Methods and devices for early warning of fire risks of electrical cabinet components

Abstract

This application provides a method and device for early warning of fire risks in electrical cabinet components. The method includes: performing condition-adaptive filtering on the original monitoring dataset of electrical cabinet components from rail transit vehicles to obtain an effective monitoring dataset; the original monitoring dataset includes fire precursor parameters of the electrical cabinet components and operating condition parameters of the rail transit vehicles; performing fire risk correlation feature parameter screening on the fire risk-related feature parameters extracted from the effective monitoring dataset to obtain a set of fire risk key feature parameters; determining the target weight of each fire risk key feature parameter in the set of fire risk key feature parameters; and calculating the real-time fire risk value of the electrical cabinet components based on each fire risk key feature parameter and its corresponding target weight. This application solves the problems of existing early warning systems having a single monitoring dimension, delayed early warning, and weak anti-interference capability.

Description

Technical Field

[0001] This application relates to the field of fire early warning technology, and more specifically, to a method and device for early warning of fire risks of electrical cabinet components. Background Technology

[0002] As the core transportation tool of rail transit, the operational safety of rail transit vehicles is directly related to the safety of passengers' lives and property and the stability of transportation order. The electrical cabinet of a rail transit vehicle, as the core power distribution and control unit, integrates various electrical components such as circuit breakers, contactors, and power modules. It is a key piece of equipment ensuring the normal operation of the traction, braking, and control systems of rail transit vehicles. Operating continuously under extreme conditions of high load, vibration, high and low temperature cycles, and strong electromagnetic interference, the stability of its operational status is crucial to the overall safety of the rail transit vehicle. A fire in any component of the electrical cabinet will directly lead to the shutdown or malfunction of the rail transit vehicle, and may even cause a major safety accident.

[0003] Currently, fire risk warning for electrical cabinet components in rail transit vehicles mainly relies on traditional early warning systems. These systems use a single type of sensor as the core monitoring component, mostly employing only temperature or smoke sensors. By monitoring temperature changes or smoke generation inside the electrical cabinet, they can achieve preliminary early warning of fire risks. Some systems will simply collect a small number of operating parameters to assist in judging the operating status of the electrical cabinet. The overall early warning logic is mainly based on monitoring a single parameter to trigger an alarm.

[0004] However, existing methods are insufficient to meet the actual needs of early warning of fire risks in electrical cabinet components of rail transit vehicles. First, the monitoring dimensions are limited, relying solely on temperature or smoke sensors, resulting in delayed early warning responses. Alarms are often triggered only when fire signs are obvious, missing the best opportunity for response. Second, the data integration capabilities are insufficient, and the data utilization efficiency is low, leading to a high false alarm rate and a high missed alarm rate in the early warning system. Third, the anti-interference capability is weak, making it susceptible to external interference that affects the accuracy of monitoring data, impacting the reliability of early warnings, and consequently affecting driving safety and maintenance efficiency. Summary of the Invention

[0005] In view of this, the purpose of this application is to provide a method and device for early warning of fire risks of electrical cabinet components. In a first aspect, embodiments of this application provide a method for early warning of fire risks of electrical cabinet components, the method comprising: The original monitoring dataset of electrical cabinet components of rail transit vehicles is subjected to working condition adaptation filtering to obtain an effective monitoring dataset. The original monitoring dataset includes fire precursor parameters of electrical cabinet components and operating condition parameters of rail transit vehicles. The fire risk correlation feature parameters extracted from the effective monitoring dataset are subjected to fire risk key feature screening processing to obtain a set of fire risk key feature parameters; Determine the target weight corresponding to each key fire risk feature parameter in the set of key fire risk feature parameters; The real-time fire risk value of electrical cabinet components is calculated based on each key fire risk characteristic parameter and the target weight corresponding to each key fire risk characteristic parameter.

[0006] Secondly, this application also provides a fire risk early warning device for electrical cabinet components, the device comprising: The data processing module is used to perform condition-adaptive filtering on the raw monitoring dataset of electrical cabinet components of rail transit vehicles to obtain an effective monitoring dataset. The raw monitoring dataset includes fire precursor parameters of electrical cabinet components and operating condition parameters of rail transit vehicles. The feature filtering module is used to perform fire risk key feature filtering on the fire risk-related feature parameters extracted from the effective monitoring dataset to obtain a set of fire risk key feature parameters. The weight determination module is used to determine the target weight corresponding to each fire risk key feature parameter in the fire risk key feature parameter set. The risk value calculation module is used to calculate the real-time fire risk value of electrical cabinet components based on each key fire risk characteristic parameter and the target weight corresponding to each key fire risk characteristic parameter.

[0007] Thirdly, embodiments of this application also provide an electronic device, including: a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, the steps of the electrical cabinet component fire risk warning method described above are performed.

[0008] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the above-described electrical cabinet component fire risk warning method.

[0009] The fire risk early warning method and device for electrical cabinet components provided in this application can solve the technical problems of existing early warning systems, such as single monitoring dimensions, delayed early warning, and weak anti-interference ability. It improves the accuracy of early warning through multi-source data acquisition and working condition adaptation filtering, and improves the accuracy of risk assessment through key feature screening and weight allocation. It can achieve early identification of fire precursors, provide differentiated handling solutions for operation and maintenance personnel, and balance driving safety and operation and maintenance economy.

[0010] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0011] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is one of the flowcharts for a fire risk warning method for electrical cabinet components provided in the embodiments of this application; Figure 2 A second flowchart illustrating a fire risk warning method for electrical cabinet components provided in this application embodiment; Figure 3 A flowchart of a fire risk warning method for electrical cabinet components provided in this application embodiment; Figure 4 Flowchart four of a fire risk warning method for electrical cabinet components provided in this application embodiment; Figure 5 The fifth flowchart of a fire risk warning method for electrical cabinet components provided in the embodiments of this application; Figure 6 This is a schematic diagram of the structure of an electrical cabinet component fire risk early warning device provided in an embodiment of this application; Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0013] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. Based on the embodiments of this application, every other embodiment obtained by those skilled in the art without inventive effort falls within the scope of protection of this application.

[0014] Please see Figure 1 , Figure 1This is a flowchart illustrating a fire risk warning method for electrical cabinet components provided in an embodiment of this application. Figure 1 As shown in the embodiments of this application, the method includes: Step S1: Perform condition adaptation filtering on the collected raw monitoring dataset of electrical cabinet components of rail transit vehicles to obtain an effective monitoring dataset. The raw monitoring dataset includes fire precursor parameters of electrical cabinet components and operating condition parameters of rail transit vehicles. Step S2: Perform fire risk key feature screening on the fire risk correlation feature parameters extracted from the effective monitoring dataset to obtain a set of fire risk key feature parameters; Step S3: Determine the target weight corresponding to each key fire risk characteristic parameter in the set of key fire risk characteristic parameters; Step S4: Calculate the real-time fire risk value of electrical cabinet components based on each key fire risk characteristic parameter and the target weight corresponding to each key fire risk characteristic parameter.

