Fault positioning method and device based on association rules, equipment and program product

Abstract

The embodiment of the invention provides a fault positioning method and device based on association rules, equipment and a program product. Relates to the field of energy storage system fault diagnosis. The method comprises the following steps: acquiring a state code, a fault code and an operation parameter of the energy storage thermal management equipment; based on the state code, the fault code and the operation parameter, generating a fault rule base by adopting an association rule mining algorithm; unsupervised clustering analysis is carried out on the operation parameters, and abnormal data deviating from the standard are screened out; fusing the abnormal data with the fault rule base to generate a fault map; and based on the fault map, a fault positioning report of the energy storage thermal management equipment is output, the fault positioning report at least comprises a fault propagation path and a maintenance scheme, and the fault propagation path is a path of a causal relationship between the fault code and the abnormal mode. The method is used for achieving the effect of improving the fault positioning efficiency and accuracy.

Description

Technical Field

[0001] This application relates to the field of fault diagnosis of energy storage systems, and in particular to a fault location method, device, equipment and program product based on association rules. Background Technology

[0002] In energy storage systems, the internal battery clusters generate a significant amount of heat during charging and discharging. This requires real-time temperature regulation by energy storage thermal management equipment to prevent overheating that could lead to thermal runaway or equipment damage. Failure of the energy storage thermal management equipment can disrupt the safe and stable operation of the battery clusters; therefore, fault location for this equipment is crucial.

[0003] Currently, existing fault location methods mainly rely on threshold-based anomaly detection. This method primarily determines whether equipment is malfunctioning by setting preset threshold ranges for parameters such as temperature, pressure, and current. However, because this method depends on experience, existing fault location methods suffer from low efficiency and poor accuracy. Summary of the Invention

[0004] This application provides a fault location method, apparatus, device, and program product based on association rules, which aims to improve the efficiency and accuracy of fault location.

[0005] Firstly, this application provides a fault location method based on association rules, the method comprising:

[0006] Acquire the status code, fault code, and operating parameters of the energy storage thermal management device; the status code indicates the identifier of the current operating status of the device, the fault code indicates the identifier of the fault that has occurred in the energy storage thermal management device, and the operating parameters indicate the continuous parameters collected during the operation of the energy storage thermal management device;

[0007] Based on the status code, the fault code, and the operating parameters, an association rule mining algorithm is used to generate a fault rule base; the fault rule base contains association rules between the status code, the fault code, and the operating parameters, and the association rule mining algorithm is used to discover implicit associations between data item sets.

[0008] By performing unsupervised cluster analysis on the operating parameters, abnormal data that deviates from the standard are filtered out.

[0009] The abnormal data is fused with the fault rule base to generate a fault map;

[0010] Based on the fault map, a fault location report of the energy storage thermal management equipment is output. The fault location report includes at least the fault propagation path and the maintenance plan. The fault propagation path is the path of causal relationship between the fault code and the abnormal mode.

[0011] In one possible implementation, the step of generating a fault rule base using an association rule mining algorithm based on the status code, the fault code, and the operating parameters includes:

[0012] Based on support, confidence, and lift, association rules between the status code and the operating parameters are selected; the support represents the frequency of the simultaneous occurrence of the status code and the operating parameters, the confidence represents the probability that the operating parameters will occur simultaneously when the status code occurs, and the lift is used to measure the independence between the status code and the operating parameters.

[0013] Based on the association rules and the association rule mining algorithm, a fault rule base is generated;

[0014] The confidence threshold is dynamically adjusted based on the equipment aging index; the equipment aging index is a comprehensive indicator reflecting the operating status of the energy storage thermal management equipment.

[0015] The fault rule base is updated based on the adjusted confidence threshold.

[0016] In one possible implementation, after filtering the association rules between the status code and the operating parameters based on support, confidence, and lift, the method further includes:

[0017] The confidence threshold is adjusted based on changes in the operating environment of the energy storage thermal management device; these changes include ambient temperature or load fluctuations.

[0018] The weights of the association rules are dynamically adjusted based on historical accuracy; the historical accuracy represents historical data on the fault location accuracy of the association rules in the fault rule base.

[0019] The fault rule base is updated based on the confidence threshold and the weight of the association rule.

[0020] In one possible implementation, the step of filtering out outlier data that deviates from the standard by performing unsupervised clustering analysis on the operating parameters includes:

[0021] The operating parameters are extracted as multidimensional features, including time series features, spatial distribution features, and equipment aging indicators; the time series features reflect the changes of operating parameters over time, and the spatial distribution features characterize the layout relationship between the equipment and the battery cluster.

[0022] Unsupervised clustering analysis is performed based on the aforementioned multidimensional features to filter out outlier data that deviate from the criteria.

[0023] In one possible implementation, the step of fusing the abnormal data with the fault rule base to generate a fault map includes:

[0024] Abnormal patterns of the operating parameters were determined through unsupervised clustering analysis;

[0025] Based on the association rules in the fault rule base and the abnormal patterns, a fault map is constructed, which includes a device status code layer, an operating parameter layer and an abnormal pattern layer.

[0026] In one possible implementation, after generating the fault map, the method further includes:

[0027] The fault propagation paths in the fault map are filtered based on the causal reasoning algorithm to eliminate false association paths caused by environmental factors.

[0028] In one possible implementation, the method further includes:

[0029] The fault rule base is adjusted based on a feedback reinforcement learning algorithm, and the weight of the association rule is adjusted according to the verification results.

[0030] Secondly, this application provides a fault location device based on association rules, the device comprising:

[0031] The acquisition module is used to acquire the status code, fault code, and operating parameters of the energy storage thermal management device; the status code indicates the identifier of the current operating status of the device, the fault code indicates the identifier of the fault that has occurred in the energy storage thermal management device, and the operating parameters indicate the continuous parameters collected during the operation of the energy storage thermal management device.

