Photovoltaic connector fault diagnosis method and system based on three-point temperature difference longitudinal comparison

Abstract

The application provides a photovoltaic connector fault diagnosis method and system based on three-point temperature difference longitudinal comparison, relates to the technical field of fault diagnosis, and the method comprises the following steps: collecting multi-point temperature data of a photovoltaic connector node to perform thermal topology modeling and generating a longitudinal temperature distribution sequence; identifying a heat flow direction based on the longitudinal temperature distribution sequence and generating a candidate thermal conduction path; calculating a fault feature confidence of the candidate thermal conduction path; generating a fault explanation text according to the fault feature confidence and constructing a fault diagnosis result list. Through the application, the technical problem that maintenance is blind and components are replaced mistakenly due to the fact that single-point temperature measurement cannot distinguish whether a heat source comes from a connector or a cable can be solved, the heat flow direction is identified and the fault position is attributed through three-point synchronous temperature measurement, accurate positioning is achieved, and the technical effect of reducing invalid maintenance and replacement loss is achieved.

Description

Technical Field

[0001] This application relates to the field of fault diagnosis technology, and in particular to a method and system for fault diagnosis of photovoltaic connectors based on longitudinal comparison of three-point temperature differences. Background Technology

[0002] With the continuous growth of photovoltaic power generation capacity and the rapid development of photovoltaic power plants towards large-scale and intensive operations, photovoltaic connectors, as key components for electrical connection in photovoltaic strings, have a decisive impact on the overall power generation efficiency and fire safety of the power plant due to their operational safety and reliability. During long-term operation, factors such as pin and socket oxidation, mechanical loosening, or improper installation can lead to increased contact resistance and localized abnormal heating in photovoltaic connectors, becoming one of the main causes of fires in photovoltaic power plants. Currently, existing photovoltaic connector temperature monitoring solutions mainly employ the method of installing a single temperature sensor inside or on the surface of the connector housing. This involves measuring the absolute temperature of the connector body and comparing it with a preset threshold to determine if a thermal fault has occurred. This monitoring method only focuses on the temperature value of the individual connector and fails to include temperature changes at the cable crimping points at both ends of the connector within the monitoring scope.

[0003] In summary, existing technologies suffer from the problem that the temperature sensors are limited to the connector body and lack the ability to simultaneously sense the temperature at the cable crimping points at both ends. This makes it impossible to distinguish whether the heat source is poor contact within the connector itself or conduction from cable overload or crimping failure when the connector body temperature rises. Furthermore, this makes it difficult for maintenance personnel to accurately determine the root cause of the fault and identify the specific components that need to be replaced or repaired after receiving a temperature alarm. As a result, maintenance becomes more blind, fault handling time is prolonged, and unnecessary economic losses are caused by mistakenly replacing intact components. Summary of the Invention

[0004] The purpose of this application is to provide a photovoltaic connector fault diagnosis method and system based on three-point temperature difference longitudinal comparison. This method aims to solve the technical problems in the prior art, where the temperature sensor is limited to the connector body and lacks the ability to synchronously sense the temperature at the cable crimping points at both ends. As a result, when the connector body temperature rises, it is impossible to distinguish whether the heat source is caused by poor contact of the connector itself or by overload or crimping failure of one side of the cable. This further affects the maintenance personnel's inability to accurately determine the root cause of the fault and identify the specific parts that need to be replaced or repaired after receiving a temperature alarm. Consequently, it leads to increased blind maintenance, prolonged fault handling time, and unnecessary economic losses caused by mistakenly replacing intact parts.

[0005] In view of the above problems, this application provides a photovoltaic connector fault diagnosis method and system based on three-point temperature difference longitudinal comparison.

[0006] Firstly, this application provides a photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison, implemented through a photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison, including: collecting multi-point temperature data of photovoltaic connector nodes to perform thermal topology modeling and generate a longitudinal temperature distribution sequence; identifying the heat flow direction based on the longitudinal temperature distribution sequence and generating candidate heat conduction paths; calculating the fault feature confidence of the candidate heat conduction paths; generating fault explanation text according to the fault feature confidence, and constructing a fault diagnosis result list.

[0007] Preferably, the photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison further includes: defining a longitudinal monitoring topology, which includes a photovoltaic connector node, a left cable node, and a right cable node; collecting the multi-point temperature data and mapping it to the longitudinal monitoring topology to construct the longitudinal temperature distribution sequence.

[0008] Preferably, the photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison further includes: taking the photovoltaic connector node as the center, calculating the temperature difference between it and the left cable node and the right cable node based on the longitudinal temperature distribution sequence, and constructing a longitudinal temperature difference gradient vector; identifying the heat source center node according to the directional characteristics of the longitudinal temperature difference gradient vector; and generating candidate thermal conduction paths in the direction of heat flow by extending from the heat source center node to the low temperature node.

[0009] Preferably, the photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison further includes: if the longitudinal temperature difference gradient vector points to the left cable node or the right cable node, then the center position of the heat source is identified as the photovoltaic connector node, and a center diffusion type heat conduction path is generated; if the longitudinal temperature difference gradient vector points to the photovoltaic connector node, then the center position of the heat source is identified as the corresponding left cable node or the right cable node, and a single-sided injection type heat conduction path is generated; if the value of the longitudinal temperature difference gradient vector is within a preset error range, then an environmentally balanced type heat conduction path is generated.

[0010] Preferably, the photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison further includes: obtaining fault feature confidence indicators, wherein the fault feature confidence indicators include longitudinal temperature difference significance indicators, three-phase temperature balance indicators, temperature change rate consistency indicators, and ambient temperature correlation indicators; extracting fault features from the candidate heat conduction paths based on the fault feature confidence indicators to obtain a set of path feature vectors; performing weighted calculations on the set of path feature vectors to output the fault feature confidence of the candidate heat conduction paths.