[0015] The method provided in this embodiment can solve the technical problems of existing early warning systems, such as single monitoring dimensions, delayed early warning, and weak anti-interference ability. It improves the accuracy of early warning by acquiring multi-source data and adapting to operating conditions and filtering, and improves the accuracy of risk assessment by filtering key features and assigning weights. It enables early identification of fire precursors, provides differentiated handling solutions for operation and maintenance personnel, and balances driving safety and operation and maintenance economy.

[0016] The following is a detailed explanation of steps S1 to S4 above: In step S1, the original monitoring dataset of the electrical cabinet components of the rail transit vehicle is subjected to operating condition adaptation filtering to obtain an effective monitoring dataset. The original monitoring dataset includes fire precursor parameters of the electrical cabinet components and operating condition parameters of the rail transit vehicle.

[0017] Here, the raw monitoring dataset refers to the basic data set used to reflect the fire precursors and operating environment of electrical cabinet components in rail transit vehicles. Specifically, the raw monitoring dataset can be formed by synchronously processing fire precursor-related parameters collected by multiple types of sensors and operating condition parameters synchronously acquired by the on-board monitoring system of rail transit vehicles. This provides comprehensive and accurate basic data for fire risk early warning, ensuring the reliability of subsequent processing and assessment, and avoiding early warning deviations due to missing or isolated data.

[0018] Among them, fire precursor parameters are various physical parameters that can reflect the abnormal state of electrical cabinet components before a fire occurs. For example, fire precursor parameters may include component surface temperature parameters, coil high-frequency current parameters, smoke concentration parameters inside the electrical cabinet, and component insulation resistance parameters. These parameters correspond to different fire precursors such as electrical component aging, poor contact, and insulation damage, and can capture abnormal changes before a fire occurs from multiple dimensions, making up for the limitations of single-parameter monitoring in existing technologies.

[0019] Among them, the operating condition parameters of rail transit vehicles are parameters that reflect the real-time operating status of rail transit vehicles. They are used to adapt parameter processing under different operating scenarios and improve the filtering effect. For example, the operating condition parameters of rail transit vehicles may include operating speed, load current, ambient temperature, and operating time. These parameters can reflect the operating load, environmental conditions, and other operating condition information of rail transit vehicles. The operating status and interference parameters of electrical cabinet components differ under different operating conditions. Combining these parameters for filtering can effectively eliminate noise interference caused by operating condition fluctuations and improve the data purification effect.

[0020] In the above steps, differentiated filtering methods can be used to eliminate noise interference for different types of raw parameters and different operating conditions, thereby improving the effectiveness of monitoring data. Specifically, the operating condition-adaptive filtering process can dynamically adjust the filtering parameters according to the parameter type and operating conditions. This includes moving average filtering for surface temperature parameters, wavelet threshold filtering for coil high-frequency current parameters, median filtering for smoke concentration and insulation resistance parameters, and subsequent normalization processing. This can eliminate noise caused by vibration of rail transit vehicles, electromagnetic interference, and fluctuations in operating conditions, purifying the raw monitoring data into effective data that truly reflects the status of electrical cabinet components.

[0021] The effective monitoring dataset is a collection of purification data obtained by adapting the original monitoring data to the operating conditions. It is the core data used to extract feature parameters and conduct risk assessment. Specifically, the effective monitoring dataset is obtained by normalizing parameters such as filtered surface temperature, coil high-frequency current, smoke concentration, and insulation resistance. The data values ​​can be mapped to the [0, 1] interval, which not only retains the feature information of the original parameters but also eliminates the influence of differences in the dimensions of different parameters. This facilitates subsequent feature extraction and weight calculation, ensuring the accuracy and rationality of the risk assessment.

[0022] In one alternative embodiment, such as Figure 2 As shown, the raw monitoring dataset can be collected through the following steps: Step S11: Collect fire precursor parameters by using various sensors placed at the electrical cabinet components of the rail transit vehicle, and simultaneously collect the operating condition parameters of the rail transit vehicle through the on-board monitoring system of the rail transit vehicle; the fire precursor parameters include the surface temperature parameters of the components, the high-frequency current parameters of the coils, the smoke concentration parameters inside the electrical cabinet, and the insulation resistance parameters of the components; the operating condition parameters include the operating speed, load current, ambient temperature, and operating time. Here, various types of sensors can include temperature sensors, high-frequency current sensors, smoke sensors, and insulation resistance sensors. These can be deployed in the core areas of components (including contactors, relays, terminals, and fuses) and inside the electrical cabinet cavity to ensure accurate collection of fire precursor parameters for each component.

[0023] Among them, the collected fire precursor parameters can specifically include the surface temperature parameters of components (reflecting the heating state of components), the high-frequency current parameters of coils (reflecting abnormal electrical performance of components), the smoke concentration parameters inside the electrical cabinet (reflecting the initial signs of a fire), and the insulation resistance parameters of components (reflecting the insulation state of components).

[0024] Step S12: Synchronously process the fire precursor-related parameters and operating condition parameters to obtain the original monitoring dataset of the electrical cabinet components of the rail transit vehicle. Each data entry in the original monitoring dataset includes the collection timestamp, sensor number, fire precursor-related parameters, and corresponding operating condition parameters.

[0025] Here, the operating condition parameters of rail transit vehicles can be synchronously acquired through the on-board monitoring system of rail transit vehicles to ensure the time synchronization with parameters related to fire precursors. After synchronization processing, the data can be stored at a preset sampling frequency of 1~10Hz to form a complete original monitoring dataset. Each data entry includes a collection timestamp, sensor number, fire precursor related parameters and corresponding operating condition parameters, which facilitates subsequent traceability and operating condition adaptation processing.

[0026] In one alternative embodiment, such as Figure 3 As shown, step S1 specifically includes the following steps: Step S13: Perform a moving average filtering process on the surface temperature parameters in each data entry to obtain the surface temperature purification parameters.

[0027] Here, for the surface temperature parameter, a moving average filtering method is used to eliminate random noise. The size of the filtering window can be dynamically adjusted according to the operating conditions of the rail transit vehicle. For example, when the load current is ≥ 80% of the rated current, the window size can be set to 5 to 8 sampling points; when the load current is < 80% of the rated current, the window size can be set to 3 to 5 sampling points, ensuring that noise can be effectively eliminated and temperature change characteristics can be preserved under different load conditions.

