An AI analysis-based charging pile fault backtracking method and system

Abstract

The application is suitable for the technical field of intelligent operation and maintenance of charging piles, and provides a charging pile fault backtracking method and system based on AI analysis, which comprises the following steps: when a target maintenance engineer is about to carry out maintenance on the target relay contact adhesion abnormity of a target charging device, the specific model and operation condition of the target charging device are determined, and the historical maintenance record of the target maintenance engineer is obtained.The technical scheme of the application can reduce the fault recurrence rate caused by the quality fluctuation of relay welding, make the operation stability of the repaired relay closer to the standard level of excellent maintenance cases, and reduce the virtual welding or excessive welding caused by unreasonable welding time setting.Through optimizing the welding process parameters, improving the maintenance consistency and reducing the equipment repair rate, the overall operation stability and reliability of the charging device are improved, the operation and maintenance cost is reduced, and the long-term use efficiency of the equipment is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent operation and maintenance of charging piles, and in particular relates to a charging pile fault backtracing method and system based on AI analysis. Background Art

[0002] Charging piles are a crucial component of new energy vehicle charging infrastructure. Relays, one of their core control components, are responsible for controlling the on / off state of circuits. However, over long periods of operation, relays are susceptible to current surges, environmental factors, and welding quality, making them prone to contact adhesion anomalies. This can prevent the relays from opening or closing properly, impacting the normal operation of the charging piles. Currently, charging pile maintenance typically relies on fault alarm systems to prompt engineers to conduct investigations. Based on experience, engineers determine whether the relays are faulty and take repair measures, such as replacing the relays or re-welding the relay contacts. However, due to the lack of accurate traceability analysis and standardized maintenance optimization methods during the maintenance process, different engineers may adopt different maintenance methods, resulting in significant differences in maintenance results. Some equipment may still experience the same or similar faults within a short period of time, increasing the equipment's operating and maintenance costs and difficulty in repair.

[0003] Existing technologies fail to establish a performance comparison model before and after maintenance during relay maintenance, resulting in a lack of data support for maintenance optimization. Furthermore, current maintenance methods cannot accurately trace the performance changes of relays under different maintenance strategies, making it difficult for maintenance engineers to effectively assess the impact of their maintenance actions on the long-term operational stability of the relay. Even though some optimization solutions incorporate maintenance data logging, it is limited to recording the status at the time of the fault and fails to effectively match appropriate historical comparison cases, making it impossible to provide a scientific and reasonable basis for maintenance adjustments. In this case, problems such as improper welding time settings and welding process deviations may still occur after the maintenance is completed, causing the relay to experience contact adhesion anomalies again in the short term, affecting the overall operational stability of the equipment. Summary of the Invention

[0004] The purpose of the present invention is to provide a charging pile fault backtracing method and system based on AI analysis, aiming to solve the problems raised in the background technology.

[0005] The present invention is implemented as follows: a charging pile fault backtracking method based on AI analysis, the method comprising:

[0006] When a target maintenance engineer intends to repair a target relay contact adhesion abnormality of a target charging device, determine the specific model and operating condition of the target charging device and obtain the target maintenance engineer's historical maintenance records;

[0007] determine whether the target maintenance engineer needs to perform a maintenance behavior comparison analysis based on historical maintenance records, and if so, filter out a specific maintenance case that matches the specific model and operating conditions of the target charging device;

[0008] Call the data storage system to obtain an optimal maintenance case that matches the specific model and operating conditions of the target charging device;

[0009] Extract the first performance change trend and the second performance change trend of the charging device after the maintenance time point in the specific maintenance case and the optimal maintenance case respectively, and generate a correction factor based on the deviation degree of the two;

[0010] Determine the welding process parameters of the target maintenance engineer when dealing with the target relay contact adhesion anomaly based on historical maintenance records, and modify the welding process parameters through the correction factor.