[0011] Preferably, the photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison further includes: obtaining the maximum and minimum temperature values ​​in the longitudinal temperature distribution sequence; calculating the absolute value of the difference between the maximum and minimum temperature values, and normalizing it in conjunction with the ambient temperature benchmark to obtain a longitudinal temperature difference significance index, which is used to measure the severity of thermal faults.

[0012] Preferably, the photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison further includes: calculating the first absolute value of the temperature difference between the left cable node and the photovoltaic connector node, and the second absolute value of the temperature difference between the right cable node and the photovoltaic connector node, based on the longitudinal temperature distribution sequence; and calculating the ratio of the difference between the first absolute value of the temperature difference and the second absolute value of the temperature difference as a three-phase temperature balance index.

[0013] Preferably, the photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison further includes: obtaining a longitudinal temperature distribution sequence within a preset time window, calculating the first derivative of the temperature change over time for the photovoltaic connector node, the left cable node, and the right cable node respectively, to obtain a temperature change rate sequence; and calculating the Pearson correlation coefficient between the temperature change rate of the photovoltaic connector node and the corresponding temperature change rates of the left and right cable nodes as a temperature change rate consistency index.

[0014] Preferably, the photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison further includes: sorting the candidate heat conduction paths in descending order, and determining the candidate heat conduction path with the highest confidence of the fault feature as the target fault path; if the target fault path is the center diffusion type heat conduction path, the fault type is determined to be poor contact of the photovoltaic connector; if the target fault path is the single-sided injection type heat conduction path, the fault type is determined to be the corresponding side cable crimping fault; if the target fault path is the environmental uniform type heat conduction path, the fault type is determined to be abnormal ambient temperature or current overload; and generating the fault explanation text based on the heat source center node and longitudinal temperature difference gradient vector of the target fault path.

[0015] Secondly, this application also provides a photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison, used to execute the photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison as described in the first aspect, including: a thermal topology modeling module, used to collect multi-point temperature data of photovoltaic connector nodes to perform thermal topology modeling and generate a longitudinal temperature distribution sequence; a heat flow direction identification module, used to identify the heat flow direction based on the longitudinal temperature distribution sequence and generate candidate heat conduction paths; a fault feature confidence calculation module, used to calculate the fault feature confidence of the candidate heat conduction paths; and a fault explanation text generation module, used to generate fault explanation text according to the fault feature confidence and construct a fault diagnosis result list.

[0016] The technical solution provided in this application has at least the following technical effects or advantages: by achieving the technical goal of accurately identifying the direction of heat conduction based on three-point synchronous temperature measurement and attributing the fault to the connector body or the corresponding side cable crimping point, it achieves the technical effect of providing clear and operable fault location information for operation and maintenance personnel, reducing ineffective maintenance time, and avoiding incorrect replacement of intact parts due to misjudgment, thereby reducing the operation and maintenance cost of power plants.

[0017] The above description is merely an overview of the technical solution of this application. To enable a clearer understanding of the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison, as described in this application.

[0020] Figure 2 This is a cross-sectional view of the internal structure of the photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison in this application.

[0021] Figure 3 This is a schematic diagram of the photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison in this application.

[0022] Figure labeling: Thermal topology modeling module 1, heat flow direction identification module 2, fault feature confidence calculation module 3, fault explanation text generation module 4, temperature measuring device 201, left cable 202, photovoltaic connector 203, right cable 204. Detailed Implementation

[0023] This application provides a photovoltaic connector fault diagnosis method and system based on three-point temperature difference longitudinal comparison. It solves the technical problem in existing technologies where the temperature sensor's placement is limited to the connector body, lacking synchronous sensing capability at the cable crimping points at both ends. This leads to the inability to distinguish whether the heat source is poor contact within the connector itself or conduction from cable overload or crimping fault on one side when the connector body temperature rises. Furthermore, this hinders maintenance personnel from accurately identifying the root cause of the fault and determining the specific components requiring replacement or repair after receiving temperature alarms, resulting in increased blind maintenance, prolonged fault handling time, and unnecessary economic losses due to the incorrect replacement of intact components. The method achieves the technical goal of accurately identifying the direction of heat conduction based on three-point synchronous temperature measurement and attributing the fault to the connector body or the corresponding cable crimping point. This provides maintenance personnel with clear and actionable fault location information, reduces ineffective maintenance time, and avoids the incorrect replacement of intact components due to misdiagnosis, thereby reducing power plant operation and maintenance costs.

[0024] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0025] Example 1, please refer to Figure 1 and Figure 2 This application provides a photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison, which is applied to a photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison, and specifically includes the following steps: S1: Collect multi-point temperature data of photovoltaic connector nodes to perform thermal topology modeling and generate a longitudinal temperature distribution sequence.

[0026] Furthermore, this application also includes: defining a longitudinal monitoring topology, the longitudinal monitoring topology including photovoltaic connector nodes, left cable nodes and right cable nodes; collecting the multi-point temperature data and mapping it to the longitudinal monitoring topology to construct the longitudinal temperature distribution sequence.

[0027] Specifically, the longitudinal monitoring topology refers to the spatial arrangement of temperature measurement points for the photovoltaic connector and its adjacent cables, serving as a spatial reference frame for subsequent temperature data acquisition. A photovoltaic connector node is defined as a temperature acquisition location located on the outer surface or internal heat-conducting medium of the photovoltaic connector body, reflecting the actual thermal state of the photovoltaic connector during operation. A left-side cable node is defined as a temperature acquisition location located on the surface of the insulation layer or near the crimp terminal of the left-side cable of the photovoltaic connector, reflecting the thermal effects of current conduction or contact resistance in the left-side cable. A right-side cable node is defined as a temperature acquisition location located on the surface of the insulation layer or near the crimp terminal of the right-side cable of the photovoltaic connector, reflecting the thermal effects of the right-side cable. The longitudinal monitoring topology includes left-side and right-side cable nodes arranged sequentially along the longitudinal axis of the photovoltaic connector, located on the left and right sides of the photovoltaic connector node, respectively. These three nodes together form a set of spatially fixed monitoring points, providing the spatial positioning basis for simultaneous three-point temperature measurement.