[0032] The first generation module is used to generate a fault rule base based on the status code, the fault code and the operating parameters using an association rule mining algorithm; the fault rule base contains association rules between the status code, the fault code and the operating parameters, and the association rule mining algorithm is used to discover implicit associations between data item sets.

[0033] The filtering module is used to filter out abnormal data that deviates from the standard by performing unsupervised clustering analysis on the operating parameters;

[0034] The second generation module is used to integrate the abnormal data with the fault rule base to generate a fault map;

[0035] The output module is used to output a fault location report of the energy storage thermal management equipment based on the fault map. The fault location report includes at least the fault propagation path and the maintenance plan. The fault propagation path is the path of causal relationship between the fault code and the abnormal mode.

[0036] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0037] The memory stores computer-executed instructions;

[0038] The processor executes computer execution instructions stored in the memory, causing the processor to perform the method described in any of the first aspects above.

[0039] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the method described in any of the first aspects above.

[0040] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the method described in any of the first aspects above.

[0041] This application provides a fault location method, apparatus, equipment, and program product based on association rules. By acquiring the status codes, fault codes, and operating parameters of an energy storage thermal management device, it provides data for subsequent generation of a fault map. Based on the status codes, fault codes, and operating parameters, an association rule mining algorithm can generate a fault rule base, reducing manual intervention and improving fault location efficiency. This fault rule base contains association rules between status codes, fault codes, and operating parameters, and the association rule mining algorithm is used to discover implicit associations between data itemsets. Unsupervised clustering analysis is performed on the operating parameters to filter out abnormal data deviating from the standard, reducing missed detections due to data sparsity. Integrating this abnormal data with the fault rule base generates a fault map, achieving comprehensive association between status codes, fault codes, and operating parameters, improving fault location accuracy. Based on the fault map, a fault location report for the energy storage thermal management device is output. This report includes at least the fault propagation path and a maintenance plan. The fault propagation path is the causal path between the fault code and the abnormal pattern, further improving fault location efficiency and accuracy. Attached Figure Description

[0042] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0043] Figure 1 A flowchart illustrating a fault location method based on association rules provided in an embodiment of this application;

[0044] Figure 2 A flowchart illustrating another fault location method based on association rules provided in this application embodiment;

[0045] Figure 3 A schematic diagram of a fault location device based on association rules provided in an embodiment of this application;

[0046] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0047] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0048] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0049] In this application, the term "comprising" and its variations can refer to non-limiting inclusion; the term "or" and its variations can refer to "and / or". The terms "first", "second", etc., in this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. In this application, "multiple" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0050] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of the relevant data all comply with the relevant laws, regulations, and standards of the relevant countries and regions, have taken necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation access points for users to choose to authorize or refuse.

[0051] Furthermore, the technical solution involved in this application, which involves big data analysis of user information (including but not limited to personal biometrics, identity data, consumption data, asset data, electronic terminal operation data, etc.) and the use of artificial intelligence technology for automated decision-making, and makes decisions that have a significant impact on personal rights based on the results of automated decision-making, provides users with corresponding operation entry points for users to choose to agree to or reject the results of automated decision-making; if the user chooses to reject, the process will proceed to the expert decision-making process.

[0052] First, let's explain some of the terms used in this application:

[0053] Energy storage thermal management equipment refers to the equipment used to regulate temperature in an energy storage system, including fans, liquid cooling units, and air conditioning-related equipment inside energy storage containers.

[0054] Support refers to the proportion of events where both event A and event B occur simultaneously out of the total number of events. Optionally, the formula for calculating support can be: , where n is the total number of events.

[0055] Confidence refers to the frequency with which event B occurs simultaneously when event A occurs. The formula for calculating confidence is: .

[0056] Lift: This refers to the frequency with which events A and B occur together, but the frequency of each event occurring individually must also be considered. Optionally, the lift calculation formula can be: .

[0057] If the lift is greater than 1, events A and B are positively correlated, and the rule is valid. If the lift is equal to 1, events A and B are independently positively correlated, and the rule is invalid. If the lift is less than 1, events A and B are independently negatively correlated, and the rule is misleading.

[0058] Minimum support threshold: refers to setting the minimum support (min_sup) for a task.

[0059] Frequent itemsets are those whose support is greater than or equal to the minimum support (min_sup).

[0060] In energy storage systems, the internal battery clusters generate a significant amount of heat during charging and discharging. This requires real-time temperature regulation by energy storage thermal management equipment to prevent overheating that could lead to thermal runaway or equipment damage. Failure of the energy storage thermal management equipment can disrupt the safe and stable operation of the battery clusters; therefore, fault location for this equipment is crucial.

[0061] The existing fault location methods mainly include the following:

[0062] 1. Threshold-based fault location method: This method mainly determines whether the equipment is abnormal by setting threshold ranges for parameters such as temperature, pressure, and current. However, this method relies on experience, resulting in low efficiency and poor accuracy in fault location.

[0063] 2. Outlier analysis is a fault location method that primarily identifies anomalies in equipment operating parameters using statistical or machine learning methods. However, this method suffers from poor fault location accuracy because it analyzes data from individual devices in isolation and fails to consider the coupling relationships between devices.

[0064] Therefore, existing fault location methods suffer from low efficiency and poor accuracy. Considering these problems, this application proposes a fault location method based on association rules. This method improves both the efficiency and accuracy of fault location.

[0065] The entity executing this association rule-based fault location method can be, for example, a fault location system. Optionally, the fault location system can be any existing electronic device with processing capabilities, such as a terminal or a server. In some embodiments, the fault location system can also be deployed in a server cluster or cloud environment. This application does not limit the deployment environment of the fault location system.

[0066] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments.

[0067] The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will be described below with reference to the accompanying drawings.

