating components in the current path step by step to identify the corresponding busbar connection area and the specific location of the terminal. For example, if a heating component is numbered A12, its branch can be located in the right terminal area of ​​the L1 section of the main busbar by designing the circuit diagram. Within the located area, retrieve the current density acquisition node number data belonging to the area from the monitoring records. Based on the actual component layout labels, such as sensor numbers ICD01 to ICD06, determine the list of nodes participating in the monitoring of the area in the current monitoring cycle. Since the current sensors in the electrical cabinet are deployed in a matrix or linear array, it is necessary to find the numbers in combination with the specific layout model. For example, the distributed current density monitoring unit on the busbar deploys a node every 0.5 meters, recording the numbers ICD01, ICD02, etc. By calling the node-location mapping table in the system, filter all monitoring node numbers covering the component path area, arrange them in numerical order, and obtain the density monitoring node list.

[0073] The mutation amplitude calculation submodule calls the density monitoring node list, calls the real-time current density value of the node and the current density value of the previous monitoring period, performs difference processing on the current density change of the same node in two periods and records the results, extracts the adjacent node numbers of the node, compares the mutation amplitude value according to the spatial position relationship between the nodes, and obtains the node mutation amplitude ratio sequence.

[0074] The system retrieves the current density data of the node in the current monitoring cycle and simultaneously retrieves the data of the same node in the previous monitoring cycle. Assuming the monitoring value of ICD03 in cycle T1 is 430 A / cm² and in cycle T0 is 395 A / cm², the change in the current cycle is 35 A / cm². All nodes are processed in a similar manner, calculating the change in current density and storing it in a temporary list. The system retrieves the adjacent node numbers for each node; for example, the adjacent nodes of ICD03 are ICD02 and ICD04. Combining this with the physical distances of each node in the 3D layout model of the control cabinet, indirect adjacent nodes are excluded. The system retains only the actual adjacent nodes. For each pair of adjacent nodes, the difference in current density change within the same period is calculated. For example, if ICD03 changes to 35A / cm², while ICD02 changes to only 10A / cm², the ratio is 3.5. The monitoring nodes are traversed in this way, and the ratio of the change amplitude between nodes is recorded in numerical order. This serves as the basic data for subsequent risk point identification. The ratio calculation is required to consider the effects of node spacing, path overlap area, and wiring lead overlap, ensuring that the comparison is only performed between nodes with correlation to prevent misjudgment and obtain the node change amplitude ratio sequence.

[0075] The risk point identification submodule filters records whose node amplitude ratio is more than three times that of their corresponding adjacent nodes based on the node mutation amplitude ratio sequence, counts the number of markings in consecutive periods, and marks the node as a risk mutation point if the mutation condition is met for two or more consecutive periods. It then summarizes the risk mutation point numbers and corresponding monitoring times to generate a pathway mutation anomaly record table.

[0076] The system filters data items to find node records with a ratio significantly higher than three times that of adjacent nodes. The judgment threshold is set as a mutation amplitude ratio exceeding 3.0 times. Starting from each monitoring node, the system gradually scans the ratio sequence. When a node is found to have a mutation amplitude ratio greater than 3.0 and the ratios of its upstream and downstream adjacent nodes are all below the value, the node is marked as a suspected risk mutation point in the current period. To avoid misjudgment caused by instantaneous fluctuations, the system counts the mutation situation of nodes in multiple consecutive periods. If a node meets the mutation condition in two or more consecutive periods, that is, it is repeatedly marked as a risk mutation point, then it is confirmed as an actual risk node. The node number and timestamp are recorded. For example, if ICD03 meets the mutation ratio condition on October 1 and October 2, 2025, a pathway mutation anomaly record table is generated.

[0077] Specifically, such as Figure 2 , 5 As shown, the contact evolution annotation module includes:

[0078] The micro-resistance sequence submodule collects the voltage and current measurements of the connected contact points within the monitoring period based on the risk mutation points marked in the path mutation anomaly record table and the corresponding connected contact point numbers. It performs a ratio operation on the voltage and current values ​​within the period and uses the ratio as the periodic micro-contact resistance value of the connected contact point. The ratios are arranged in chronological order to generate a micro-contact resistance value sequence for the numbered contacts.