[0028] Furthermore, multi-point temperature data refers to the temperature measurements collected by the temperature sensor, i.e., the temperature measuring device 201, at three locations: the left cable node, the photovoltaic connector node, and the right cable node. These three temperature measurements correspond to three spatial points. For example... Figure 2 As shown, the temperature measuring device 201 is in close contact with the insulation layer of the left cable 202, the photovoltaic connector 203, and the right cable 204.

[0029] Each temperature measurement value is assigned the corresponding node coordinates in the longitudinal monitoring topology according to its spatial location of acquisition, thus establishing a one-to-one correspondence between each temperature value and its associated node location. The three mapped temperature values ​​are then arranged in a fixed spatial order: left cable node, photovoltaic connector node, and right cable node, forming a spatially ordered array of temperature values, known as the longitudinal temperature distribution sequence. This sequence preserves the temperature information of the three measurement points at the same time and their relative spatial relationships, providing the basic data input for subsequent thermal gradient calculations and heat flow direction identification.

[0030] S2: Based on the longitudinal temperature distribution sequence, the heat flow direction is identified, and candidate heat conduction paths are generated.

[0031] Furthermore, this application also includes: taking the photovoltaic connector node as the center, calculating the temperature difference between it and the left cable node and the right cable node based on the longitudinal temperature distribution sequence, and constructing a longitudinal temperature difference gradient vector; identifying the heat source center node according to the directional characteristics of the longitudinal temperature difference gradient vector; and generating candidate thermal conduction paths in the direction of heat flow by extending from the heat source center node to the low temperature node.

[0032] Furthermore, this application also includes: if the longitudinal temperature difference gradient vector points to the left cable node or the right cable node, then the center position of the heat source is identified as the photovoltaic connector node, and a center-diffusion type heat conduction path is generated; if the longitudinal temperature difference gradient vector points to the photovoltaic connector node, then the center position of the heat source is identified as the corresponding left cable node or the right cable node, and a unilateral injection type heat conduction path is generated; if the value of the longitudinal temperature difference gradient vector is within a preset error range, then an environmentally balanced type heat conduction path is generated.

[0033] Specifically, the photovoltaic connector node is set as the reference center point for temperature difference calculation. Based on the temperature values ​​of the left cable node, the photovoltaic connector node, and the right cable node stored in the longitudinal temperature distribution sequence, a first subtraction operation and a second subtraction operation are performed respectively. The first subtraction operation is the photovoltaic connector node temperature value minus the left cable node temperature value, and the second subtraction operation is the photovoltaic connector node temperature value minus the right cable node temperature value. The results of the two subtraction operations are the two temperature difference values. The longitudinal temperature difference gradient vector is a two-dimensional array. The first dimension of this two-dimensional array is the temperature difference value between the photovoltaic connector node and the left cable node, and the second dimension is the temperature difference value between the photovoltaic connector node and the right cable node. The direction of this two-dimensional array is determined by the sign and relative magnitude of the two temperature difference values. The constructed longitudinal temperature difference gradient vector fully expresses the potential flow trend of heat along the longitudinal axis from the high-temperature region to the low-temperature region.

[0034] Next, the directional feature refers to the pointing pattern jointly presented by the sign attribute and the comparison result of the absolute value of the two temperature difference values ​​in the longitudinal temperature difference gradient vector. If the first dimension element is positive and the second dimension element is positive, the directional feature is characterized by positive values ​​on both sides; if the first dimension element is negative and the second dimension element is negative, the directional feature is characterized by negative values ​​on both sides; if the first dimension element is positive and the second dimension element is negative, or the first dimension element is negative and the second dimension element is positive, the directional feature is characterized by opposite signs on one side. The heat source center node refers to the starting point pointed to by the longitudinal temperature difference gradient vector, that is, the spatial point with the highest temperature and outward heat dissipation. It is identified based on directional characteristics. If the directional characteristic shows positive values ​​on both sides, it means that the temperature of the photovoltaic connector node is higher than that of the left and right cable nodes, and the photovoltaic connector node is identified as the heat source center node. If the directional characteristic shows negative values ​​on both sides, it means that the temperature of the photovoltaic connector node is lower than that of the left and right cable nodes, and the one with the higher temperature is identified as the heat source center node. If the directional characteristic shows opposite signs on one side, it means that the temperature of the photovoltaic connector node is higher than that of the cable nodes on one side but lower than that of the cable nodes on the other side, and the node with the highest temperature and higher than the other two nodes is identified as the heat source center node.

[0035] Then, the heat source center node is determined as the starting point, and the low-temperature node refers to the node with the lowest temperature value in the longitudinal temperature distribution sequence. Starting from the heat source center node, the spatial connection is formed by sequentially passing through intermediate nodes and reaching the low-temperature node according to the sign direction of the longitudinal temperature difference gradient vector. The candidate heat conduction path is an ordered sequence of nodes connected in descending order of temperature, consisting of the heat source center node, intermediate nodes, and low-temperature node. Each adjacent node connection in this ordered sequence is consistent with the heat flow direction indicated by the longitudinal temperature difference gradient vector. This ordered node sequence completely represents the possible routes for heat to be transferred from the high-temperature source along the longitudinal axis to the low-temperature region in the actual physical space, which is the original basis for subsequent fault type determination.