[0028] Step S14: Perform wavelet threshold filtering on the high-frequency current parameters of the coil in each data entry to obtain the high-frequency current purification parameters of the coil.

[0029] Here, wavelet threshold filtering can be used to eliminate electromagnetic interference for the high-frequency current parameters of the coil. For example, a db4 wavelet basis can be selected, the number of decomposition layers can be set to 3 to 5, and interference components can be eliminated by setting a preset threshold, retaining the characteristic frequency components of 10kHz to 1MHz that reflect abnormal heating of the components, which is suitable for the strong electromagnetic interference operating conditions of rail transit vehicles.

[0030] Step S15: Perform median filtering on the smoke concentration parameters and component insulation resistance parameters in each data entry to obtain the smoke concentration purification parameters and component insulation resistance purification parameters in the electrical cabinet.

[0031] Here, median filtering is used to eliminate transient interference caused by vibration, targeting the smoke concentration and component insulation resistance parameters inside the electrical cabinet. For example, the filtering window can be set to three sampling points to remove abnormal data exceeding reasonable ranges. The reasonable range for smoke concentration inside the electrical cabinet can be set to 0~100ppm, and the reasonable range for component insulation resistance can be set to 1MΩ~100MΩ, preventing abnormal data from affecting subsequent processing.

[0032] Step S16: Normalize the surface temperature purification parameters, coil high-frequency current purification parameters, smoke concentration purification parameters in the electrical cabinet, and component insulation resistance purification parameters to obtain an effective monitoring dataset.

[0033] Here, the normalization process can be performed using the following formula:

[0034] Where, x norm The normalized effective monitoring data is represented by x, which is the original parameter value after filtering. min x is the minimum reasonable value for this parameter. max This is the maximum reasonable value for this parameter.

[0035] By normalizing all parameter values ​​to the [0, 1] interval, the dimensional differences can be eliminated, and the purified and effective monitoring data can be obtained.

[0036] In step S2, the fire risk-related feature parameters extracted from the effective monitoring dataset are subjected to fire risk key feature screening processing to obtain a set of fire risk key feature parameters.

[0037] In this step, the fire risk-related feature parameters are various feature parameters extracted from the effective monitoring dataset that are related to the fire risk of electrical cabinet components, and are used to reflect the fire risk status. Optionally, the fire risk-related feature parameters can include static feature parameters and dynamic feature parameters. Static feature parameters can reflect the steady-state operation of the component, while dynamic feature parameters can reflect the changing trend of the component's operating status. The combination of the two can comprehensively capture the static manifestations and dynamic changes of the component's fire risk, providing a comprehensive feature basis for the selection of key features.

[0038] Specifically, the screening and processing of key fire risk features can be a process of selecting characteristic parameters that significantly affect the fire risk status using specific methods. The aim is to eliminate irrelevant or less influential features, thereby improving the efficiency and accuracy of subsequent risk assessments. Optionally, the screening and processing of key fire risk features can employ sensitivity analysis. By calculating the sensitivity of the fire risk status to each associated characteristic parameter, characteristic parameters that meet the sensitivity criteria are selected as key features. This reduces redundant features and computational complexity, while ensuring that the selected key features accurately reflect the fire risk status, avoiding risk assessment bias due to feature redundancy or omissions.

[0039] Here, the set of key fire risk characteristic parameters is a collection of characteristic parameters that have a significant impact on the fire risk status, obtained after screening. It serves as the input for subsequent risk assessment model construction and risk value calculation. Optionally, the parameters in the set of key fire risk characteristic parameters are all features with high sensitivity to the fire risk status, capable of accurately capturing key changes in fire precursors, ensuring the accuracy and relevance of the risk assessment.

[0040] In one optional embodiment, the fire risk associated characteristic parameters include static characteristic parameters and dynamic characteristic parameters. The static characteristic parameters include: the average value of the surface temperature parameter, the maximum value of the surface temperature parameter, the peak value of the surface temperature parameter, the effective value of the coil high-frequency current parameter, the peak factor of the coil high-frequency current parameter, the average value of the smoke concentration parameter inside the electrical cabinet, the minimum value of the component insulation resistance parameter, and the stable value of the component insulation resistance parameter.

[0041] Specifically, the average value of the surface temperature parameter reflects the average heating level of the component over a period of time, the maximum value reflects the highest heating degree of the component, and the peak value reflects the instantaneous highest temperature of the component. The combination of the three can comprehensively reflect the steady-state heating state of the component. The effective value of the coil high-frequency current parameter reflects the average work capacity of the current, and the peak factor reflects the degree of distortion of the current waveform, which can reflect the steady-state anomalies of the component's electrical performance. The average value of the smoke concentration parameter inside the electrical cabinet reflects the average accumulation degree of smoke inside the cabinet, which is an important steady-state characteristic in the early stage of a fire. The minimum value of the component insulation resistance parameter reflects the lowest level of the component's insulation performance, and the stable value reflects the steady-state state of the component's insulation performance, which can reflect the steady-state degree of insulation aging of the component.

[0042] In addition, the dynamic characteristic parameters include: the rate of increase of surface temperature parameter, the difference between surface temperature and ambient temperature, the harmonic distortion rate of coil high-frequency current parameter, the fluctuation amplitude of coil high-frequency current parameter, the rate of increase of smoke concentration parameter in electrical cabinet, and the rate of decrease of component insulation resistance parameter.

[0043] Specifically, the rate of increase of surface temperature parameters reflects the speed of component temperature change and can capture the trend of abnormal heating of components. The difference between surface temperature and ambient temperature reflects the degree of abnormal heating of components relative to the environment. The combination of the two can accurately capture the dynamic anomalies of component heating. The harmonic distortion rate of coil high-frequency current parameters reflects the degree of distortion of current waveform, and the fluctuation amplitude reflects the degree of instability of current and can reflect the dynamic anomalies of component electrical performance. The rate of increase of smoke concentration parameters in electrical cabinets reflects the speed of smoke accumulation and is an important characteristic of dynamic changes in the early stage of a fire. The rate of decrease of component insulation resistance parameters reflects the rate of deterioration of component insulation performance and can capture the dynamic trend of insulation aging.

[0044] Optionally, such as Figure 4 As shown, step S2 specifically includes the following steps: Step S21: Establish a mapping function between fire risk status and fire risk-related characteristic parameters. Fire risk status includes no risk, low risk, medium risk and high risk. Step S22: Calculate the sensitivity of fire risk status to each fire risk-related characteristic parameter based on the mapping relationship function; Step S23: Collect fire risk-related feature parameters with a sensitivity not lower than the preset sensitivity threshold into a set of key fire risk feature parameters.