[0079] The system extracts the identification information of the risk mutation points and locates the corresponding contact point numbers. For example, the contact point number corresponding to the mutation point ICD05 is CT05. The system retrieves the real-time voltage and current measurement values ​​of CT05 within the current monitoring cycle, with a sampling frequency of once every 10 seconds. It ensures that the voltage and current data sources come from the same sensor or time synchronization source. At each sampling moment, the corresponding voltage and current values ​​are extracted and paired. For example, the voltage at the first sampling point is 2.4mV and the current is 80A, and the voltage at the second point is 2.5mV and the current is 79.5A. The system performs division on each pair of voltage and current data to obtain the micro-contact resistance value at the time point. After the sampling points are processed, each micro-contact resistance value is recorded as an item according to the sampling order. The system continuously records data from multiple cycles by day or hour to form a cross-cycle time series. The series uses the contact point number as the index item. Each record includes a timestamp, the measured resistance value, and the source sensor number. This process is performed in parallel on multiple contact point numbers to generate a micro-contact resistance value sequence for the numbered contacts.

[0080] The failure trend identification submodule calls the continuous periodic values ​​of the measuring points in the micro-contact resistance value sequence of the numbered contact point, identifies the direction of change within five or more periods, determines whether it shows a monotonically increasing trend, calls the corresponding periodic connection contact temperature change value sequence, and performs monotonic trend identification operation on the connection contact temperature change value sequence within the same period range. If both sets of sequences continue to rise for more than five periods, the connection contact point is marked as a failure evolution risk node, and a contact point failure trend mark is generated.

[0081] The system analyzes the values ​​of the latest five or more consecutive monitoring cycles. It identifies the increasing trend of the five resistance values ​​in chronological order. For example, if the resistance values ​​of the previous five cycles are 29μΩ, 31μΩ, 33μΩ, 34μΩ, and 36μΩ, the system identifies this as a monotonically increasing trend. If the resistance value in a certain cycle is equal to the previous value, such as 32μΩ, 32μΩ, 33μΩ, 34μΩ, and 35μΩ, it does not constitute a monotonically increasing trend and the data is directly discarded. After completing the resistance trend identification, the system retrieves the temperature change value sequence of the corresponding contact within the same cycle range. For example, if the temperature is 62℃, 64℃, 66℃, 68℃, and 70℃, it also constitutes a monotonically increasing trend. If both sequences meet the condition of increasing for more than five consecutive cycles, the system immediately marks the contact point as having a failure evolution risk status. After the marking is completed, the system generates a contact point failure trend mark with fields such as contact number, number of consecutive increasing cycles, start date, and end resistance temperature.

[0082] Specifically, such as Figure 2 , 6 As shown, the structure aggregation determination module includes:

[0083] The location clustering submodule is based on the contact point failure trend marker, statistically analyzes the spatial positioning coordinates corresponding to the contact point number, calculates the spatial offset feature value, and performs location classification operation on the connection contacts with failure trends according to the preset area grid division in the electrical control cabinet structure diagram, generating a regional failure contact distribution table.

[0084] The spatial offset feature value is expressed by the formula:

[0085] ;

[0086] in, Representing the The contact point of the first connection is on the X-axis and the second... The spatial offset characteristic value of the contact point on the Y-axis. Representing the The X-axis coordinates of the contact points are connected. This represents the average value of the X-axis coordinate of the contact point. Representing the The Y-axis coordinate value of the contact point. This represents the average value of the Y-axis coordinate of the contact point. Representing the The Z-axis coordinate value of the contact point. This represents the average value of the Z-axis coordinates of all connected contact points. This represents the total number of contact points. , , These are the indexed variables in the three-dimensional coordinate space;

[0087] Formula calculation logic: By calculating the three-dimensional offset of the contact point relative to the spatial center, a geometric feature value that comprehensively reflects the positional difference is constructed. In the numerator, the coordinate difference of the contact point in the X-axis and Y-axis directions is taken and multiplied to represent the dual-axis coordinated offset trend on the horizontal plane. In the denominator, the absolute offset of the contact point in the Z-axis direction is summed, 1 is added, and then squared to form a weighted normalization factor in the depth direction, so that the spatial height fluctuation has a constraining effect on the planar offset. By taking the absolute value and the square root, a non-negative spatial offset scale is obtained to quantify the degree of overall deviation of the contact point from the center. The overall calculation logic takes into account the coordinated difference of the three-dimensional coordinates and the normalization correction, realizes the structured expression of the positional offset intensity, and makes the spatial distribution difference of each contact point comparable and clusterable in numerical form.