[0036] If the first temperature difference obtained by subtracting the temperature of the left cable node from the temperature of the photovoltaic connector node is positive, and the second temperature difference obtained by subtracting the temperature of the right cable node from the temperature of the photovoltaic connector node is also positive, or if the first temperature difference is positive and the second temperature difference is negative but the absolute value of the first temperature difference is greater than the absolute value of the second temperature difference, then the direction of the longitudinal temperature gradient vector points to the side with the lower temperature, either the left or right cable node. If the center of the heat source is identified as a photovoltaic connector node, the determination condition indicates that the temperature of the photovoltaic connector node is simultaneously higher than both the left and right cable nodes, or higher than the cable node with the higher temperature. Therefore, the heat originates from the photovoltaic connector node itself. A center-diffusion type heat conduction path refers to an ordered sequence of paths where heat is transferred from the photovoltaic connector node to both the left and right cable nodes simultaneously, starting from the photovoltaic connector node. This path sequence is characterized by the heat source being located at the geometric center of the three-point monitoring structure, with heat diffusing symmetrically or asymmetrically to the lower-temperature regions on both sides.

[0037] Next, if the reciprocal of the first temperature difference obtained by subtracting the photovoltaic connector node temperature from the left cable node temperature is positive, or if the reciprocal of the second temperature difference obtained by subtracting the photovoltaic connector node temperature from the right cable node temperature is positive, the direction of the longitudinal temperature difference gradient vector points towards the lower-temperature photovoltaic connector node, indicating that the heat does not originate from the photovoltaic connector node itself. When the left cable node temperature is higher than both the photovoltaic connector node temperature and the right cable node temperature, the center of the heat source is identified as the left cable node. When the right cable node temperature is higher than both the photovoltaic connector node temperature and the left cable node temperature, the center of the heat source is identified as the right cable node. Taking the higher-temperature cable node (left or right) as the starting point of the heat source, heat is unidirectionally transferred from this starting point along the longitudinal axis to the photovoltaic connector node, and further to the opposite cable node in an ordered path sequence. This path sequence is characterized by the heat source being located at one end of the three-point monitoring structure, with heat being transferred unidirectionally to the center node and the opposite node.

[0038] Then, the preset error range refers to the pre-set temperature difference judgment threshold range, which is used to measure whether the temperature difference between the three nodes is statistically significant. If the absolute value of the temperature difference between the photovoltaic connector node and the left cable node is less than the pre-set first threshold, and the absolute value of the temperature difference between the photovoltaic connector node and the right cable node is less than the pre-set second threshold, or the ratio of the temperature difference values ​​on both sides is close to 1, a heat transfer path is directly generated without specifying a specific heat source center node. The characteristic of this path is that the temperature values ​​of the three nodes are close to each other, there is no significant temperature gradient, so the heat does not flow in a clear direction between the three nodes, and the temperature distribution shows an overall uniformity.

[0039] S3: Calculate the confidence level of the fault characteristics of the candidate thermal conduction path.

[0040] Furthermore, this application also includes: obtaining a fault feature confidence index, wherein the fault feature confidence index includes a longitudinal temperature difference significance index, a three-phase temperature balance index, a temperature change rate consistency index, and an ambient temperature correlation index; extracting fault features from the candidate heat conduction path based on the fault feature confidence index to obtain a path feature vector set; performing a weighted calculation on the path feature vector set to output the fault feature confidence of the candidate heat conduction path.

[0041] Furthermore, this application also includes: obtaining the maximum and minimum temperature values ​​in the longitudinal temperature distribution sequence; calculating the absolute value of the difference between the maximum and minimum temperature values, and normalizing it in conjunction with an ambient temperature benchmark to obtain a longitudinal temperature difference significance index, wherein the longitudinal temperature difference significance index is used to measure the severity of thermal failure.

[0042] Furthermore, this application also includes: based on the longitudinal temperature distribution sequence, calculating the first absolute value of the temperature difference between the left cable node and the photovoltaic connector node, and the second absolute value of the temperature difference between the right cable node and the photovoltaic connector node; calculating the ratio of the difference between the first absolute value of the temperature difference and the second absolute value of the temperature difference as a three-phase temperature balance index.

[0043] Furthermore, this application also includes: obtaining a longitudinal temperature distribution sequence within a preset time window, calculating the first derivative of the temperature change over time for the photovoltaic connector node, the left cable node, and the right cable node respectively, to obtain a temperature change rate sequence; and calculating the Pearson correlation coefficient between the temperature change rate of the photovoltaic connector node and the corresponding temperature change rates of the left and right cable nodes as a consistency index of temperature change rate.

[0044] Specifically, the fault characteristic confidence index is a set of pre-defined quantitative parameters used to measure the probability and severity of faults reflected by candidate thermal conduction paths from different dimensions. The fault characteristic confidence index consists of four sub-indicators: longitudinal temperature difference significance index, three-phase temperature balance index, temperature change rate consistency index, and ambient temperature correlation index.

[0045] The maximum temperature value refers to the temperature reading with the highest numerical value in the vertical temperature distribution sequence, representing the highest temperature among the three temperature measurement points. The minimum temperature value refers to the temperature reading with the lowest numerical value in the vertical temperature distribution sequence, representing the lowest temperature among the three temperature measurement points. Obtaining the maximum and minimum temperature values ​​in the vertical temperature distribution sequence requires iterating through all three temperature values ​​in the array, comparing each pair to determine the maximum and minimum values, thus preparing input data for subsequent difference calculations.