[0045] The specific processes of steps S21 to S23 above are as follows: The response quantity Y is defined as four fire risk states of electrical cabinet components: no risk, low risk, medium risk, and high risk; and the static and dynamic characteristic parameters are defined as influencing factors X, where X = (x1, x2, ..., x...). i ,…,xn Let represent the parameter vector composed of all fire risk-related characteristic parameters. Establish a mapping function F(X) between fire risk states and fire risk-related characteristic parameters. This function characterizes the correspondence between each fire risk-related characteristic parameter and the fire risk state. Based on this mapping function, calculate Y with respect to x. i Sensitivity S i The calculation formula is as follows:

[0046] In the formula, F(X) is the mapping function between fire risk status and fire risk-related characteristic parameters, x i S represents the fire risk-related characteristic parameter i. i This represents the sensitivity corresponding to the i-th fire risk-related feature parameter; the preset sensitivity threshold can be set to 0.3. If the sensitivity |S i If |≥0.3, then the fire risk-related characteristic parameter x is determined. i It has a significant impact on fire risk status and is included in the set of key characteristic parameters of fire risk; if |S i If |<0.3, then the fire risk-related characteristic parameter is eliminated, and the final set of key fire risk characteristic parameters is formed.

[0047] This screening method can accurately retain characteristic parameters that have a significant impact on fire risk, eliminate redundant features, and improve the efficiency and accuracy of subsequent risk assessments.

[0048] In one optional embodiment, the method provided in this application further includes: A fire risk assessment model is constructed to determine the target weights corresponding to each key fire risk characteristic parameter in the set of key fire risk characteristic parameters. The fire risk assessment model includes a target layer, a criterion layer, and an indicator layer. The target layer is used to characterize the real-time fire risk value. The criterion layer includes surface temperature characteristics, coil high-frequency current characteristics, cabinet smoke concentration characteristics, and component insulation resistance characteristics. The indicator layer corresponds to each key characteristic parameter of fire risk.

[0049] The fire risk assessment model is used to calculate fire risk values ​​and classify early warning levels. It quantifies the degree of fire risk by assigning key fire risk characteristic parameters and their weights. Optionally, the fire risk assessment model can be constructed based on the fuzzy hierarchical analysis method, adopting a hierarchical structure design divided into a target layer, a criterion layer, and an indicator layer. This enables reasonable weight allocation and risk quantification of multi-dimensional key fire risk characteristic parameters, adapting to the multi-factor and multi-dimensional assessment needs of fire risks in electrical cabinet components of rail transit vehicles, and improving the rationality and accuracy of risk assessment.

[0050] Here, the target layer is the fire risk assessment result of the electrical cabinet components of rail transit vehicles (i.e., real-time fire risk value), the criterion layer is the surface temperature characteristics, coil high-frequency current characteristics, smoke concentration characteristics inside the electrical cabinet, and component insulation resistance characteristics (corresponding to the four categories of key fire risk characteristic parameters respectively), and the indicator layer is the key fire risk characteristic parameters after screening in step S2 (i.e., the specific key fire risk characteristic parameters under each criterion layer). This hierarchical structure model can clearly define the hierarchical relationship of each characteristic parameter and provide a clear framework for weight allocation.

[0051] Then, in step S3, the target weights corresponding to each fire risk key characteristic parameter in the fire risk key characteristic parameter set are determined.

[0052] The target weight represents the proportion of each key fire risk characteristic parameter in the risk assessment, reflecting the degree of influence of different key fire risk characteristics on fire risk. Specifically, the target weight can be calculated using fuzzy hierarchical analysis, combined with expert experience and mathematical methods to ensure the rationality and scientific nature of the weight allocation. This ensures that key characteristic parameters with a greater impact on fire risk have higher weights, improving the accuracy of risk assessment and avoiding risk assessment bias caused by unreasonable weight allocation.

[0053] In one alternative embodiment, such as Figure 5 As shown, step S3 specifically includes the following steps: Step S31: Construct the first fuzzy complementary judgment matrix of the criterion layer relative to the target layer based on the criterion layer judgment results, and construct the second fuzzy complementary judgment matrix of each index layer relative to its respective criterion layer based on the index layer judgment results; the criterion layer judgment results represent the interval scores obtained by pairwise comparison of surface temperature characteristics, coil high-frequency current characteristics, cabinet smoke concentration characteristics, and component insulation resistance characteristics, and the index layer judgment results represent the interval scores obtained by pairwise comparison of key fire risk characteristic parameters under the same criterion.

[0054] Here, we can invite experts in the field of rail transit electrical safety to score the relative importance of indicators at each level using a scaling method of 0.1 to 0.9, and construct the first fuzzy complementary judgment matrix R of the criterion layer to the target layer. a And the second fuzzy complementary judgment matrix R of each index layer to its corresponding criterion layer. b (Each criterion layer corresponds to a second fuzzy complementary judgment matrix); the core of the 0.1~0.9 scaling method is to determine the relative importance of indicators through pairwise comparisons. The first fuzzy complementary judgment matrix R in this application... a And the second fuzzy complementary judgment matrix R b The expression is as follows:

[0055]

[0056] In the formula, r ii =0.5 indicates that the indicators are equally important. ij >0.5 indicates that the index r j R i Important, r ij <0.5 indicates that the index r i R j Important, and the matrix elements satisfy the fuzzy complementarity relation r. ij +r ji =1, where n is the number of indicators at the corresponding level; the evaluation results of the criterion layer are the interval scores obtained by experts through pairwise comparison of surface temperature characteristics, coil high-frequency current characteristics, cabinet smoke concentration characteristics, and component insulation resistance characteristics, which are used to construct the first fuzzy complementary judgment matrix; the evaluation results of the indicator layer are the interval scores obtained by experts through pairwise comparison of each key feature parameter under the same criterion, which are used to construct the second fuzzy complementary judgment matrix corresponding to each criterion layer.

[0057] Step S32: Solve for the criterion layer weight vector based on the first fuzzy complementary judgment matrix and solve for the weight vector of each index layer based on the second fuzzy complementary judgment matrix. Construct the criterion layer feature matrix from the criterion layer weight vector and construct the index layer feature matrix from the weight vector of each index layer.