[0088] Spatial offset characteristic value represents the degree of deviation of a single contact point from the spatial center of the contact point. It comprehensively considers the coordinate changes of the contact point in the horizontal and vertical directions, reflecting the overall positional differences in the three-dimensional structure. The larger the value, the more the contact point deviates from the center position, and the smaller the value, the closer the contact point distribution is to the overall equilibrium state.

[0089] Parameter acquisition and calculation instructions:

[0090] , , The spatial positions of the connection points in the power distribution control cabinet were obtained point by point using an industrial camera and 3D measuring equipment. For the five connection points numbered 1 to 5, the following coordinates were obtained through 3D scanning:

[0091] ;

[0092] ;

[0093] ;

[0094] ;

[0095] ;

[0096] ;

[0097] Example calculation process (in) , For example, calculate the spatial offset characteristic value of the third contact point):

[0098] Molecular part:

[0099] ;

[0100] Denominator:

[0101] ;

[0102] Normalization factor:

[0103] ;

[0104] Substitute into the calculation:

[0105] ;

[0106] Calculation data display:

[0107] Table 1: Spatial Coordinate Sampling and Eigenvalue Calculation Table

[0108] Contact Number (mm) (mm) (mm) 1 122.5 75.4 63.2 — 2 125.0 74.9 65.1 — 3 128.3 76.8 64.5 0.3109 4 124.1 75.5 66.0 — 5 126.7 75.0 62.8 —

[0109] As shown in Table 1, the offset value of the third contact point is 0.3109, which is within the preset benchmark threshold of 0.25 to 0.35, and belongs to the slight offset level.

[0110] Threshold setting instructions and example explanations:

[0111] Offset strength level classification is based on failure data statistics and is set as follows: Normal: Slight deviation: Moderate offset: Severe offset: ;

[0112] The third contact in the example Therefore, it is classified as a slightly offset contact and will be marked for failure clustering and regional hot spot analysis;

[0113] Example of quantization for non-numerical data (illustrated explanation)

[0114] To introduce non-numerical characteristic parameters such as "assembly tension" or "electrical connection material," these can be converted into dimensionless values ​​using quantification standards. For example:

[0115] The "tension state" can be quantified as follows: relaxed = 0.2, normal = 0.5, overly tight = 0.8 (based on the tension distribution statistics of the dynamic tension sensor).

[0116] The "Material Type" is set according to conductivity: Copper = 1.0, Aluminum = 0.6, Stainless Steel = 0.3 (based on normalized unit resistivity).

[0117] If contact 3 is made of copper and the tension is normal, then the material coefficient = 1.0 and the tension coefficient = 0.5, which can be used in the next stage of model building.

[0118] The advantage of the formula is that by introducing the sum of the absolute values ​​of the Z-axis offset and squaring them, it is integrated with the offset trend in the XY direction. Without having to directly construct the three-dimensional vector distance, a normalized scalar value that reflects the differences in spatial disturbance distribution is constructed. This makes the classification and partitioning of contact point offset under complex structures more distinguishable, which is beneficial to the subsequent hot spot clustering and failure localization process.

[0119] The results show that the third contact point is significantly deviated from the center position in the XY plane, and the offset is suppressed and adjusted by the spatial fluctuation of the Z-axis. The spatial offset feature value of 0.3109 reaches the set slight offset range. The spatial offset feature value will be used as a spatial offset index to participate in the generation of cluster labels when classifying the distribution of regional failure contacts in the next stage, supporting the construction of the failure trend distribution map.

[0120] The focus area identification submodule calls the number of failure trend contact points in the area from the regional failure contact point distribution table, performs a cumulative operation on the number of regional contact points, filters the area numbers whose number of failure trend contact points exceeds the preset structural focus threshold, marks them as structural focus areas in the spatial distribution, and integrates the focus area location information with the number to generate a fault concentration distribution mapping map.