[0046] Furthermore, the absolute value of the difference refers to the non-negative value of the difference between the maximum and minimum temperature values. This absolute value reflects the maximum temperature span in the three-point temperature distribution. The ambient temperature reference is a pre-set reference temperature value. This reference temperature value can be the ambient air temperature collected by the ambient temperature sensor at the current moment, or it can be the average ambient temperature after filtering, used to eliminate the influence of the ambient temperature level on the absolute value of the temperature difference. Dividing the absolute value of the difference by the sum of the ambient temperature reference and a preset constant, or multiplying it by a preset proportional coefficient, ensures that the result falls within a uniform dimensionless interval, making the temperature difference data under different environmental conditions comparable. The longitudinal temperature difference significance index is a dimensionless value obtained after normalization. The magnitude of this value directly characterizes the significance of the temperature difference between the three temperature measurement points. The severity of the thermal fault is a quantitative description of the heat intensity generated by the photovoltaic connector or cable due to increased contact resistance, overload current, or insulation aging. A higher longitudinal temperature difference significance index value indicates a stronger thermal fault, while a lower value indicates a milder thermal fault or the absence of a thermal fault. For example, define the temperature measurement values ​​at three temperature measurement points, namely the temperature value T of the left cable node. L Photovoltaic connector node temperature value T C and the temperature value T of the cable node on the right R Maximum temperature value T max The expression is T max =max(T L ,T C ,T R This indicates that the element with the largest value among three temperature values ​​is selected. The minimum temperature value is T. min The expression is T min =min(T L ,T C ,T R This represents selecting the element with the smallest value from three temperature values. The absolute value of the difference is ΔT. diff The expression is ΔT diff =|T max -T min |, that is, the absolute value of the difference between the maximum and minimum temperature values, which is the fundamental parameter for measuring the maximum temperature span in a three-point temperature distribution. The absolute value of the difference is ΔT. diff The range of values ​​for ΔT is a real number greater than or equal to 0. diff A value of 0 indicates that the temperatures at the three temperature measurement points are exactly the same, ΔT diff The larger the value, the more significant the temperature difference between the three points. Ambient temperature reference T envDefined as a pre-set reference temperature value, this reference temperature value can be taken as the ambient air temperature collected by the ambient temperature sensor at the current moment, or as the average ambient temperature after filtering. The ambient temperature reference is used to eliminate the interference of the ambient temperature level on the absolute value of the temperature difference. The first method in the normalization process is to divide the absolute value of the difference by the sum of the ambient temperature reference and the preset constant. The preset constant is set to K1, where K1 is a real number greater than 0, used to prevent the denominator from being too small when the ambient temperature reference approaches zero, which would lead to numerical instability. Then the expression for the normalization result S1 is S1=( The second method in the normalization process is to multiply the absolute value of the difference by a preset scaling factor, K2, which is a real number greater than zero. This scaling factor is used to map the normalization result to a uniform dimensionless interval. The expression for the normalized result S2 is then S2 = ΔT. diff ×K2. The expression for the longitudinal temperature difference significance index S is S=S1 or S=S2, depending on whether the first or second method is selected in the actual implementation. This index S is a dimensionless value obtained after normalization, and its value directly represents the significance of the temperature difference between the three temperature measurement points. The correspondence between the value of the longitudinal temperature difference significance index S and the severity of the thermal fault is as follows: the larger the S value, the stronger the severity of the thermal fault; the smaller the S value, the weaker the severity of the thermal fault; when S approaches zero, it indicates that there is no significant temperature difference between the three temperature measurement points, that is, there is no obvious local thermal fault.

[0047] The temperature values ​​corresponding to the left cable node and the photovoltaic connector node are read from the longitudinal temperature distribution sequence. The temperature value of the left cable node is subtracted from the temperature value of the photovoltaic connector node, and the absolute value of the subtraction is taken. The resulting non-negative value is the first absolute temperature difference, which is used to quantify the degree of temperature deviation between the left cable node and the photovoltaic connector node. The temperature values ​​corresponding to the right cable node and the photovoltaic connector node are read from the longitudinal temperature distribution sequence. The temperature value of the right cable node is subtracted from the temperature value of the photovoltaic connector node, and the absolute value of the subtraction is taken. The resulting non-negative value is the second absolute temperature difference, which is also used to quantify the degree of temperature deviation between the right cable node and the photovoltaic connector node.

[0048] Next, the difference ratio refers to the relative proportion between the absolute values ​​of the first and second temperature differences. This difference ratio is calculated by dividing the first temperature difference (numerator) by the second temperature difference (denominator) and performing a division operation, or vice versa. The quotient obtained is the difference ratio. The three-phase temperature balance index is defined as the aforementioned difference ratio itself. This index characterizes the temperature symmetry between the left and right cable nodes and the photovoltaic connector node. When the difference ratio equals 1, it indicates that the temperature differences between the left and right cable nodes and the photovoltaic connector node are completely equal, and the temperature distribution exhibits perfect left-right symmetry. When the difference ratio is greater than 1, it indicates that the temperature difference between the left and right cable nodes and the photovoltaic connector node is greater than that between the right and right cable nodes and the photovoltaic connector node, and the temperature distribution exhibits a left-side deviation. When the difference ratio is less than 1, it indicates that the temperature difference between the right and right cable nodes and the photovoltaic connector node is greater than that between the left and right cable nodes and the photovoltaic connector node, and the temperature distribution exhibits a right-side deviation.

[0049] A preset time window refers to a pre-defined continuous time interval with a fixed start and end time. Between the start and end times are multiple equally spaced sampling times, each corresponding to a set of longitudinal temperature distribution sequences. Each sampling time contains three data points: the temperature values ​​of the photovoltaic connector node, the left cable node, and the right cable node. The first derivative of temperature over time is the differential of the temperature value with respect to time. It is calculated by dividing the temperature difference between two adjacent sampling times within the preset time window by the corresponding time interval; the quotient is the rate of temperature change at that moment. The sign of the first derivative indicates a rising or falling trend in temperature, and its absolute value indicates the speed of temperature change. Calculating the first derivatives of temperature over time for the photovoltaic connector node, the left cable node, and the right cable node involves performing differential operations on the temperature time series data of each of the three nodes within the preset time window. Each node yields a set of temperature change rate values ​​corresponding to a specific sampling time. The three time series are formed by arranging a set of temperature change rate values ​​of each of the three nodes in the same sampling time order. Each time series corresponds to one node. The three time series together form a temperature change rate sequence, which fully records the dynamic behavior of temperature change at the three temperature measurement points within a preset time window.