[0068] Figure 1 This is a flowchart illustrating a fault location method based on association rules, provided as an embodiment of this application. Figure 1 As shown, the method includes:

[0069] S101, acquire the status code, fault code and operating parameters of the energy storage thermal management device. The status code indicates the identifier of the current operating status of the device, the fault code indicates the identifier of the fault that has occurred in the energy storage thermal management device, and the operating parameters indicate the continuous parameters collected during the operation of the energy storage thermal management device.

[0070] For example, status codes can be used to represent normal operation, standby, fault shutdown, or maintenance mode, etc.

[0071] For example, fault codes can be numeric codes or a combination of letters and numbers. Fault codes can be used to characterize temperature sensor failure, cooling fan failure, or overloaded loading modules, etc.

[0072] Optionally, the fault location system can collect operating parameters in real time based on the sensors built into the energy storage thermal management device, generate status codes through the controller, and generate fault codes when the energy storage thermal management device malfunctions. The status codes, fault codes, and operating parameters of the energy storage thermal management device can be obtained through a local interface based on the controller.

[0073] Optionally, the fault location system can also query the server through an application programming interface (API) based on a preset period and the unique identifier of the energy storage thermal management device to determine the status code, fault code, and operating parameters of the energy storage thermal management device.

[0074] S102, Based on the status code, the fault code, and the operating parameter, an association rule mining algorithm is used to generate a fault rule base. The fault rule base contains association rules between the status code, the fault code, and the operating parameter. The association rule mining algorithm is used to discover implicit associations between data item sets.

[0075] For example, a data item set can be a set of the smallest data units extracted from status codes, fault codes, and operating parameters.

[0076] Optionally, the fault location system can pre-set minimum support and minimum confidence. Based on the data itemset composed of the aforementioned status codes, fault codes, and operating parameters, it can determine frequent itemsets and extract association rules that meet the confidence requirements from these frequent itemsets. Alternatively, the fault location system can construct a frequent pattern tree based on the aforementioned status codes, fault codes, and operating parameters, arrange the data itemsets in descending order of support, extract frequent itemsets from the frequent pattern tree according to mining conditions, and filter out association rules based on confidence requirements.

[0077] Optionally, the fault location system can construct a fault rule library by mapping and storing the mined association rules in a relational database according to core fields such as support, confidence, and rule generation time.

[0078] S103, by performing unsupervised cluster analysis on the operating parameters, outlier data that deviates from the standard are screened out.

[0079] For example, unsupervised clustering analysis can be a label-free clustering algorithm that identifies anomalous patterns in runtime parameters.

[0080] For example, anomalous data can be data points that deviate from normal operating patterns. For instance, anomalous data could be sudden changes in the pressure of a liquid chiller unit, or abnormal fluctuations in air conditioning temperature.

[0081] As one possible implementation, the fault location system can extract multidimensional features of the operating parameters. These multidimensional features include time-series features, spatial distribution features, and equipment aging indicators. The time-series features reflect the changes in operating parameters over time, and the spatial distribution features characterize the layout relationship between the equipment and the battery clusters. Then, based on these multidimensional features, unsupervised clustering analysis is performed to filter out outlier data that deviates from the criteria.

[0082] Optionally, the fault location system can use a sliding window method to divide time segments and calculate statistical characteristics of the operating parameters within each target time window. For example, the fault location system can set the target time window size to 5 minutes, calculating the mean, variance, peak value, and valley value of the voltage data every minute. Alternatively, Fourier transform can be used to perform frequency domain analysis on the time series of operating parameters, extracting periodic features to obtain time series characteristics. It should be understood that this application does not limit the duration of the target time window, nor its start and end times.

[0083] Optionally, the fault location system can calculate the Euclidean distance between a single energy storage thermal management device and the battery cluster based on the physical layout coordinates of the energy storage thermal management device and the battery cluster, thus obtaining the spatial distribution characteristics. Alternatively, it can calculate the density of energy storage thermal management devices within the battery cluster to obtain the spatial distribution characteristics. For example, based on the center location of the battery cluster, the number of energy storage thermal management devices within a 5-meter radius can be calculated to obtain the spatial distribution characteristics.

[0084] Optionally, the fault location system can extract the cumulative runtime feature from the operating parameters and obtain the equipment aging index by the ratio of the cumulative runtime to the design life of the energy storage thermal management equipment.

[0085] Optionally, the fault location system can also construct prompt words based on operating parameters and input these prompt words into a feature extraction model to obtain the time-series features and spatial distribution features of the operating parameters, as well as equipment aging indicators. The prompt words are used to obtain the time-series features, spatial distribution features, and equipment aging indicators of the operating parameters. For example, the feature extraction model can be any pre-trained deep learning model or machine learning model with feature extraction capabilities.

[0086] Optionally, the fault location system can calculate the distance from each feature data point to its cluster center based on the cluster center distance. If the distance from a feature data point to its cluster center exceeds the preset distance range of the cluster, the feature data point is determined to be abnormal data within the cluster. If an isolated point that is not assigned to any cluster appears, the isolated point is determined to be abnormal data.

[0087] Optionally, the fault location system can also calculate the average similarity between the data to be judged and all data in the normal cluster. If the average similarity between the data to be judged and all data in the normal cluster is lower than a preset similarity threshold, then the abnormal data that deviates from the normal standard is judged.

[0088] The above method extracts multidimensional features of the operating parameters, laying the foundation for subsequent unsupervised clustering analysis. Based on these multidimensional features, unsupervised clustering analysis filters out outlier data that deviates from the standard, avoiding missed detections and thus improving the accuracy of fault location.

[0089] S104. The abnormal data is integrated with the fault rule base to generate a fault map.

[0090] For example, a fault map can be a directed graph that displays fault paths with fault codes as nodes and association rules as edges.

[0091] In some embodiments, the fault location system can determine the abnormal pattern of the operating parameter through unsupervised clustering analysis. Then, based on the association rules in the fault rule base and the abnormal pattern, a fault map is constructed, which includes a device status code layer, an operating parameter layer, and an abnormal pattern layer.