[0121] The system counts the number of failure trend contact points within each grid number. It iterates through the area numbers and calculates the corresponding number of failure contacts in real time. For example, area number G015 contains 6 failure contacts, including CT03, CT08, and CT11, so the area's failure contact count is 6. The system pre-sets a structural focus threshold, set to 5 contacts based on the component density, heat dissipation capacity, and typical layout of the electrical control cabinet. If the number of failure contacts within an area number exceeds the threshold, it is identified as a spatial structural focus area. After identification, the system marks the focus area number and extracts the geometric center coordinates of the corresponding grid as positioning information. This information is then bound to the number to form a complete location label. The focus area number and location information are summarized through a visualization module. Each focus area is marked with a specific color and includes an area number and a label indicating the number of failure contacts. This assists maintenance personnel in determining the range of concentrated fault risks, realizing a spatial focus presentation of local degradation trends within the structure, and generating a fault concentration distribution mapping map.

[0122] Specifically, such as Figure 2 , 7 As shown, the fault risk output module includes:

[0123] The risk element aggregation submodule calls the component number and contact point number information included in the region according to the area number of the structural focus area in the fault concentration distribution mapping map, extracts the thermal degradation status label of the corresponding component and the contact point failure trend mark of the corresponding contact, counts the number of risk elements by region dimension, and generates a risk element statistics table for the focus area.

[0124] The system sequentially retrieves the component numbers and contact point numbers covered within the corresponding area. After locating the area number in the fault mapping map, it sequentially searches for the components and contacts bound to the number. For example, area G008 includes components M05 and M09 and contacts CT04, CT05, and CT07. It calls the thermal degradation status label corresponding to each component and extracts the label status (e.g., a status value of 1 represents thermal degradation, and a value of 0 represents no degradation). It reads the contact point failure trend label status corresponding to the contact point number within the area and extracts the label values ​​in the same format. The system records the above two types of label information by number to form a basic dataset. The system accumulates the number of component degradation labels and contact point failure labels for each focused area and counts the third type of risk element, i.e., the number of combinations of degradation and failure labels. All statistical results are indexed by "area number" and the number of the three types of risk elements are used as values ​​to generate a statistical table of risk elements in the focused area.

[0125] The diagnostic trigger judgment submodule calls the risk element statistics table of the focus area, and compares the number of risk elements with the set thermal degradation element judgment value, failure trend mark judgment value and combined threshold standard in turn. If all three conditions are met, the area number is recorded as the area with diagnostic output conditions, and the set of diagnostic output trigger area numbers is obtained.

[0126] According to the established judgment logic, the statistical values ​​of the three types of risk elements under each area number are compared sequentially. The judgment criteria set by the system include: the judgment value of thermal degradation element is set to ≥2, the judgment value of failure trend mark is set to ≥2, and the threshold of composite item combination is ≥1. The system judges each record in the statistical table. If all three values ​​meet the set conditions, the area number is marked as an area with diagnostic output conditions. For example, the record of area number G010 shows that the number of thermal degradation elements is 3, the number of failure contacts is 4, and the number of composite items is 2. All three values ​​are greater than the set standards, so the area meets the trigger conditions. The system scans the records one by one to obtain the set of diagnostic output trigger area numbers.

[0127] The conclusion construction submodule integrates the component number, contact point number, associated thermal degradation status label and contact point failure trend mark in the corresponding area according to the diagnostic output trigger area number set, constructs structured record entries, summarizes and outputs diagnostic entries according to area number, and generates a set of regional fault diagnosis conclusions.

[0128] The system sequentially retrieves the component numbers and contact point numbers under each number, then integrates the corresponding thermal degradation status labels and contact point failure trend marker values ​​to form structured record entries. Each entry uses the region number as the top-level index and includes component number and label fields, contact point number and marker fields. It also records the periodic time range of the acquired labels and the corresponding monitoring number. For example, if components M01 and M04 in region G015 have thermal degradation status and contact points CT02 and CT03 have failure trend markers, then the system constructs a record containing "region number G015", "thermal degradation components = [M01, M04]", "failed contact = [CT02, CT03]", and "label period = period 20 to 25". All region numbers that meet the diagnostic output conditions are organized into such structured entries. The system integrates the entry set according to the number order to generate a set of regional fault diagnosis conclusions.