[0050] Next, the time series of temperature change rates of the photovoltaic connector node within a preset time window and the time series of temperature change rates of the left cable node within the same preset time window are covariated. The result is then divided by the product of the standard deviations of the two time series; the quotient is the Pearson correlation coefficient between them. The Pearson correlation coefficient ranges from -1 to 1. A positive correlation coefficient close to 1 indicates that the temperature change rates of the two nodes are in the same direction and highly synchronized. A negative correlation coefficient close to -1 indicates that the temperature change rates of the two nodes are in opposite directions. A correlation coefficient close to 0 indicates that there is no linear correlation between the temperature change rates of the two nodes. The temperature change rate consistency index is defined as either one of the two Pearson correlation coefficients or the average of both. This index quantifies the degree of synchronization of temperature change rates between the photovoltaic connector node and the left cable node, and between the photovoltaic connector node and the right cable node. The closer the value is to 1, the more consistent the thermal inertia behavior of the three nodes. The greater the deviation from 1, the more significant the difference in thermal response behavior between a certain node and the other nodes.

[0051] The ambient temperature correlation index is defined as the difference or ratio between the average of three temperature values ​​in the longitudinal temperature distribution sequence and the current ambient temperature collected by the ambient temperature sensor. This difference or ratio is used to eliminate the interference of ambient temperature fluctuations on fault determination.

[0052] Next, for each candidate heat conduction path, four sub-indicators are calculated: longitudinal temperature difference significance index, three-phase temperature balance index, temperature change rate consistency index, and ambient temperature correlation index. The four sub-indicator values ​​corresponding to each candidate heat conduction path are arranged in a preset order to form a multi-dimensional array. Each candidate heat conduction path generates a multi-dimensional array, and the multi-dimensional arrays of all candidate heat conduction paths together constitute a path feature vector set. Each multi-dimensional array in this set corresponds to a unique candidate heat conduction path, and each multi-dimensional array has four dimensions.

[0053] Then, different weighting coefficients are pre-assigned to the longitudinal temperature difference significance index, three-phase temperature balance index, temperature change rate consistency index, and ambient temperature correlation index. The values ​​of the four sub-indicators corresponding to each candidate heat conduction path are multiplied by their respective weighting coefficients, and the four products are then summed. The magnitude of the weighting coefficients is determined based on prior knowledge of the contribution of each sub-indicator to fault diagnosis. A comprehensive score is output after weighted calculation to quantitatively assess the probability or credibility of the fault type represented by each candidate heat conduction path. A higher comprehensive score indicates a greater likelihood of the fault corresponding to that candidate heat conduction path, and vice versa.

[0054] S4: Generate fault explanation text according to the confidence level of the fault characteristics, and construct a fault diagnosis result list.

[0055] Furthermore, this application also includes: sorting the candidate heat conduction paths in descending order, and determining the candidate heat conduction path with the highest confidence level of the fault characteristics as the target fault path; if the target fault path is the center diffusion type heat conduction path, then the fault type is determined to be poor contact of the photovoltaic connector; if the target fault path is the single-sided injection type heat conduction path, then the fault type is determined to be the corresponding side cable crimping fault; if the target fault path is the environmental equalization type heat conduction path, then the fault type is determined to be abnormal ambient temperature or current overload; and generating the fault explanation text based on the heat source center node and longitudinal temperature difference gradient vector of the target fault path.

[0056] Specifically, each candidate heat conduction path has a corresponding fault feature confidence score. All candidate heat conduction paths are sorted in descending order of their fault feature confidence scores, with the path at the top of the sorted sequence having the highest fault feature confidence score. After descending sorting, the candidate heat conduction path with the highest fault feature confidence score is determined as the sole basis for subsequent fault determination, as this path has the highest confidence among all candidate paths.

[0057] Next, the path type where the heat source is centered at the photovoltaic connector node and heat diffuses from the photovoltaic connector node to both the left and right cable nodes corresponds to the case where the longitudinal temperature gradient vector points to the left or right cable node. Due to the abnormally increased contact resistance between the pins and sockets inside the photovoltaic connector caused by oxidation, wear, or mechanical loosening, the increased contact resistance generates Joule heating when current flows, making the temperature of the photovoltaic connector node significantly higher than that of the cable nodes on both sides. When the target fault path is determined to be a center-diffusion type heat conduction path, it indicates that the heat source is located at the photovoltaic connector node, and the fault type is determined to be poor contact of the photovoltaic connector.

[0058] Then, a unilateral injection-type heat conduction path refers to a path where the heat source center is located at the left or right cable node, and heat is injected unidirectionally from this unilateral endpoint to the photovoltaic connector node and the opposite cable node. This path type corresponds to the case where the longitudinal temperature gradient vector points towards the photovoltaic connector node. A corresponding-side cable crimping fault refers to a physical state where the cable on the same side as the heat source center experiences increased crimping resistance at the crimp terminal due to insecure crimping, improper crimping tools, or long-term vibration. Increased crimping resistance generates Joule heating when current flows, causing the temperature of this side's cable node to be significantly higher than that of the photovoltaic connector node and the opposite cable node. When the target fault path is determined to be a unilateral injection-type heat conduction path, it indicates that the heat source is located at the left or right cable node, and the fault type is determined to be a corresponding-side cable crimping fault.