[0011] As a further limitation of the technical scheme of the embodiment of the application, the step of determining whether the target maintenance engineer needs to perform a maintenance behavior comparison analysis based on historical maintenance records, and if so, filtering out a specific maintenance case that matches the specific model and operating conditions of the target charging device, comprises:

[0012] Analyze historical maintenance records and extract a number of relay contact adhesion anomaly maintenance records of the same type as the target charging device handled by the target maintenance engineer;

[0013] Call the data storage system to count the proportion of maintenance records in the extracted number of maintenance records that have short-term recurrent faults, and determine whether the proportion exceeds a preset threshold;

[0014] If it is determined that the proportion exceeds the preset threshold, perform a maintenance behavior comparison analysis for the target maintenance engineer, and filter out a specific maintenance case that matches the specific model and operating conditions of the target charging device.

[0015] As a further limitation of the technical scheme of the embodiment of the application, the step of calling the data storage system to obtain an optimal maintenance case that matches the specific model and operating conditions of the target charging device, comprises:

[0016] Call the data storage system to retrieve historical maintenance records of other maintenance engineers, and count the proportion of short-term recurrent faults in the maintenance tasks handled by each maintenance engineer for the same type of charging device as the target charging device;

[0017] Filter out the best maintenance engineer with the lowest proportion of short-term recurrent faults;

[0018] In the historical maintenance records of the selected excellent maintenance engineers, find maintenance cases that match the specific model and operating conditions of the target charging equipment and have no short-term recurring failures, and select them as preferred maintenance cases.

[0019] As a further limitation of the technical solution of the embodiment of the present invention, the steps of respectively extracting the first performance change trend and the second performance change trend of the charging equipment after the maintenance time point in the specific maintenance case and the preferred maintenance case, and generating a correction factor based on the degree of deviation between the two include:

[0020] Based on the data storage system and specific maintenance cases, the relay on-off voltage data and relay coil pull-in current data of the target charging device within a preset time period after the maintenance time point are obtained, and the obtained data are integrated, arranged in time series, and plotted to obtain a first on-off voltage change curve and a first pull-in current change curve;

[0021] Based on the data storage system and the preferred maintenance case, the relay on-off voltage data and the relay coil pull-in current data of the charging device corresponding to the preferred maintenance case within a preset time period after the maintenance time point are obtained, and the obtained data are integrated, arranged in time series, and plotted to obtain a second on-off voltage change curve and a second pull-in current change curve;

[0022] A correction factor is generated based on the degree of deviation between the first on-off voltage variation curve and the second on-off voltage variation curve, and the degree of deviation between the first pickup current variation curve and the second pickup current variation curve.

[0023] As a further limitation of the technical solution of the embodiment of the present invention, the steps of determining the welding process parameters of the target maintenance engineer when handling the target relay contact adhesion abnormality based on the historical maintenance record, and correcting the welding process parameters using the correction factor include:

[0024] Analyze historical maintenance records to extract the welding process parameters used by target maintenance engineers to address contact adhesion anomalies on target relays;

[0025] Call the preset correction formula and correct the welding process parameters based on the correction factor;

[0026] The corrected welding process parameters are applied to the target maintenance engineer to perform the target relay contact adhesion abnormality repair work of the target charging equipment.

[0027] As a further limitation of the technical solution of the embodiment of the present invention, the preset correction formula is: ;

[0028] in Refers to the corrected welding process parameters, Refers to the uncorrected welding process parameters, Refers to the average slope of the first on-off voltage change curve, Refers to the average slope of the second on-off voltage change curve, Refers to the degree of deviation between the first on-off voltage change curve and the second on-off voltage change curve, Refers to the adjustment weight of the deviation degree between the first on-off voltage change curve and the second on-off voltage change curve, Refers to the average slope of the first pickup current change curve, Refers to the average slope of the second pickup current change curve, Refers to the degree of deviation between the first pickup current change curve and the second pickup current change curve. Refers to the adjustment weight of the deviation between the first pickup current variation curve and the second pickup current variation curve.

[0029] A charging pile fault backtracking system based on AI analysis, comprising: a data acquisition module, a specific maintenance case acquisition module, a preferred maintenance case acquisition module, a correction factor generation module, and a parameter correction module, wherein:

[0030] A data acquisition module is used to determine the specific model and operating condition of the target charging device and obtain the historical maintenance records of the target maintenance engineer when the target maintenance engineer intends to repair the target relay contact adhesion abnormality of the target charging device;

[0031] A specific maintenance case acquisition module is used to determine whether a maintenance behavior comparison analysis is needed for the target maintenance engineer based on historical maintenance records. If so, it selects specific maintenance cases that match the specific model and operating conditions of the target charging equipment.