[0129] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An intelligent fault diagnosis system for an electrical control cabinet, characterized in that, The system comprises: The heat trend identification module obtains the running initial temperature data and running time data of the heating element in the electrical control cabinet, performs adjacent point temperature rise rate calculation on the temperature change sequence of the heating element, synchronously maps the temperature rise rate sequence and the heat hysteresis sequence on the time axis, performs difference change trend judgment between corresponding points, and generates an element heat degradation state label; The density mutation monitoring module locates the current density monitoring nodes at the busbar and the terminal end in the corresponding element connection path based on the element heat degradation state label, performs difference processing on the real-time and last current density values and records the mutation amplitude, and generates a path mutation abnormal record table; The contact evolution labeling module collects the periodic voltage and current pairs of the connection contact according to the path mutation abnormal record table, performs voltage and current ratio operation to form a micro-contact resistance value sequence, identifies the continuous change trend of the micro-contact resistance value sequence with the period, and generates a contact point failure trend marker; The structure aggregation determination module counts the spatial distribution position of the connection contact inside the electrical control cabinet based on the contact point failure trend marker, divides the position of the failure trend contact point, and generates a fault concentrated distribution mapping diagram.

2. The intelligent fault diagnosis system of an electrical control cabinet according to claim 1, characterized in that: The element heat degradation state label comprises heat hysteresis sequence change characteristics, temperature rise rate synchronization state, and heat stability decline signal, the path mutation abnormal record table comprises current density mutation amplitude, current density mutation ratio, and mutation time period, the contact point failure trend marker comprises micro-contact resistance value change trend, temperature synchronization rising period number, and failure evolution risk state, and the fault concentrated distribution mapping diagram comprises failure trend contact point spatial distribution, structure focusing area position, and area failure contact point number.

3. The intelligent fault diagnostic system for electrical control cabinets of claim 1, wherein: The heat trend identification module comprises: The temperature sequence acquisition submodule obtains the running initial temperature data and running time data of the heating element in the electrical control cabinet, performs temperature difference and time difference ratio calculation between adjacent temperature sampling points of any heating element, records the temperature rise rate per unit time represented by the ratio in the time period, and generates a temperature rise rate change sequence; The heat hysteresis sequence construction submodule calls the running time interval corresponding to the sampling points in the temperature rise rate change sequence, obtains the temperature release delay time value in the interval at the node in the energy release path, and sequentially accumulates and records the temperature release delay time value on the time axis to generate a heat hysteresis change sequence; The heat stability determination submodule filters the interval in which the temperature release delay time value continuously presents a monotone decreasing trend based on the heat hysteresis change sequence, calls the temperature rise rate change sequence value in the corresponding time period, judges whether it is continuously in the rising state, and if the above two conditions are met, records the heat stability state identification value of the heating element, and obtains the element heat degradation state label.

4. The intelligent fault diagnostic system for electrical control cabinets of claim 3, wherein: The density mutation monitoring module comprises: The monitoring node positioning submodule locates the busbar and terminal end region in the current connection path of the heating element based on the heating element number information corresponding to the heat degradation state label, obtains the current density acquisition node number corresponding to the monitoring period in the region, and obtains a density monitoring node list; The mutation amplitude calculation submodule calls the density monitoring node list, calls the real-time current density value of the node and the current density value of the last monitoring period, performs difference processing on the current density change of the same node in two periods and records the result, extracts the adjacent node number of the node, compares the mutation amplitude values according to the spatial position relationship between the nodes, and obtains a node mutation amplitude ratio sequence; The risk point identification submodule filters the records of the node amplitude ratio greater than three times the corresponding adjacent node according to the node mutation amplitude ratio sequence, counts the number of marks in the continuous period, marks the node as a risk mutation point if the mutation condition is met in two or more continuous periods, and generates a path mutation abnormal record table by summarizing the risk mutation point number and the corresponding monitoring time.

5. The intelligent fault diagnostic system for electrical control cabinets of claim 4, wherein: The contact evolution labeling module comprises: The micro-resistance sequence submodule collects the voltage and current measurement values of the connection contact point in the monitoring period according to the connection contact point number corresponding to the risk mutation point marked in the path mutation abnormal record table, performs item-by-item ratio operation on the voltage and current values in the period, takes the ratio as the periodic micro-contact resistance value of the connection contact point, arranges the ratio in time sequence, and generates a micro-contact resistance value sequence of the numbered contact point; The failure trend identification submodule calls the continuous period values of the measurement point in the micro-contact resistance value sequence of the numbered contact point, identifies the change direction in five or more periods, judges whether it shows a monotonic increasing trend, calls the connection contact point temperature change value sequence of the corresponding period, performs a monotonic trend identification operation on the connection contact point temperature change value sequence in the same period range, and marks the connection contact point as a failure evolution risk node if both sequences continue to rise for more than five periods, and generates a contact point failure trend mark.