[0092] For example, the exception pattern layer could be an exception pattern used to store a fault rule base. For instance, a coupling pattern between abnormal air conditioning temperatures and localized overheating of battery clusters.

[0093] Optionally, the fault location system can obtain abnormal clusters based on unsupervised clustering, extract common features of all data within the clusters, and determine the abnormal pattern corresponding to the abnormal cluster by combining the common features. For example, the common features could be time series features with a variance of less than 5, or spatial distribution features such as the edge region of the battery cluster.

[0094] Optionally, the fault location system can also classify abnormal modes based on the root cause of abnormal operating parameters or their manifestation, and assign an abnormality identifier to each abnormal mode. For example, the fault location system can classify measurement value deviations caused by sensor aging as parameter drift-type abnormal modes, sudden speed drops caused by fan failure as time-series mutation-type abnormal modes, or performance degradation caused by component wear as aging-related abnormal modes.

[0095] Optionally, the fault location system can preset the status codes of the energy storage thermal management equipment as core nodes. Each core node is associated with a status name, status trigger condition, and status duration. These core nodes are stored in a graph database, and a status code encoding index is created for each core node to obtain the equipment status code layer. For example, the associated status name could be normal operation, standby, fault shutdown, or maintenance mode; the status trigger condition could be a fault shutdown trigger condition; and the status duration could be, for example, 3 seconds.

[0096] Optionally, the fault location system can define original operating parameter nodes based on parameter units, normal threshold ranges, and acquisition cycles, and define multi-dimensional feature nodes based on feature calculation methods and normal feature ranges. By using feature extraction algorithms, the original operating parameter nodes and corresponding multi-dimensional feature nodes can be associated to obtain the operating parameter layer.

[0097] For example, the original operating parameter node could be coolant temperature, pump flow rate, or fan speed. Multidimensional feature nodes could be time-series features defining the temperature trend slope, spatial distribution features defining the equipment area, or aging indicators defining the cooling power attenuation rate. Through feature extraction algorithms, for example, the correlation between the original operating parameter node (coolant temperature) and the corresponding temperature trend slope in the time-series features of the multidimensional feature nodes can be constructed.

[0098] Optionally, the fault location system can construct an abnormal mode layer based on the above-mentioned abnormal modes as core nodes, and associate them by abnormal mode codes, abnormal mode types, suspected fault codes, and abnormal severity.

[0099] Optionally, the fault location system can establish matching associations between multi-dimensional feature nodes in the operating parameter layer and abnormal pattern nodes in the abnormal pattern layer through feature matching. Using association rules in the fault rule base, it can establish trigger associations between points in the operating parameter layer and the device status code layer. Based on fault codes and mapping criteria, it can establish mapping associations between the abnormal pattern layer and the device status code layer. Based on these matching associations, trigger associations, and mapping associations, a visualized fault map is generated.

[0100] Optionally, the fault location system can use a graph database visualization tool to display the topology of the equipment status code layer, operating parameter layer, and abnormal mode layer in the fault map. The fault map can also be indexed by status codes, abnormal modes, and fault codes. For example, the fault map might include scenarios such as liquid chiller communication interruption, battery cluster overheating, and abnormal air conditioning temperature.

[0101] Using the methods described above, unsupervised clustering analysis is employed to identify the abnormal patterns of the operating parameters, laying the foundation for subsequent fault mapping and improving the efficiency of fault location. Then, based on the association rules in the fault rule base and the abnormal patterns, a fault map is constructed. This fault map includes a device status code layer, an operating parameter layer, and an abnormal pattern layer, reducing missed fault detections and improving the comprehensiveness of fault location.

[0102] As one possible implementation, after generating the fault map, the fault location system can also filter the fault propagation paths in the fault map based on the causal reasoning algorithm to eliminate false association paths caused by environmental factors.

[0103] For example, a causal reasoning algorithm could model the causal relationship between device status codes and operating parameters to eliminate spurious associations caused by statistical correlation.

[0104] For example, a propagation path can be a path that represents the causal relationship between a fault code and a runtime anomaly.

[0105] For example, environmental factors can be the indirect influence of external conditions on the operating status of energy storage thermal management equipment. External conditions can be, for example, temperature, or load fluctuations.

[0106] Optionally, the fault location system can pre-select nodes of the fault map and environmental factors as nodes of the causal inference model, align the nodes based on time sequence, learn the conditional probabilities between nodes through maximum likelihood estimation, and construct spurious correlation paths caused by environmental factors and causal propagation paths of equipment faults themselves.

[0107] Optionally, the fault location system can introduce environmental factors as conditional variables into the causal inference model for each propagation path in the fault map, and examine whether the association of core nodes in the propagation path satisfies the condition of independence. If the condition of independence is satisfied, the path is determined to be a false association path caused by environmental factors and is excluded. If the condition of independence is not satisfied, it is determined to be a causal propagation path of the equipment's own fault, retained, and marked as a valid path.

[0108] For example, given an environmental factor E, determine whether the association between node P and node M exists. If the association between node P and node M disappears after removing environmental factor E, the path between node P and node M is determined to be a false association path caused by the environmental factor. If a stable association still exists between node P and node M after removing environmental factor E, the path between node P and node M is determined to be a causal propagation path of the device's own fault.

[0109] Optionally, the fault location system can also construct prompt words based on the propagation path of the fault map, and input the prompt words into a causal scoring model to obtain a causal strength score for each propagation path. The prompt words are used to obtain a causal strength score for each propagation path based on the propagation path of the fault map. It should be understood that this causal scoring model can be any pre-trained deep learning model or machine learning model with the function of obtaining causal strength scores based on propagation paths, etc., which will not be elaborated further here.

[0110] Optionally, the fault location system can compare the above causal strength score with a preset causal threshold. If the causal strength score is greater than or equal to the preset causal threshold, the fault propagation path is determined to be a causal propagation path of the device itself. If the causal strength score is less than the preset causal threshold, the fault propagation path is determined to be a false association path caused by environmental factors.