[0059] Subsequently, an environmentally balanced thermal conduction path refers to a path type where no specific heat source center node needs to be specified, and the temperature values ​​of the three temperature measuring points are close to each other with no significant temperature gradient. This path type corresponds to the situation where the value of the longitudinal temperature gradient vector falls within a preset error range. An abnormal ambient temperature refers to a state where the temperature of the environment surrounding the photovoltaic connector exceeds the normal operating temperature range. An increase in ambient temperature will simultaneously affect the three temperature measuring points, causing the overall temperature of the three nodes to rise while the temperature gradient remains within the preset error range. Current overload refers to a state where the operating current flowing through the photovoltaic connector and cable exceeds the rated current carrying capacity limit. Current overload will cause the three temperature measuring points to rise synchronously due to the overall Joule heating effect, while the temperature gradient remains within the preset error range. When the target fault path is determined to be an environmentally balanced thermal conduction path, it indicates that the temperatures of the three nodes are generally consistent without a significant heat source direction. In this case, the fault type is determined to be either an abnormal ambient temperature or a current overload.

[0060] Next, based on the fault type corresponding to the target fault path, the location of the heat source center node, and the direction of the longitudinal temperature difference gradient vector, a diagnostic conclusion text described in natural language is generated. The fault explanation text includes the fault type name, the location of the fault, and the temperature difference value.

[0061] Finally, a structured data list is generated by organizing the fault type, heat source center node location, fault characteristic confidence value, and fault explanation text of the target fault path according to a preset format. Each field in the list occupies a fixed column position, and each diagnostic result occupies a row. This structured data list is used to output to maintenance personnel or upload to the monitoring system.

[0062] In summary, the photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison provided in this application has the following technical effects: by achieving the technical goal of accurately identifying the direction of heat conduction based on three-point synchronous temperature measurement and attributing the fault to the connector body or the corresponding side cable crimping point, it provides clear and operable fault location information for operation and maintenance personnel, reduces ineffective maintenance time, and avoids incorrect replacement of intact parts due to misjudgment, thereby reducing the technical effect of power plant operation and maintenance costs.

[0063] Example 2: Based on the same inventive concept as the photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison in the previous examples, this application also provides a photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison. Please refer to the appendix. Figure 3 The system includes: a thermal topology modeling module 1, used to collect multi-point temperature data of photovoltaic connector nodes for thermal topology modeling and generate a longitudinal temperature distribution sequence; a heat flow direction identification module 2, used to identify the heat flow direction based on the longitudinal temperature distribution sequence and generate candidate heat conduction paths; a fault feature confidence calculation module 3, used to calculate the fault feature confidence of the candidate heat conduction paths; and a fault explanation text generation module 4, used to generate fault explanation text according to the fault feature confidence and construct a fault diagnosis result list.

[0064] Furthermore, the photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison is also used to: define a longitudinal monitoring topology, which includes a photovoltaic connector node, a left cable node, and a right cable node; collect the multi-point temperature data and map it to the longitudinal monitoring topology to construct the longitudinal temperature distribution sequence.

[0065] Furthermore, the photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison is also used to: calculate the temperature difference between the photovoltaic connector node and the left and right cable nodes respectively based on the longitudinal temperature distribution sequence, and construct a longitudinal temperature difference gradient vector; identify the heat source center node according to the directional characteristics of the longitudinal temperature difference gradient vector; and generate candidate thermal conduction paths in the direction of heat flow by extending from the heat source center node to the low temperature node.

[0066] Furthermore, the photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison is also used for: if the longitudinal temperature difference gradient vector points to the left cable node or the right cable node, then the center position of the heat source is identified as the photovoltaic connector node, and a center diffusion type heat conduction path is generated; if the longitudinal temperature difference gradient vector points to the photovoltaic connector node, then the center position of the heat source is identified as the corresponding left cable node or the right cable node, and a single-sided injection type heat conduction path is generated; if the value of the longitudinal temperature difference gradient vector is within a preset error range, then an environmentally balanced type heat conduction path is generated.

[0067] Furthermore, the photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison is also used for: obtaining fault feature confidence indicators, which include longitudinal temperature difference significance indicators, three-phase temperature balance indicators, temperature change rate consistency indicators, and ambient temperature correlation indicators; extracting fault features from the candidate heat conduction paths based on the fault feature confidence indicators to obtain a set of path feature vectors; and performing weighted calculations on the set of path feature vectors to output the fault feature confidence of the candidate heat conduction paths.

[0068] Furthermore, the photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison is also used to: obtain the maximum and minimum temperature values ​​in the longitudinal temperature distribution sequence; calculate the absolute value of the difference between the maximum and minimum temperature values, and perform normalization processing in conjunction with the ambient temperature benchmark to obtain a longitudinal temperature difference significance index, which is used to measure the severity of thermal faults.

[0069] Furthermore, the photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison is also used to: calculate the first absolute value of the temperature difference between the left cable node and the photovoltaic connector node, and the second absolute value of the temperature difference between the right cable node and the photovoltaic connector node, based on the longitudinal temperature distribution sequence; and calculate the ratio of the difference between the first absolute value of the temperature difference and the second absolute value of the temperature difference as a three-phase temperature balance index.

[0070] Furthermore, the photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison is also used to: obtain the longitudinal temperature distribution sequence within a preset time window, calculate the first derivative of the temperature change over time of the photovoltaic connector node, the left cable node, and the right cable node respectively, and obtain the temperature change rate sequence; calculate the Pearson correlation coefficient between the temperature change rate of the photovoltaic connector node and the corresponding temperature change rates of the left and right cable nodes as a temperature change rate consistency index.

[0071] Furthermore, the photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison is also used for: sorting the candidate heat conduction paths in descending order, and determining the candidate heat conduction path with the highest confidence level of the fault feature as the target fault path; if the target fault path is the center diffusion type heat conduction path, the fault type is determined to be poor contact of the photovoltaic connector; if the target fault path is the single-sided injection type heat conduction path, the fault type is determined to be the corresponding side cable crimping fault; if the target fault path is the environmental uniformity type heat conduction path, the fault type is determined to be abnormal ambient temperature or current overload; and generating the fault explanation text based on the heat source center node and longitudinal temperature difference gradient vector of the target fault path.