[0032] The preferred maintenance case acquisition module is used to call the data storage system to obtain the preferred maintenance case that matches the specific model and operating conditions of the target charging equipment;

[0033] A correction factor generation module is used to extract the first performance change trend and the second performance change trend of the charging equipment after the maintenance time point in the specific maintenance case and the preferred maintenance case, respectively, and generate a correction factor based on the degree of deviation between the two;

[0034] The parameter correction module is used to determine the welding process parameters of the target maintenance engineer when handling the target relay contact adhesion anomaly based on historical maintenance records, and to correct the welding process parameters through correction factors.

[0035] As a further limitation of the technical solution of the embodiment of the present invention, the specific maintenance case acquisition module specifically includes:

[0036] A record parsing unit is used to parse historical maintenance records and extract maintenance records of several relay contact adhesion anomalies of the same type as the target charging equipment handled by the target maintenance engineer;

[0037] A proportion judgment unit is used to call the data storage system, count the proportion of maintenance records with short-term recurring faults in the extracted maintenance records, and judge whether the proportion exceeds a preset threshold;

[0038] The specific maintenance case determination unit is used to perform a comparative analysis of the maintenance behavior of the target maintenance engineer if it is determined that the proportion exceeds a preset threshold, and to screen out specific maintenance cases that match the specific model and operating conditions of the target charging equipment.

[0039] As a further limitation of the technical solution of the embodiment of the present invention, the preferred maintenance case acquisition module specifically includes:

[0040] A percentage statistics unit is used to call the data storage system, retrieve the historical maintenance records of other maintenance engineers, and calculate the percentage of short-term recurring faults handled by each maintenance engineer when handling maintenance tasks of the same type as the target charging equipment;

[0041] Excellent maintenance engineer identification unit, used to screen out excellent maintenance engineers with the lowest proportion of short-term recurring failures;

[0042] The preferred maintenance case determination unit is used to search for maintenance cases that match the specific model and operating conditions of the target charging equipment and have no short-term recurring faults in the historical maintenance records of the selected excellent maintenance engineers, and use them as preferred maintenance cases.

[0043] As a further limitation of the technical solution of the embodiment of the present invention, the correction factor generation module specifically includes:

[0044] A first curve drawing unit is configured to obtain relay on / off voltage data and relay coil pull-in current data of a target charging device within a preset time period after a maintenance time point based on a data storage system and a specific maintenance case, integrate the obtained data, arrange them in a time series, and draw a first on / off voltage change curve and a first pull-in current change curve;

[0045] A second curve drawing unit is configured to obtain, based on the data storage system and the preferred maintenance case, relay on / off voltage data and relay coil pull-in current data of the charging device corresponding to the preferred maintenance case within a preset time period after the maintenance time point, integrate the obtained data, arrange them in time series, and draw a second on / off voltage change curve and a second pull-in current change curve;

[0046] The correction factor generating unit is used to generate a correction factor based on the deviation degree between the first on-off voltage change curve and the second on-off voltage change curve, and the deviation degree between the first pickup current change curve and the second pickup current change curve.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] The present invention can effectively improve the repair quality and reliability of abnormal relay contact adhesion through a charging pile fault backtracking method based on AI analysis. By extracting historical maintenance records, analyzing the maintenance methods of target maintenance engineers, and combining the proportion of short-term recurring failures, it is ensured that the optimization of maintenance behavior has data support. Compared with the traditional maintenance method that relies on experience and judgment, the present invention extracts the on-off voltage change curve and the pull-in current change curve of the equipment after maintenance through the data storage system, compares them with excellent maintenance cases, calculates the degree of deviation, and corrects the welding process parameters based on the correction factor, making the maintenance process more accurate and standardized.