[0111] After generating the fault map using the above method, the fault location algorithm filters the fault propagation paths in the fault map based on causal reasoning. This can eliminate false correlation paths caused by environmental factors and reduce the one-sidedness of fault location caused by isolated data analysis, thereby improving the accuracy of fault location.

[0112] S105, based on the fault map, output a fault location report for the energy storage thermal management equipment. The fault location report includes at least the fault propagation path and the maintenance plan. The fault propagation path is the causal relationship between the fault code and the abnormal mode.

[0113] For example, a maintenance plan can be a device maintenance plan developed based on the root cause of the fault pointed to by the fault propagation path.

[0114] Optionally, the fault location system can locate nodes based on fault maps, fault codes, abnormal patterns, or multiple dimensions. Based on the located node, it extracts the associated nodes between the device status code layer, operating parameter layer, and abnormal pattern layer. According to the association logic of abnormal patterns, operating parameters, and fault codes, the fault propagation path is determined. Based on this fault propagation path, the local solution rule base is queried, and the corresponding maintenance plan is output.

[0115] Optionally, the fault location system can extract node data, fault propagation path information, and maintenance plans through the interface, fill them into a preset structured template, and output a fault location report for the energy storage thermal management equipment.

[0116] As one possible implementation, the fault location system can adjust the fault rule base based on a feedback reinforcement learning algorithm, which adjusts the weight of the association rule according to the verification results.

[0117] Optionally, the fault location system can obtain verification results through an API interface, or supplement them with manual feedback. For example, verification results may include maintenance pass rate, location accuracy rate, and parameter compliance rate.

[0118] Optionally, the fault location system can dynamically adjust the weight of the association rule based on the verification result. If the fault location verification result corresponding to the association rule indicates that the fault location is accurate, the weight of the association rule is increased by a preset upward adjustment amount; if the fault location verification result corresponding to the association rule indicates that the fault location is incorrect, the weight of the association rule is decreased by a preset downward adjustment amount; if the fault location verification result corresponding to the association rule indicates that the fault location is partially matched, the weight of the association rule remains unchanged.

[0119] For example, if the fault location verification result corresponding to association rule X is that the fault location is accurate, the weight of association rule X is adjusted from 0.6 to 0.7 based on an upward adjustment of 10%; if the fault location verification result corresponding to association rule X is that the fault location is incorrect, the weight of association rule X is adjusted from 0.55 to 0.5 based on a downward adjustment of 5%; if the fault location verification result corresponding to association rule X is that the fault location is partially matched, the weight of association rule X remains unchanged at 0.8.

[0120] Using the above method, the fault rule base is adjusted based on the feedback reinforcement learning algorithm. The feedback reinforcement learning algorithm adjusts the weight of the association rule according to the verification results, thereby improving the accuracy of fault location.

[0121] In this embodiment, the status code, fault code, and operating parameters of the energy storage thermal management equipment are acquired to provide data for the subsequent generation of a fault map. Based on the status code, fault code, and operating parameters, an association rule mining algorithm can be used to generate a fault rule base, reducing manual intervention and improving fault location efficiency. By performing unsupervised clustering analysis on the operating parameters, abnormal data deviating from the standard is filtered out, reducing the problem of missed detections due to data sparsity. By fusing the abnormal data with the fault rule base, a fault map can be generated, realizing a comprehensive association between the status code, fault code, and operating parameters, improving the accuracy of fault location. Based on the fault map, a fault location report of the energy storage thermal management equipment is output. The fault location report includes at least the fault propagation path and the maintenance plan. The fault propagation path is the causal path between the fault code and the abnormal pattern, improving the efficiency and accuracy of fault location.

[0122] The following section provides a detailed explanation of how the fault location system generates a fault rule base based on the status code, the fault code, and the operating parameters using an association rule mining algorithm.

[0123] In some embodiments, the fault location system can filter association rules between the status code and the operating parameter based on support, confidence, and lift. Support represents the frequency of the simultaneous occurrence of the status code and the operating parameter; confidence represents the probability that the operating parameter will occur simultaneously when the status code occurs; and lift measures the independence between the status code and the operating parameter. Then, based on the association rules and the association rule mining algorithm, a fault rule base is generated. Next, the confidence threshold is dynamically adjusted according to equipment aging indicators, which are comprehensive indicators reflecting the operating status of the energy storage thermal management equipment. Finally, the fault rule base is updated based on the adjusted confidence threshold.

[0124] Optionally, the fault location system can pre-set a support threshold, an initial confidence threshold, and a lift threshold. Based on these thresholds, it can screen association rules in a step-by-step manner. For example, the fault location system can select first candidate association rules with a support greater than the support threshold of 5%. Based on the first candidate association rules, it can select rules with a confidence greater than the initial confidence threshold of 80% to obtain second candidate association rules. Based on the second candidate association rules, it can select rules with a lift greater than the lift threshold of 1.2 to obtain the target association rule set.

[0125] Optionally, the fault location system can also assign weights to support, confidence, and lift, calculate a comprehensive score for association rules based on these weights, and filter out association rules with a comprehensive score greater than a preset score threshold. For example, the fault location system can assign a support weight of 30%, a confidence weight of 40%, and a lift weight of 30%. If a certain association rule has a comprehensive score of 82, which is greater than the preset score threshold of 70, then the association rule is retained.

[0126] Optionally, the fault location system can classify association rules based on the aforementioned comprehensive score and the association rule mining algorithm. For example, association rules with a comprehensive score greater than or equal to 80 are defined as high-reliability rules, stored in an in-memory database, and prioritized for fault location; association rules with a comprehensive score less than 80 but greater than or equal to 60 are defined as medium-reliability rules, stored in a disk database, and used to assist in fault diagnosis; and association rules with a comprehensive score less than 60 are defined as rules to be verified and stored in a temporary partition.