[0058] Here, the criterion layer includes four fixed evaluation indicators: surface temperature characteristics, coil high-frequency current characteristics, cabinet smoke concentration, and component insulation resistance characteristics. The corresponding number of indicators is set to n=4. The weight vector W of the criterion layer is calculated using a general formula. a =(W a1 W a2 W a3 W a4 ), where W a1 For surface temperature characteristic weights, W a2 For the characteristic weights of the high-frequency current of the coil, W a3 Weights of smoke concentration characteristics inside the cabinet, W a4 The insulation resistance characteristic weight of the component is used. Based on the number of key fire risk characteristic parameters within each indicator layer, the total number of corresponding indicators n is determined. The weight of each individual indicator is calculated and then combined to obtain the weight vector W for each indicator layer. b W b =(W b1 W b2 ,…,W bn This refers to all individual weights W obtained sequentially under the same criterion. b1 W b2 ,…,W bnBy integrating them according to the preset arrangement order, the indicator layer weight vector W specific to this criterion can be formed. b For example, the criterion layer weight vector W a Single item weight and the weight vector W of each indicator layer b Single item weight The general calculation formula is as follows:

[0059]

[0060] In the formula, k is the index of the current single indicator within the corresponding level, n is the total number of indicators in the corresponding level, and r kj This corresponds to the element in the k-th row and j-th column of the fuzzy complementary judgment matrix. Subsequently, the criterion layer weight vector... Construct the characteristic matrix W of the criterion layer a *, determined by the weight vector W of each indicator layer b Construct the feature matrix W of the index layer b *, Criterion layer feature matrix W a *and the indicator layer feature matrix W b The formula for calculating * is as follows:

[0061]

[0062] Step S33: Compare the first fuzzy complementary judgment matrix with the feature matrix of the criterion layer, and compare the second fuzzy complementary judgment matrix with the feature matrix of the index layer to determine whether the weight allocation results of each level are valid.

[0063] Step S33 above specifically includes the following steps: Step S331: Calculate the first compatibility index between the first fuzzy complementary judgment matrix and the feature matrix of the criterion layer. If the value of the first compatibility index is less than the first preset index threshold, the weight allocation result of the constructed first fuzzy complementary judgment matrix is ​​determined to be valid. Step S332: Calculate the second compatibility index between the second fuzzy complementary judgment matrix and the indicator layer feature matrix. If the value of the second compatibility index is less than the second preset index threshold, the weight allocation result of the constructed second fuzzy complementary judgment matrix is ​​determined to be valid.

[0064] For example, calculate the first fuzzy complementary judgment matrix R. a Its characteristic matrix W a *The first compatibility index I(R) a W a *), and the second fuzzy complementary judgment matrix Rb Its corresponding characteristic matrix W b *The second compatibility index I(R) b W b *), First compatibility index I(R) a W a *) and the second compatibility index I(R) b W b The general calculation formula for *) is as follows:

[0065]

[0066] here, This represents the element in the i-th row and j-th column of the first fuzzy complementary judgment matrix. This represents the element in the i-th row and j-th column of the second fuzzy complementary judgment matrix. This represents the feature components at corresponding positions within the feature matrix of the criterion layer. , This represents the feature component at the corresponding position within the feature matrix of the indicator layer. Both the first and second preset indicator thresholds are set to 0.1. If I(R) a W a If *) < 0.1, then the first fuzzy complementary judgment matrix is ​​determined to be a satisfactory consistency matrix, and the weight allocation of the criterion layer is reliable; if I(R) < 0.1, then the first fuzzy complementary judgment matrix is ​​determined to be a satisfactory consistency matrix, and the weight allocation of the criterion layer is reliable; b W b If *) < 0.1, the second fuzzy complementary judgment matrix of the corresponding criterion layer is determined to be a satisfactory consistency matrix, and the weight allocation of the index layer under this criterion layer is reliable. If any compatibility index is ≥ 0.1, the consistency of the judgment matrix at this level is determined to be unsatisfactory. Experts in the field of rail transit electrical safety need to be invited again to adjust the scoring results of the relative importance of the indicators, reconstruct the fuzzy complementary judgment matrix, and repeat the weight solution, feature matrix construction, and compatibility test process until all level compatibility indicators meet the threshold requirements. This ensures the rationality and scientific nature of the weight allocation results of the feature parameters at each level throughout the process.

[0067] Step S34: After all levels have passed the consistency test, the weight vector of the criterion layer and the weight vector of the corresponding indicator layer are weighted and fused together in combination with the preset hierarchical weight constraint relationship to obtain the target weight corresponding to each fire risk key characteristic parameter.

[0068] The preset hierarchical weight constraint relationship means that the sum of the weights corresponding to the surface temperature characteristics, coil high-frequency current characteristics, cabinet smoke concentration characteristics, and component insulation resistance characteristics of the criterion layer is 1, and the cumulative weight of the key fire risk characteristic parameters corresponding to all index layers under each criterion layer is consistent with the weight of the corresponding criterion layer.

[0069] Combined with the preset hierarchical weight constraints, namely the surface temperature characteristic weight W in the criterion layer t Coil current characteristic weight W c Weight of smoke characteristics inside electrical cabinet W s Component insulation characteristic weight W r Satisfy W t +W c +W s +W r =1, the sum of the weights of each key feature parameter in the index layer is equal to the weight of the corresponding criterion layer (i.e., the sum of the weights of the key parameters under the surface temperature feature is W). t The weights of the key parameters under the coil current characteristics are W. c The weights of the key parameters under the smoke characteristics inside the electrical cabinet are W. s The weights of the key parameters under the insulation characteristics of the component are W. r The weight vector of the criterion layer and the weight vector of the corresponding index layer are weighted and fused to obtain the target weight of each key fire risk feature parameter. That is, the target weight of each key feature parameter of the index layer = the weight of the corresponding criterion layer × the weight of the index layer parameter in its criterion layer. Finally, the target weight of all key feature parameters is obtained, ensuring that the weight allocation conforms to expert experience and meets the requirements of mathematical consistency, and can accurately reflect the degree of influence of each key feature on fire risk.

[0070] In step S4, the real-time fire risk value of the electrical cabinet components is calculated based on each fire risk key characteristic parameter and the target weight corresponding to each fire risk key characteristic parameter.

[0071] The real-time fire risk value is a numerical value used to quantify the current fire risk level of electrical cabinet components in rail transit vehicles. It is the core basis for classifying warning levels and generating response recommendations. Optionally, the real-time fire risk value can be calculated using a weighted summation method. By combining the normalized values ​​of each key fire risk characteristic parameter with the target weight, the degree of fire risk can be accurately quantified. The higher the value, the higher the fire risk. Its function is to transform multi-dimensional key fire risk characteristic parameters into a single quantitative indicator, which facilitates the subsequent classification of warning levels and the generation of response recommendations.