6. The intelligent fault diagnostic system of an electrical control cabinet according to claim 5, characterized in that: The process of identifying the change direction in five or more periods is: Collect the micro-contact resistance values of any six consecutive periods in the micro-contact resistance value sequence of the numbered contact point, and judge whether the micro-contact resistance values of the six periods sequentially satisfy the condition that the difference between any adjacent periods is greater than 0.2 milliohm; If at least five periods in the six periods satisfy the condition, it is determined that a monotonic increasing trend is shown; The process of calling the connection contact point temperature change value sequence of the corresponding period and performing a monotonic trend identification operation on the connection contact point temperature change value sequence in the same period range is: Collect the connection contact point temperature change value sequence corresponding to the six periods, judge whether the temperature value difference between any adjacent periods is greater than 1.5 degrees Celsius, and if at least five periods in the six periods satisfy the condition, it is determined that the temperature value shows a monotonic increasing trend.

7. The intelligent fault diagnostic system for electrical control cabinets of claim 5, wherein: The structure aggregation determination module comprises: The position clustering division submodule counts the spatial positioning coordinate value corresponding to the connection contact point number based on the contact point failure trend mark, calculates the spatial offset characteristic value, and performs a position classification operation on the connection contact point with a failure trend according to the preset regional grid division in the electrical control cabinet structure diagram, and generates a regional failure contact point distribution table; The focus area identification submodule calls the number of failure trend contact points in the region in the region failure contact point distribution table, performs an accumulation operation on the number of region contact points, filters the region numbers whose number of failure trend contact points exceeds the preset structure focus threshold, marks the spatial distribution as a structure focus area, and maps and integrates the focus area position information and the number to generate a fault concentrated distribution mapping diagram.

8. The intelligent fault diagnostic system of an electrical control cabinet according to claim 7, characterized in that: In the region failure contact point distribution table, the number of corresponding failure trend connection contact points is extracted for each region number, and the total number of the numbers is counted; The process of filtering the region numbers whose number of failure trend contact points exceeds the preset structure focus threshold is: comparing the number of failure trend connection contact points corresponding to the region number with the set structure focus threshold 20, and only keeping the region numbers whose number of failure trend connection contact points is greater than 20; The process of mapping and integrating the focus area position information and the number is: Respectively extract the spatial positioning coordinate range of each region number filtered in the electrical control cabinet structure diagram, one-to-one correspondingly integrate according to the region number, and form the integrated result into the fault concentrated distribution mapping diagram.

9. The intelligent fault diagnostic system for electrical control cabinets of claim 1, wherein: The system further comprises a fault risk output module: The fault risk output module records and counts the number of risk elements in the region according to the fault concentrated distribution mapping diagram, combines the element thermal degradation state label and the contact point failure trend mark of the corresponding elements in the structure focus area, and triggers the diagnostic output logic when the region simultaneously meets the set conditions to generate a region fault diagnosis conclusion set; The region fault diagnosis conclusion set includes structure focus area identification results, element degradation risk levels, and contact point failure trend indicators.

10. The intelligent fault diagnostic system of an electrical control cabinet according to claim 9, characterized in that: The fault risk output module comprises: The risk element aggregation submodule calls the element number and connection contact point number information included in the region according to the region number of the structure focus area in the fault concentrated distribution mapping diagram, extracts the thermal degradation state label of the corresponding element and the contact point failure trend mark of the corresponding contact point, counts the number of risk elements according to the region dimension, and generates a focus area risk element statistical table; The diagnostic trigger judgment submodule calls the focus area risk element statistical table, and sequentially matches the risk element number value with the set thermal degradation element judgment value, the failure trend mark judgment value, and the combined threshold standard. If all three meet the set conditions, the region number is recorded as a region with diagnostic output conditions, and a diagnostic output trigger area number set is obtained; The conclusion construction submodule integrates the element number, contact point number, and associated thermal degradation state label and contact point failure trend mark content in the corresponding region according to the diagnostic output trigger area number set, assembles a structured record item, outputs the diagnostic item according to the region number, and generates a region fault diagnosis conclusion set.

Images