[0127] Optionally, the fault location system can classify different aging levels based on equipment aging indicators. Based on the above aging levels, a mapping table between aging levels and confidence thresholds is preset. By collecting equipment aging indicators in real time, the system matches the confidence threshold corresponding to the aging level with the equipment aging indicators.

[0128] For example, a fault location system can classify equipment aging indices from 0 to 0.3 as slightly aged, from 0.3 to 0.7 as moderately aged, and from 0.7 to 1 as severely aged. A preset mapping table between aging levels and confidence thresholds includes a confidence threshold of 80% for slightly aged, 70% for moderately aged, and 60% for severely aged. If the current equipment aging index is 0.5, corresponding to a moderately aged level, the confidence threshold can be adjusted to 70% by querying the mapping table.

[0129] Optionally, the fault location system can re-traverse all rules in the fault rule base based on the adjusted confidence threshold and a preset update cycle (e.g., 24 hours). If the confidence of the original association rule is greater than or equal to the adjusted confidence threshold, the rule is retained; if the confidence of the original association rule is less than the adjusted confidence threshold, it is marked as a failed association rule and moved to the historical association rule table for archiving. The fault location system can also re-run the association rule mining algorithm to filter out new association rules with a confidence greater than the adjusted confidence threshold from the newly added running data and add them to the rule fault base.

[0130] Using the methods described above, association rules between the status code and the operating parameter are filtered based on support, confidence, and lift, laying the foundation for generating a fault rule base. Based on these association rules and the association rule mining algorithm, a fault rule base is generated, improving system resource utilization. The confidence threshold is dynamically adjusted according to equipment aging indicators to improve system stability. Based on the adjusted confidence threshold, the fault rule base is updated to improve the accuracy of fault location.

[0131] As one possible implementation, after filtering the association rules between the status code and the operating parameter based on support, confidence, and lift, the fault location system can adjust the confidence threshold based on changes in the operating environment of the energy storage thermal management device, including ambient temperature or load fluctuations. Then, the weight of the association rule is dynamically adjusted based on historical accuracy, which represents historical data on the fault location accuracy of the association rule in the fault rule base. Finally, the fault rule base is updated based on the confidence threshold and the weight of the association rule.

[0132] Optionally, the fault location system can preset an ambient temperature range and a corresponding confidence threshold, and determine the confidence threshold based on changes in the operating environment of the energy storage thermal management device. For example, if the ambient temperature is below 0 degrees Celsius, i.e., a low-temperature region, the confidence threshold is set to 85%; if the ambient temperature is between 0 degrees Celsius and 35 degrees Celsius, i.e., a normal temperature range, the confidence threshold is set to 80%; and if the ambient temperature is above 0 degrees Celsius, i.e., a high-temperature range, the confidence threshold is set to 75%.

[0133] Optionally, the fault location system can determine the confidence threshold based on the load fluctuation range of the energy storage thermal management equipment. The load fluctuation range can be determined by calculating the ratio of the difference between the maximum and minimum load to the rated load. For example, if the load fluctuation range is less than 20%, i.e., low fluctuation, the confidence threshold is set to 80%; if the load fluctuation range is between 20% and 50%, i.e., medium fluctuation, the confidence threshold is set to 70%; and if the load fluctuation range is greater than 50%, i.e., high fluctuation, the confidence threshold is set to 60%.

[0134] Optionally, the fault location system can also determine the confidence threshold based on changes in the operating environment of the energy storage thermal management device and the load fluctuation range. For example, if the operating environment of the energy storage thermal management device changes to a high-temperature range and the load fluctuation range is high, then the joint adjustment threshold is 68%.

[0135] Optionally, in the above method for adjusting the confidence threshold, the adjusted confidence threshold should be greater than the set lower limit of the confidence threshold.

[0136] Optionally, the fault location system can statistically analyze the historical accuracy of the association rule 30 days after the current time, assign a historical accuracy of 0.6 to the historical accuracy of the current time 7 days after the current time, and assign a weight of 0.4 to the historical accuracy of the current time within 8 to 30 days after the current time, calculate the weighted historical accuracy, establish the correspondence between the weighted historical accuracy and the weight of the association rule, and adjust the weight coefficient of the association rule to the weighted historical accuracy.

[0137] Optionally, the fault location system can retain association rules with a confidence level greater than the adjusted confidence threshold and a weight greater than 0.5 in the fault rule base, and delete association rules with a confidence level less than or equal to the adjusted confidence threshold, or association rules with a weight less than 0.1, from the fault rule base to obtain an updated fault rule base.

[0138] Using the above method, after filtering the association rules between the status code and the operating parameter based on support, confidence, and lift, the confidence threshold is adjusted according to changes in the operating environment of the energy storage thermal management equipment to improve the environmental adaptability of the association rules. The weight of the association rule can be dynamically adjusted based on historical accuracy to improve fault location efficiency. Based on the confidence threshold and the weight of the association rule, the fault rule base is updated to improve the rule matching efficiency and accuracy of fault location.

[0139] Figure 2 This is a flowchart illustrating another fault location method based on association rules provided in an embodiment of this application. Figure 2 As shown, the method includes:

[0140] Step 1: Data Acquisition and Preprocessing

[0141] (1) Read historical thermal management equipment fault records, corresponding to the historical operating data of battery clusters and thermal management equipment, as well as fault codes and status codes.

[0142] (2) Discard invalid fault code data, that is, fault codes that cannot be parsed through the communication protocol.

[0143] (3) Based on the principle of association rules.

[0144] (4) Clean and preprocess the data, mark the fault records and actual operation fault data in different time periods, and normalize the data.

[0145] (5) Continuous data through parameter discretization; for example, segmenting and quantizing continuous operating parameters such as temperature change rate, water pump pressure, voltage, and current.

[0146] Step 2: Create a fault rule base

[0147] (1) The processed data is used to discover data with frequent characteristics in the database based on the support. If it is necessary to expand the original database, new data is collected and merged with the original data. Finally, association rules are generated from the frequent itemset based on prior experience.