[0072] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The photovoltaic connector fault diagnosis method and specific examples based on three-point temperature difference longitudinal comparison in the foregoing embodiment 1 are also applicable to the photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison in this embodiment. Through the foregoing detailed description of the photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison, those skilled in the art can clearly understand the photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

[0073] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0074] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison, characterized in that, include: Multi-point temperature data of photovoltaic connector nodes are collected to perform thermal topology modeling and generate a longitudinal temperature distribution sequence. Based on the longitudinal temperature distribution sequence, heat flow direction is identified to generate candidate heat conduction paths; Calculate the failure feature confidence level of the candidate heat conduction path; Based on the confidence level of the fault characteristics, generate fault explanation text and construct a list of fault diagnosis results.

2. The photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison as described in claim 1, characterized in that, Multi-point temperature data of photovoltaic connector nodes are collected for thermal topology modeling to generate a longitudinal temperature distribution sequence, including: Define a longitudinal monitoring topology, which includes photovoltaic connector nodes, left-side cable nodes, and right-side cable nodes; The collected multi-point temperature data is mapped to the longitudinal monitoring topology to construct the longitudinal temperature distribution sequence.

3. The photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison as described in claim 1, characterized in that, Based on the longitudinal temperature distribution sequence, heat flow direction identification is performed to generate candidate heat conduction paths, including: Centered on the photovoltaic connector node, the temperature difference between the connector node and the left and right cable nodes is calculated based on the longitudinal temperature distribution sequence, and a longitudinal temperature difference gradient vector is constructed. Based on the directional characteristics of the longitudinal temperature difference gradient vector, the central node of the heat source is identified; Starting from the central node of the heat source and extending towards the low-temperature node, candidate thermal conduction paths are generated in the direction of heat flow.

4. The photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison as described in claim 3, characterized in that, If the longitudinal temperature difference gradient vector points to the left cable node or the right cable node, the center position of the heat source is identified as the photovoltaic connector node, and a center-diffusion type heat conduction path is generated. If the longitudinal temperature gradient vector points to the photovoltaic connector node, the center position of the heat source is identified as the corresponding left cable node or the right cable node, and a single-sided injection heat conduction path is generated. If the value of the longitudinal temperature difference gradient vector is within the preset error range, an environmentally balanced thermal conduction path is generated.

5. The photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison as described in claim 1, characterized in that, Calculating the failure feature confidence of the candidate heat conduction path includes: Obtain the confidence index of fault characteristics, which includes the longitudinal temperature difference significance index, the three-phase temperature balance index, the temperature change rate consistency index, and the environmental temperature correlation index. Based on the fault feature confidence index, fault features are extracted from the candidate heat conduction path to obtain a set of path feature vectors. The set of path feature vectors is weighted and calculated to output the fault feature confidence level of the candidate thermal conduction path.

6. The photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison as described in claim 5, characterized in that, Calculate the significance index of longitudinal temperature difference, including: Obtain the maximum and minimum temperature values ​​in the longitudinal temperature distribution sequence; The absolute value of the difference between the maximum and minimum temperature values ​​is calculated and normalized in conjunction with the ambient temperature benchmark to obtain the longitudinal temperature difference significance index, which is used to measure the severity of thermal failure.

7. The photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison as described in claim 5, characterized in that, Calculate the three-phase temperature balance index, including: Based on the longitudinal temperature distribution sequence, the absolute value of the first temperature difference between the left cable node and the photovoltaic connector node, and the absolute value of the second temperature difference between the right cable node and the photovoltaic connector node are calculated respectively. The ratio of the absolute value of the first temperature difference to the absolute value of the second temperature difference is calculated and used as an index of the three-phase temperature balance.

8. The photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison as described in claim 5, characterized in that, Calculate the consistency index for temperature change rate, including: Obtain the longitudinal temperature distribution sequence within a preset time window, and calculate the first derivative of the temperature change over time for the photovoltaic connector node, the left cable node, and the right cable node to obtain the temperature change rate sequence. The Pearson correlation coefficient between the temperature change rate of the photovoltaic connector node and the corresponding temperature change rates of the left and right cable nodes is calculated as an indicator of temperature change rate consistency.

9. The photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison as described in claim 4, characterized in that, Generate fault explanation text based on the aforementioned fault characteristic confidence level, including: The candidate thermal conduction paths are sorted in descending order, and the candidate thermal conduction path with the highest confidence level of the fault characteristics is determined as the target fault path; If the target fault path is the center-diffusion type heat conduction path, then the fault type is determined to be poor contact of the photovoltaic connector; If the target fault path is the single-sided injection type heat conduction path, then the fault type is determined to be the corresponding side cable crimping fault. If the target fault path is the environmentally balanced thermal conduction path, then the fault type is determined to be abnormal ambient temperature or current overload. Based on the heat source center node and longitudinal temperature difference gradient vector of the target fault path, the fault explanation text is generated.

10. A photovoltaic connector fault diagnosis system based on three-point temperature difference longitudinal comparison, characterized in that, The steps for implementing the photovoltaic connector fault diagnosis method based on three-point temperature difference longitudinal comparison as described in any one of claims 1 to 9 include: The thermal topology modeling module is used to collect multi-point temperature data of photovoltaic connector nodes to perform thermal topology modeling and generate a longitudinal temperature distribution sequence. A heat flow direction identification module is used to identify the heat flow direction based on the longitudinal temperature distribution sequence and generate candidate heat conduction paths. The fault feature confidence calculation module is used to calculate the fault feature confidence of the candidate thermal conduction path; The fault explanation text generation module is used to generate fault explanation text according to the confidence level of the fault characteristics and construct a fault diagnosis result list.

Images