[0072] In one optional embodiment, the real-time fire risk value of the electrical cabinet components in a rail transit vehicle can be calculated using the following formula:

[0073] Where H represents the real-time fire risk value corresponding to the electrical cabinet components of the rail transit vehicle, m is the total number of key fire risk characteristic parameters obtained after sensitivity analysis screening, and W iThe final target weight, x, is obtained by hierarchical weight fusion calculation of the i-th key fire risk feature parameter. i,norm The effective monitoring data obtained after normalizing the i-th key characteristic parameter of fire risk.

[0074] Specifically, constrained by both the normalized data value range and the weight normalization constraint, the real-time fire risk value H is fixed in the range of [0,1]. The larger the value of H, the higher the risk of fire hazards and fire accidents in the corresponding rail transit vehicle electrical cabinet components. To ensure the real-time and timely nature of the fire risk assessment results, a dynamic update cycle of 1-3 minutes is set. The latest valid monitoring data, after being processed by operating condition adaptation filtering, feature parameter extraction, and key feature screening, is periodically retrieved and iteratively calculated using the above formula to obtain the real-time fire risk value H. This dynamic update of the fire risk value can accurately capture the dynamic evolution trend of fire risk in rail transit vehicle electrical cabinet components in real time, providing accurate and reliable quantitative data support for the subsequent automatic determination of fire risk warning levels and real-time adjustment of differentiated operation and maintenance suggestions.

[0075] For example, if the normalized values ​​of key characteristic parameters and their corresponding target weights at a certain moment are as follows: the normalized value of the maximum surface temperature is 0.3, and the weight is 0.2; the normalized value of the effective value of the coil high-frequency current is 0.2, and the weight is 0.15; the normalized value of the average smoke concentration is 0.1, and the weight is 0.25; and the normalized value of the minimum insulation resistance is 0.2, and the weight is 0.4, then the real-time fire risk value H = 0.3 × 0.2 + 0.2 × 0.15 + 0.1 × 0.25 + 0.2 × 0.4 = 0.195, corresponding to the subsequent Level 1 warning state.

[0076] Furthermore, the method provided in this application embodiment also includes: Step S5: Based on the real-time fire risk value, generate a fire risk warning level and corresponding handling suggestions for each warning level, and dynamically update the warning level and corresponding handling suggestions based on the real-time fire risk value.

[0077] The warning level is a fire risk level classification based on real-time fire risk values. It is used to intuitively reflect the fire risk status of electrical cabinet components and provide clear risk alerts for maintenance personnel. Optionally, the warning level can be divided into four levels, from low to high, corresponding to no risk, low risk, medium risk, and high risk, respectively. This enables graded management of fire risks, avoids a one-size-fits-all approach, and provides maintenance personnel with a clear basis for risk classification, facilitating differentiated handling measures based on different risk levels.

[0078] The response recommendations are specific measures developed for different warning levels to prevent and control fire risks. Based on the risk level, differentiated and targeted response plans are formulated to balance driving safety and operational efficiency. Optionally, the response recommendations correspond one-to-one with the warning level; the higher the risk level, the more urgent the response action. This ensures that unnecessary operational intervention is reduced when the fire risk is low, and that emergency measures are taken promptly when the fire risk is high to prevent the fire from occurring or escalating, while simultaneously reducing operational costs.

[0079] In the above steps, the warning level and corresponding handling suggestions are adjusted in real time according to the changes in the real-time fire risk value to ensure the timeliness and accuracy of the warning information and handling suggestions, thereby adapting to the dynamic changes in the fire risk of electrical cabinet components.

[0080] In addition, when the real-time fire risk value exceeds the warning level threshold, the warning level and corresponding handling suggestions are automatically updated and simultaneously pushed to the vehicle-mounted maintenance personnel and the ground monitoring center to ensure that maintenance personnel can keep abreast of the risk changes and take corresponding measures to avoid the expansion of fire risk due to the delay in warning.

[0081] In one optional embodiment, step S5 specifically includes the following steps: Step S51: Based on the real-time fire risk value and the range division corresponding to each warning level, generate the fire risk warning level and the corresponding handling suggestions for each warning level; among which, the warning levels include Level 1 warning, Level 2 warning, Level 3 warning and Level 4 warning. The higher the warning level, the more urgent the corresponding handling action.

[0082] Specifically, the rules for classifying warning levels and corresponding handling recommendations are as follows: Level 1 Warning (No Risk): When 0≤H<0.2, it indicates that the electrical cabinet components are operating normally and there is no fire risk. The corresponding handling suggestion is to maintain the normal monitoring frequency (such as the preset sampling frequency 1~10Hz), without any additional handling, only to continuously monitor data changes to ensure the normal operation of the components.

[0083] Level 2 Warning (Low Risk): When 0.2≤H<0.4, it indicates that there is a minor abnormality in the electrical cabinet components and the risk of fire is low. The corresponding handling suggestion is to increase the monitoring frequency to 5~8Hz, continuously track changes in abnormal parameters, and remind maintenance personnel to focus on checking the relevant components during the next maintenance, promptly identify minor abnormalities, and avoid escalation of risks.

[0084] Level 3 Warning (Medium Risk): When 0.4≤H<0.7, it indicates that the electrical cabinet components are obviously abnormal and there is a moderate fire risk. The corresponding handling recommendation is to immediately notify the vehicle maintenance personnel, start local cooling measures (such as turning on the cooling fan in the electrical cabinet and turning off non-essential heat-generating components), suspend non-essential loads, reduce component loads, monitor abnormal components in real time, and make emergency preparations. If the risk continues to escalate, take further handling measures in a timely manner.

[0085] Level 4 Warning (High Risk): When H≥0.7, it indicates that there is a serious abnormality in the electrical cabinet components, the fire risk is extremely high, and a fire may be about to occur. The corresponding handling recommendation is to immediately trigger the vehicle-mounted audible and visual alarm, and at the same time send an emergency warning message to the ground monitoring center, prompting the operation and maintenance personnel to shut down the machine for inspection and activate the fire fighting plan, quickly investigate fire hazards, prevent the fire from occurring or spreading, and ensure driving safety.

[0086] Furthermore, when the real-time fire risk value H crosses the warning level threshold (e.g., from 0.38 to 0.42, crossing the threshold of 0.4 between Level 2 and Level 3 warnings), the system automatically updates the warning level and corresponding handling suggestions, and simultaneously pushes them to the driver's cab display terminal and the operation and maintenance platform. This ensures that operation and maintenance personnel can promptly grasp the changes in risk and take corresponding handling measures. At the same time, with the dynamic updating of monitoring data, the real-time fire risk value is recalculated every 1 to 3 minutes. If the risk value falls back to the lower warning level threshold, the warning level and corresponding handling suggestions will also be automatically lowered to ensure the timeliness and accuracy of the warning and avoid over-handling or untimely handling.