[0148] (2) Based on the confidence level, discover related data sets with high frequency of synchronous occurrence from the set of related data, summarize effective strong association rules based on the evaluation index lift, and classify the rules.

[0149] (3) Analyze the above rules based on the lift to obtain effective association rules that conform to reality. If the lift is greater than 1, it is considered to be associated; if the lift is greater than 3, it is considered to be worthy of attention.

[0150] (4) Adjust the confidence threshold p adaptively according to the degree of equipment aging.

[0151] (5) Output an event association rule library based on the association rule algorithm.

[0152] (6) Construct a directed graph for nodes that meet the minimum confidence level. If there are nodes A and B in the directed graph, during the confidence level calculation, if the confidence level from node A to node B is greater than the confidence level threshold p, then it is considered that there is an edge from A to B. The confidence level threshold can be dynamically adjusted.

[0153] (7) Based on the generated directed graph, a certain fault is taken as input, and a rule association graph is output. Combined with the fault definition, fault analysis is performed to analyze the fault situation that appears in the historical data and the process of fault propagation in the container.

[0154] (8) Based on the rule association graph analysis process, the fault codes and operation data generated by complex faults are analyzed in association, and the existing propagation paths are given to identify the fault source, assisting maintenance and R&D personnel to quickly locate the abnormal situation on site and improve the analysis efficiency.

[0155] Step 3: Unsupervised Clustering Anomaly Identification

[0156] (1) After obtaining the association rules, perform unsupervised clustering on the relevant operational data with correlation, filter out abnormal outliers that deviate from the standard, and improve the fault identification and detection based on association rules by filtering out abnormal outliers that deviate from the standard.

[0157] (2) Dynamic fusion: The overall confidence score can be calculated by summing the weights assigned to the rule confidence score and the cluster similarity score. The weights assigned to the rule confidence score and the cluster similarity score can be dynamically adjusted based on the historical accuracy.

[0158] (3) When the overall confidence level reaches a certain threshold, a fault alarm will be triggered.

[0159] Step 4: Rule Closed-Loop Optimization

[0160] The verification results of unknown fault alarms are fed back to the rule base to determine whether new rules need to be added or confidence weights need to be adjusted, thereby constructing an evolvable fault knowledge graph and resolving the contradiction between experience solidification and the discovery of unknown faults in energy storage scenarios.

[0161] By combining association rule-based methods with outlier analysis, the above approach couples the internal fault code relationships of thermal management equipment, the association rules of thermal management status codes, the battery cluster fault codes corresponding to thermal management hot standby, and their data outlier characteristics. This assists maintenance and R&D personnel in quickly locating the main fault code, improving analysis efficiency. It also improves upon traditional fault alarms based solely on threshold judgments or data-driven approaches, constructing an evolving fault graph and resolving the contradiction between entrenched experience and the discovery of unknown faults in energy storage scenarios.

[0162] The above are the method embodiments provided in this application. The apparatus provided in this application will be described below.

[0163] Figure 3 This is a schematic diagram of a fault location device based on association rules, provided as an embodiment of this application. Figure 3 As shown, the fault location device 300 based on association rules includes: an acquisition module 301, a first generation module 302, a filtering module 303, a second generation module 304, and an output module 305. Among them,

[0164] The acquisition module 301 is used to acquire the status code, fault code and operating parameters of the energy storage thermal management device; the status code indicates the identifier of the current operating status of the device, the fault code indicates the identifier of the fault that has occurred in the energy storage thermal management device, and the operating parameters indicate the continuous parameters collected during the operation of the energy storage thermal management device.

[0165] The first generation module 302 is used to generate a fault rule base based on the status code, the fault code and the operating parameter using an association rule mining algorithm; the fault rule base contains association rules between the status code, the fault code and the operating parameter, and the association rule mining algorithm is used to discover implicit associations between data item sets.

[0166] The filtering module 303 is used to filter out abnormal data that deviates from the standard by performing unsupervised clustering analysis on the operating parameters;

[0167] The second generation module 304 is used to integrate the abnormal data with the fault rule base to generate a fault map;

[0168] The output module 305 is used to output a fault location report of the energy storage thermal management equipment based on the fault map. The fault location report includes at least the fault propagation path and the maintenance plan. The fault propagation path is the path of causal relationship between the fault code and the abnormal mode.

[0169] Optionally, the first generation module 302 is further configured to filter association rules between the status code and the operating parameter based on support, confidence, and lift; the support characterizes the frequency of the simultaneous occurrence of the status code and the operating parameter, the confidence characterizes the probability that the operating parameter will occur simultaneously when the status code occurs, and the lift is used to measure the independence between the status code and the operating parameter; a fault rule base is generated based on the association rules and the association rule mining algorithm; the confidence threshold is dynamically adjusted according to the equipment aging index; the equipment aging index is used to reflect a comprehensive index of the operating status of the energy storage thermal management equipment; and the fault rule base is updated based on the adjusted confidence threshold.

[0170] Optionally, the first generation module 302 is further configured to: filter the association rules between the status code and the operating parameter based on support, confidence, and lift; adjust the confidence threshold based on changes in the operating environment of the energy storage thermal management device; the changes in the operating environment include ambient temperature or load fluctuations; dynamically adjust the weight of the association rule based on historical accuracy; the historical accuracy represents historical data of the fault location accuracy of the association rule in the fault rule base; and update the fault rule base based on the confidence threshold and the weight of the association rule.

[0171] Optionally, the screening module 303 is also used to extract multidimensional features of the operating parameter, including time series features, spatial distribution features, and the equipment aging index; the time series features reflect the changes of the operating parameter over time, and the spatial distribution features characterize the layout relationship between the equipment and the battery cluster; based on the multidimensional features, unsupervised clustering analysis is performed to screen out abnormal data that deviate from the standard.