[0049] The technical solution of this invention can reduce the recurrence rate of faults caused by fluctuations in relay welding quality, bringing the operational stability of repaired relays closer to the standard of excellent repair cases, and reducing the occurrence of cold or excessive welds caused by improper welding time settings. By optimizing welding process parameters, repair consistency is improved, and the equipment return rate is reduced, thereby improving the overall operational stability and reliability of charging equipment, reducing operation and maintenance costs, and improving the long-term efficiency of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A flowchart of a method provided by an embodiment of the present invention;

[0051] Figure 2 A flowchart of screening specific maintenance cases in the method provided in an embodiment of the present invention;

[0052] Figure 3 A flowchart of matching preferred maintenance cases in the method provided in an embodiment of the present invention;

[0053] Figure 4 A flow chart of generating a correction factor in the method provided in an embodiment of the present invention;

[0054] Figure 5 A flow chart of correcting welding process parameters in the method provided in an embodiment of the present invention;

[0055] Figure 6 An application architecture diagram of the system provided by an embodiment of the present invention;

[0056] Figure 7 A structural block diagram of a specific maintenance case acquisition module in a system provided by an embodiment of the present invention;

[0057] Figure 8 A structural block diagram of a preferred maintenance case acquisition module in a system provided by an embodiment of the present invention;

[0058] Figure 9 This is a structural block diagram of a correction factor generation module in a system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0060] Figure 1 A flow chart of a method provided by an embodiment of the present invention is shown.

[0061] Specifically, a charging pile fault backtracking method based on AI analysis includes the following steps:

[0062] Step S100 , when a target maintenance engineer intends to repair a target relay contact adhesion abnormality of a target charging device, the specific model and operating condition of the target charging device are determined, and a historical maintenance record of the target maintenance engineer is obtained.

[0063] In an embodiment of the present invention, the target charging device refers to a charging device with a relay control function, including but not limited to AC charging piles, DC charging piles, supercharging stations, etc., which are mainly used to provide power transmission for electric vehicles. The target relay contact adhesion abnormality refers to the internal contacts of the relay failing to close or open normally under the action of the on-off control signal due to long-term use, welding quality, load impact, arc erosion, etc. during operation, thereby affecting the normal operation of the charging device. Usually, the maintenance methods for relay contact adhesion abnormality include replacing the relay, re-welding the relay contacts, adjusting the contact pressure, cleaning the oxide layer or foreign matter on the contact surface, etc. Among them, welding maintenance is a more common method. Welding process parameters such as welding time, welding temperature, and solder thickness directly affect the stability and service life of the relay after maintenance.

[0064] Determining the specific model and operating conditions of the target charging device is a crucial step in performing accurate fault tracing. The specific model is typically obtained through information on the device nameplate, records in the device management system, or the factory serial number. Operating conditions refer to the state of the target charging device during actual operation, primarily including but not limited to parameters such as cumulative usage time, load conditions, grid voltage fluctuations, ambient temperature and humidity, equipment maintenance frequency, historical load profiles of the charging device, and average charging power. These factors directly impact the relay's operating state, determining its reliability and potential failure modes. Determining operating conditions generally relies on information such as internal device sensor monitoring data, remote data collection from the charging management system, and historical device operation logs.

[0065] The data storage system is the core data management module of the present invention, containing operation records, maintenance records, and equipment maintenance information for charging equipment. Specifically, the data storage system includes the following types of data: First, equipment operation logs, which record the charging equipment's on-off voltage data, relay pull-in current, voltage and current fluctuations during the charging process, and environmental parameters; Second, equipment fault logs, which record fault information generated during equipment operation, including fault code, trigger time, affected components, and equipment status before the fault occurred; Third, equipment maintenance logs, which record historical equipment maintenance, such as contact cleaning, relay replacement, and welding repair operations; Fourth, maintenance engineer inspection logs, which record the specific maintenance actions of each maintenance engineer on different charging equipment, including repair time, repair method, welding parameters, and tools used; Fifth, post-repair equipment operating status logs, which record the equipment's operating status after repair, including the relay on-off voltage curve and relay pull-in current change curve after repair, to ensure that the equipment's status after repair can be monitored; Sixth, equipment short-term recurrence fault statistics, which record whether the same or similar fault recurs within a specific time window after repair, and calculate the short-term recurrence probability.