[0087] This application embodiment collects multi-dimensional fire precursor parameters through multiple types of sensors, combines them with rail transit vehicle operating condition parameters to form an original monitoring dataset, purifies the data through operating condition adaptation filtering, extracts static and dynamic feature parameters and filters key features, determines the weights of key features based on fuzzy hierarchical analysis, calculates real-time fire risk values ​​through weighted summation, and finally divides the four-level early warning levels and generates differentiated handling suggestions to achieve dynamic early warning.

[0088] Since existing single-sensor early warning methods can only issue alarms after a fire has occurred, this application can issue early warnings more than 15 minutes in advance, accurately capturing early signs of fire. Through multi-source parameter fusion and operating condition-adaptive filtering design, the early warning accuracy is improved to over 95%, reducing false alarm and missed alarm rates. A four-level graded early warning mechanism clarifies the priority of handling different risks, avoiding "one-size-fits-all" shutdowns and reducing operation and maintenance costs by more than 30%. This solution does not require large-scale modification of the existing electrical cabinet structure; it can be directly integrated with the TCMS system of rail transit vehicles, has strong compatibility, and is suitable for electrical cabinet monitoring scenarios of various high-speed rail transit vehicles and intercity rail transit vehicles, making it easy to use.

[0089] Based on the same inventive concept, this application also provides an electrical cabinet component fire risk warning device corresponding to the electrical cabinet component fire risk warning method. Since the principle of the device in this application is similar to the electrical cabinet component fire risk warning method described above in this application, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be described again.

[0090] Please see Figure 6 , Figure 6 This is a structural schematic diagram of an electrical cabinet component fire risk early warning device provided in an embodiment of this application. Figure 6 As shown, the device 600 includes: The data processing module 601 is used to perform condition-adaptive filtering on the original monitoring dataset of electrical cabinet components of rail transit vehicles to obtain an effective monitoring dataset. The original monitoring dataset includes fire precursor parameters of electrical cabinet components and operating condition parameters of rail transit vehicles. The feature filtering module 602 is used to perform fire risk key feature filtering on the fire risk related feature parameters extracted from the effective monitoring dataset to obtain a set of fire risk key feature parameters. The weight determination module 603 is used to determine the target weight corresponding to each fire risk key feature parameter in the fire risk key feature parameter set. The risk value calculation module 604 is used to calculate the real-time fire risk value of electrical cabinet components based on each fire risk key characteristic parameter and the target weight corresponding to each fire risk key characteristic parameter. Furthermore, the device 600 also includes: The risk warning module 605 is used to generate fire risk warning levels and corresponding handling suggestions for each warning level based on real-time fire risk values, and to dynamically update the warning levels and corresponding handling suggestions based on real-time fire risk values.

[0091] The device provided in this application collects multi-dimensional fire precursor parameters through multiple types of sensors, combines them with rail transit vehicle operating condition parameters to form an original monitoring dataset, purifies the data through operating condition adaptation filtering, extracts static and dynamic feature parameters and filters key features, determines the weights of key features based on fuzzy hierarchical analysis, calculates real-time fire risk values ​​through weighted summation, and finally divides four warning levels and generates differentiated handling suggestions to achieve dynamic early warning.

[0092] Please see Figure 7 , Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 7 As shown, the electronic device 700 includes a processor 710, a memory 720, and a bus 730.

[0093] The memory 720 stores machine-readable instructions executable by the processor 710. When the electronic device 700 is running, the processor 710 communicates with the memory 720 via the bus 730. When the machine-readable instructions are executed by the processor 710, they can perform the operations described above. Figures 1 to 5 The steps of the fire risk warning method for electrical cabinet components in the method embodiment shown are described in detail in the method embodiment, and will not be repeated here.

[0094] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, can perform the above-described actions. Figures 1 to 5 The steps of the fire risk warning method for electrical cabinet components in the method embodiment shown are described in detail in the method embodiment, and will not be repeated here.

[0095] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0096] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the shown or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0097] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0098] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0099] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0100] Finally, it should be noted that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The scope of protection of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this application. Such modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for early warning of fire risk in electrical cabinet components, characterized in that, include: The original monitoring dataset of electrical cabinet components of rail transit vehicles is subjected to working condition adaptation filtering to obtain an effective monitoring dataset. The original monitoring dataset includes fire precursor parameters of electrical cabinet components and operating condition parameters of rail transit vehicles. The fire risk correlation feature parameters extracted from the effective monitoring dataset are subjected to fire risk key feature screening processing to obtain a set of fire risk key feature parameters; Determine the target weight corresponding to each key fire risk feature parameter in the set of key fire risk feature parameters; The real-time fire risk value of electrical cabinet components is calculated based on each key fire risk characteristic parameter and the target weight corresponding to each key fire risk characteristic parameter.

2. The method according to claim 1, characterized in that, The following steps were used to collect the raw monitoring dataset of electrical cabinet components in rail transit vehicles: Fire precursor parameters are collected by various sensors installed at the electrical cabinet components of rail transit vehicles, and operating condition parameters of rail transit vehicles are collected simultaneously by the on-board monitoring system of rail transit vehicles. The fire precursor parameters include surface temperature parameters of electrical cabinet components, high-frequency current parameters of coils, smoke concentration parameters inside electrical cabinets, and insulation resistance parameters of components. The operating condition parameters include operating speed, load current, ambient temperature, and operating time. The fire precursor parameters and the operating condition parameters are processed synchronously to obtain the original monitoring dataset of the electrical cabinet components of the rail transit vehicle. Each data entry in the original monitoring dataset includes a collection timestamp, sensor number, fire precursor parameters, and corresponding operating condition parameters.

3. The method according to claim 2, characterized in that, The original monitoring dataset of the electrical cabinet components of the rail transit vehicle is subjected to condition-adaptive filtering to obtain an effective monitoring dataset, including: The surface temperature parameter in each data entry is processed by moving average filtering to obtain the surface temperature purification parameter. Wavelet threshold filtering is performed on the coil high-frequency current parameters in each data entry to obtain the coil high-frequency current purification parameters. Median filtering was performed on the smoke concentration parameters inside the electrical cabinet and the insulation resistance parameters of the components in each data entry to obtain the smoke concentration purification parameters inside the electrical cabinet and the insulation resistance purification parameters of the components. The surface temperature purification parameter, the coil high-frequency current purification parameter, the smoke concentration purification parameter inside the electrical cabinet, and the component insulation resistance purification parameter are normalized to obtain an effective monitoring dataset.