[0172] Optionally, the second generation module 304 is further configured to determine the abnormal pattern of the operating parameter through unsupervised clustering analysis; and to construct a fault map based on the association rules in the fault rule base and the abnormal pattern, the fault map including the device status code layer, the operating parameter layer and the abnormal pattern layer.

[0173] Optionally, the second generation module 304 is also used to filter the fault propagation paths in the fault map based on the causal reasoning algorithm after generating the fault map, and exclude false association paths caused by environmental factors.

[0174] Optionally, the second generation module 304 is further configured to adjust the fault rule base based on a feedback reinforcement learning algorithm, which adjusts the weight of the association rule according to the verification result.

[0175] The fault location device based on association rules provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0176] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 4 As shown, the electronic device 400 provided in this embodiment includes at least one processor 401 and a memory 402. Optionally, the electronic device 400 further includes a communication component 403. The processor 401, memory 402, and communication component 403 are connected via a bus 404.

[0177] In a specific implementation, at least one processor 401 executes computer execution instructions stored in memory 402, causing at least one processor 401 to perform the above-described method.

[0178] The specific implementation process of processor 401 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0179] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0180] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0181] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0182] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0183] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0184] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0185] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0186] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0187] 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.

[0188] In addition, the functional units in the various embodiments of the present invention 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.

[0189] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part 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 of the various embodiments of this invention. 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.

[0190] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0191] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A fault location method based on association rules, characterized in that, include: Obtain the status codes, fault codes, and operating parameters of the energy storage thermal management equipment; The status code indicates the identifier of the current operating status of the device, the fault code indicates the identifier of the fault that has occurred in the energy storage thermal management device, and the operating parameters indicate the continuous parameters collected during the operation of the energy storage thermal management device. Based on the status code, the fault code, and the operating parameters, an association rule mining algorithm is used to generate a fault rule base. The fault rule base contains association rules between status codes, fault codes and operating parameters, and the association rule mining algorithm is used to discover implicit associations between data item sets. By performing unsupervised cluster analysis on the operating parameters, abnormal data that deviates from the standard are filtered out. The abnormal data is fused with the fault rule base to generate a fault map; Based on the fault map, a fault location report of the energy storage thermal management equipment is output. The fault location report includes at least the fault propagation path and the maintenance plan. The fault propagation path is the path of causal relationship between the fault code and the abnormal mode.

2. The method according to claim 1, characterized in that, The step of generating a fault rule base using an association rule mining algorithm based on the status code, the fault code, and the operating parameters includes: Based on support, confidence, and lift, association rules between the status code and the operating parameters are selected; the support represents the frequency of the simultaneous occurrence of the status code and the operating parameters, the confidence represents the probability that the operating parameters will occur simultaneously when the status code occurs, and the lift is used to measure the independence between the status code and the operating parameters. Based on the association rules and the association rule mining algorithm, a fault rule base is generated; The confidence threshold is dynamically adjusted based on the equipment aging index; the equipment aging index is a comprehensive indicator reflecting the operating status of the energy storage thermal management equipment. The fault rule base is updated based on the adjusted confidence threshold.

3. The method according to claim 2, characterized in that, After filtering the association rules between the status code and the operating parameters based on support, confidence, and lift, the method further includes: The confidence threshold is adjusted based on changes in the operating environment of the energy storage thermal management device; these changes include ambient temperature or load fluctuations. The weights of the association rules are dynamically adjusted based on historical accuracy; the historical accuracy represents historical data on the fault location accuracy of the association rules in the fault rule base. The fault rule base is updated based on the confidence threshold and the weight of the association rule.

4. The method according to claim 1, characterized in that, The step of performing unsupervised clustering analysis on the operating parameters to filter out outlier data that deviates from the standard includes: The operating parameters are extracted as multidimensional features, including time series features, spatial distribution features, and equipment aging indicators; the time series features reflect the changes of operating parameters over time, and the spatial distribution features characterize the layout relationship between the equipment and the battery cluster. Unsupervised clustering analysis is performed based on the aforementioned multidimensional features to filter out outlier data that deviate from the criteria.

5. The method according to claim 1, characterized in that, The step of fusing the abnormal data with the fault rule base to generate a fault map includes: Abnormal patterns of the operating parameters were determined through unsupervised clustering analysis; Based on the association rules in the fault rule base and the abnormal patterns, a fault map is constructed, which includes a device status code layer, an operating parameter layer and an abnormal pattern layer.

6. The method according to claim 5, characterized in that, After generating the fault map, the process also includes: The fault propagation paths in the fault map are filtered based on the causal reasoning algorithm to eliminate false association paths caused by environmental factors.

7. The method according to any one of claims 1 to 6, characterized in that, The method further includes: The fault rule base is adjusted based on a feedback reinforcement learning algorithm, and the weight of the association rule is adjusted according to the verification results.

8. A fault location device based on association rules, characterized in that, include: The acquisition module is used to acquire the status codes, fault codes, and operating parameters of the energy storage thermal management equipment. The status code indicates the identifier of the current operating status of the device, the fault code indicates the identifier of the fault that has occurred in the energy storage thermal management device, and the operating parameters indicate the continuous parameters collected during the operation of the energy storage thermal management device. The first generation module is used to generate a fault rule base based on the status code, the fault code and the operating parameters using an association rule mining algorithm; The fault rule base contains association rules between status codes, fault codes and operating parameters, and the association rule mining algorithm is used to discover implicit associations between data item sets. The filtering module is used to filter out abnormal data that deviates from the standard by performing unsupervised clustering analysis on the operating parameters; The second generation module is used to integrate the abnormal data with the fault rule base to generate a fault map; The output module is used to output a fault location report of the energy storage thermal management equipment based on the fault map. The fault location report includes at least the fault propagation path and the maintenance plan. The fault propagation path is the path of causal relationship between the fault code and the abnormal mode.

9. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-7.

11. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method described in any one of claims 1-7.

Images