[0066] Historical maintenance records are part of the data storage system and primarily contain information about maintenance engineers' maintenance processes on various charging devices. This includes information such as repair time, device model, fault type, repair method, repair parameters (including welding process parameters such as welding time, temperature, solder thickness, contact cleaning method, and contact pressure adjustment), and repair tools. These historical maintenance records provide traceability of maintenance activities. Combined with device operation logs, they can analyze the impact of different maintenance methods on the long-term operation of the device, providing data support for subsequent maintenance optimization.

[0067] Furthermore, the charging pile fault backtracking method based on AI analysis also includes the following steps:

[0068] In step S200 , based on historical maintenance records, it is determined whether it is necessary to perform maintenance behavior comparison analysis for the target maintenance engineer. If it is determined to be necessary, specific maintenance cases that match the specific model and operating conditions of the target charging equipment are screened out.

[0069] Specifically, Figure 2 A flow chart for screening specific repair cases is shown.

[0070] The following steps are used to determine whether a maintenance behavior comparison analysis is needed for the target maintenance engineer based on historical maintenance records. If necessary, specific maintenance cases that match the specific model and operating conditions of the target charging equipment are selected:

[0071] Step S201: parsing historical maintenance records to extract several relay contact adhesion abnormality maintenance records of the same type as the target charging equipment handled by the target maintenance engineer;

[0072] Step S202: calling the data storage system to count the proportion of maintenance records with short-term recurring faults in the extracted maintenance records, and determining whether the proportion exceeds a preset threshold;

[0073] In step S203 , if it is determined that the proportion exceeds the preset threshold, a comparative analysis of the maintenance behavior of the target maintenance engineer is performed, and specific maintenance cases that match the specific model and operating conditions of the target charging equipment are screened out.

[0074] In an embodiment of the present invention, the process of parsing historical maintenance records and extracting several maintenance records of relay contact adhesion anomalies of the same type as the target charging device, which have been handled by the target maintenance engineer, is mainly based on the historical maintenance record data stored in the data storage system. First, based on the unique identifier of the target maintenance engineer, the system retrieves all maintenance records that the engineer has performed in historical maintenance tasks. Subsequently, the maintenance cases involving relay contact adhesion anomalies are screened out, and the equipment types recorded in the maintenance cases are further compared, and only maintenance records of the same type as the target charging device are retained. This screening process can adopt a query method based on information such as equipment model, operating parameters, fault code, etc. to ensure that only maintenance cases with high relevance are extracted.

[0075] After calling the data storage system, the percentage of maintenance records with short-term recurring faults in the extracted maintenance records is counted, and analysis is required based on the monitoring data of the equipment operating status after maintenance. Short-term recurring faults refer to situations where the same or similar faults occur on the same equipment within a specific time window after the maintenance is completed (for example, 30 days or 50 working cycles). By comparing the equipment operation logs after maintenance, the system analyzes key performance data such as the relay on-off voltage curve and the pull-in current curve after the maintenance is completed, and determines whether there are abnormal fluctuations or deviations from the normal operating range. Combined with the equipment fault alarm information after maintenance, statistics are made to see whether the equipment triggers the same fault alarm again in the short term. If the same maintenance engineer has a high proportion of short-term recurring faults in similar maintenance cases, it indicates that his maintenance method may be unstable or some operations do not meet the optimal maintenance standards.

[0076] The preset threshold is primarily based on industry experience, equipment design standards, and statistical analysis of actual operational data. Generally speaking, a reasonable threshold for the short-term recurrence rate can be determined by a number of factors, including the normal operating life of the equipment, the natural recurrence rate of common faults, and the distribution of repair success rates among different engineers in historical data. For example, if industry experience shows that the short-term recurrence rate of relay contact sticking anomalies is typically less than 10% after normal maintenance, the threshold can be set between 10% and 15%. If the proportion of short-term recurring faults for a maintenance engineer exceeds this threshold, it indicates that their maintenance behavior may have room for improvement.