4. The method according to claim 1, characterized in that, The fire risk-related characteristic parameters include static characteristic parameters and dynamic characteristic parameters. The static characteristic parameters include: the average value of the surface temperature parameter, the maximum value of the surface temperature parameter, the peak value of the surface temperature parameter, the effective value of the coil high-frequency current parameter, the peak factor of the coil high-frequency current parameter, the average value of the smoke concentration parameter inside the electrical cabinet, the minimum value of the component insulation resistance parameter, and the stable value of the component insulation resistance parameter. The dynamic characteristic parameters include: the rate of increase of the surface temperature parameter, the difference between the surface temperature and the ambient temperature, the harmonic distortion rate of the coil high-frequency current parameter, the fluctuation amplitude of the coil high-frequency current parameter, the rate of increase of the smoke concentration parameter inside the electrical cabinet, and the rate of decrease of the component insulation resistance parameter. The fire risk correlation feature parameters extracted from the effective monitoring dataset are subjected to fire risk key feature screening processing to obtain a set of fire risk key feature parameters, including: Establish a mapping function between fire risk status and fire risk-related characteristic parameters, wherein the fire risk status includes no risk status, low risk status, medium risk status and high risk status; Based on the mapping relationship function, calculate the sensitivity of the fire risk status to each fire risk-related characteristic parameter; The fire risk-related feature parameters with a sensitivity not lower than a preset sensitivity threshold are collected into a set of key fire risk feature parameters.

5. The method according to claim 1, characterized in that, The method further includes: A fire risk assessment model is constructed to determine the target weights corresponding to each key fire risk characteristic parameter in the set of key fire risk characteristic parameters. The fire risk assessment model includes a target layer, a criterion layer, and an indicator layer. The target layer is used to characterize the real-time fire risk value. The criterion layer includes surface temperature characteristics, coil high-frequency current characteristics, cabinet smoke concentration characteristics, and component insulation resistance characteristics. The indicator layer corresponds to each key fire risk characteristic parameter.

6. The method according to claim 5, characterized in that, The determination of the target weight corresponding to each key fire risk feature parameter in the set of key fire risk feature parameters includes: Based on the criterion-level evaluation results, a first fuzzy complementary judgment matrix is ​​constructed relative to the target layer of the criterion layer, and a second fuzzy complementary judgment matrix is ​​constructed relative to the criterion layer of each index layer based on the index-level evaluation results. The criterion-level evaluation results represent the interval scores obtained by pairwise comparison of surface temperature characteristics, coil high-frequency current characteristics, cabinet smoke concentration characteristics, and component insulation resistance characteristics. The index-level evaluation results represent the interval scores obtained by pairwise comparison of key fire risk characteristic parameters under the same criterion. The criterion layer weight vector is solved based on the first fuzzy complementary judgment matrix, and the weight vector of each index layer is solved based on the second fuzzy complementary judgment matrix. The criterion layer feature matrix is ​​constructed from the criterion layer weight vector, and the index layer feature matrix is ​​constructed from the weight vector of each index layer. The first fuzzy complementary judgment matrix is ​​compared with the feature matrix of the criterion layer for compatibility, and the second fuzzy complementary judgment matrix is ​​compared with the feature matrix of the index layer for compatibility, so as to determine whether the weight allocation results of each level are valid. After all levels have passed the consistency test, the weight vector of the criterion layer and the weight vector of the corresponding indicator layer are weighted and fused together in combination with the preset hierarchical weight constraint relationship to obtain the target weight corresponding to each key characteristic parameter of fire risk.

7. The method according to claim 6, characterized in that, The step of performing a compatibility comparison between the first fuzzy complementary judgment matrix and the criterion layer feature matrix, and a compatibility comparison between the second fuzzy complementary judgment matrix and the index layer feature matrix, to determine whether the weight allocation results at each level are valid, includes: Calculate the first compatibility index between the first fuzzy complementary judgment matrix and the feature matrix of the criterion layer. If the value of the first compatibility index is less than the first preset index threshold, then the weight allocation result of the constructed first fuzzy complementary judgment matrix is ​​determined to be valid. And calculate the second compatibility index between the second fuzzy complementary judgment matrix and the feature matrix of the index layer. If the value of the second compatibility index is less than the second preset index threshold, then the weight allocation result of the constructed second fuzzy complementary judgment matrix is ​​determined to be valid.

8. The method according to claim 6, characterized in that, The preset hierarchical weight constraint relationship means that the sum of the weights corresponding to the surface temperature characteristics, coil high-frequency current characteristics, cabinet smoke concentration characteristics, and component insulation resistance characteristics of the criterion layer is 1, and the cumulative weight of the fire risk key characteristic parameters corresponding to all index layers under each criterion layer is consistent with the weight of the corresponding criterion layer.

9. The method according to claim 1, characterized in that, The real-time fire risk value of electrical cabinet components in rail transit vehicles is calculated using the following formula: Where H represents the real-time fire risk value corresponding to the electrical cabinet components of the rail transit vehicle, m is the total number of key fire risk characteristic parameters obtained after sensitivity analysis screening, and W i x represents the final target weight obtained after hierarchical weight fusion calculation of the i-th key fire risk characteristic parameter. i,norm This represents the valid monitoring data obtained after normalization of the i-th key fire risk characteristic parameter.

10. The method according to claim 1, characterized in that, The method further includes: Based on the real-time fire risk value, a fire risk warning level and a corresponding handling suggestion are generated, and the warning level and the corresponding handling suggestion are dynamically updated based on the real-time fire risk value.

11. The method according to claim 10, characterized in that, The step of generating fire risk warning levels and corresponding handling suggestions for each warning level based on the real-time fire risk value includes: Based on the real-time fire risk value and the range of each warning level, a fire risk warning level and a corresponding handling suggestion are generated. The warning levels include Level 1, Level 2, Level 3, and Level 4. The higher the warning level, the more urgent the corresponding action recommendations.

12. A fire risk early warning device for electrical cabinet components, characterized in that, include: The data processing module is used to perform condition-adaptive filtering on the raw monitoring dataset of electrical cabinet components of rail transit vehicles to obtain an effective monitoring dataset. The raw monitoring dataset includes fire precursor parameters of electrical cabinet components and operating condition parameters of rail transit vehicles. The feature filtering module is used to perform fire risk key feature filtering on the fire risk-related feature parameters extracted from the effective monitoring dataset to obtain a set of fire risk key feature parameters. The weight determination module is used to determine the target weight corresponding to each fire risk key feature parameter in the fire risk key feature parameter set. The risk value calculation module is used to calculate the real-time fire risk value of electrical cabinet components based on each key fire risk characteristic parameter and the target weight corresponding to each key fire risk characteristic parameter.

Images