[0077] When the proportion of short-term recurring faults exceeds a preset threshold, a comparative analysis of the maintenance behavior of targeted maintenance engineers is performed to optimize repair quality and reduce recurrence rates. If a particular engineer's maintenance approach results in a high recurrence rate, this may indicate room for improvement in welding process parameters, repair procedures, tool usage, and other areas. Therefore, further analysis of their maintenance behavior is necessary, and comparison with repair cases with low recurrence rates is needed to identify potential causes of poor repair results, such as short welding time or insufficient solder thickness. By analyzing these differences, engineers can be provided with more precise repair parameter recommendations, optimizing maintenance strategies and improving repair quality.

[0078] The purpose of selecting specific maintenance cases is to provide accurate comparative references to ensure targeted maintenance optimization. When selecting specific maintenance cases, ensure they match the specific model and operating conditions of the target charging equipment to avoid deviations in maintenance strategies due to differences in equipment type, operating environment, and load conditions. Specific maintenance cases can provide target maintenance engineers with maintenance experience that is more consistent with the current equipment operating conditions, making optimized maintenance parameters more applicable and improving the accuracy and consistency of maintenance work.

[0079] Furthermore, the charging pile fault backtracking method based on AI analysis also includes the following steps:

[0080] Step S300: calling a data storage system to obtain a preferred maintenance case that matches the specific model and operating conditions of the target charging device.

[0081] Specifically, Figure 3 A flow chart for matching preferred maintenance cases is shown.

[0082] The steps of calling the data storage system to obtain the preferred maintenance case that matches the specific model and operating conditions of the target charging equipment specifically include the following steps:

[0083] Step S301: Call the data storage system to retrieve the historical maintenance records of other maintenance engineers and calculate the proportion of short-term recurring faults in the maintenance tasks of the same type as the target charging equipment handled by each maintenance engineer;

[0084] Step S302: Screen out outstanding maintenance engineers with the lowest proportion of short-term recurring failures;

[0085] Step S303 : Search the historical maintenance records of the selected excellent maintenance engineers for maintenance cases that match the specific model and operating conditions of the target charging equipment and have no short-term recurring faults, and select them as preferred maintenance cases.

[0086] In this embodiment of the present invention, a data storage system is called to retrieve the historical maintenance records of other maintenance engineers and to calculate the percentage of short-term recurring faults encountered by each maintenance engineer during maintenance tasks for the same type of target charging equipment. This requires first establishing a historical maintenance data analysis model based on the data storage system. This model filters all maintenance cases matching the target charging equipment type and classifies the data according to the maintenance engineer's unique identity, extracting all maintenance records for each maintenance engineer for the same equipment type. Subsequently, from these maintenance records, maintenance cases involving relay contact sticking anomalies are screened, and the number of such maintenance tasks performed by each maintenance engineer is counted.

[0087] When calculating the percentage of short-term recurring failures for each maintenance engineer handling this type of repair, it is necessary to analyze the equipment's short-term operational status after the repair, combining the equipment's operation logs and post-repair equipment status records stored in the data system. If the equipment experiences the same or similar relay contact sticking anomaly within a specific time window (such as 30 days or 50 charging cycles), the repair case is considered a short-term recurring failure. For each maintenance engineer, the proportion of short-term recurring failures across all their repair cases is calculated to determine the short-term recurring failure percentage for each engineer. A recurring failure percentage distribution table is then created. This distribution table is used to subsequently screen outstanding maintenance engineers.

[0088] When selecting outstanding maintenance engineers with the lowest percentage of short-term recurrent failures, it is necessary to ensure that the calculation of this percentage is based on a sufficient number of maintenance cases, that is, a certain threshold number of maintenance cases must be met. For example, if a maintenance engineer has only performed a small number of maintenance tasks, even if his or her percentage of short-term recurrent failures is low, the statistical results may not be representative due to insufficient sample size. Therefore, a minimum threshold number of maintenance cases needs to be set during the screening process. For example, it is required that at least 10 repairs for abnormal relay contact adhesion have been performed to have sufficient statistical significance. Finally, the engineers with the lowest short-term recurrence rate and the number of maintenance cases that meet the preset standards are selected as outstanding maintenance engineers by sorting them from low to high according to the percentage of short-term recurrent failures.

[0089] In the historical maintenance records of the selected excellent maintenance engineers, we search for maintenance cases that match the specific model and operating conditions of the target charging equipment and have not experienced short-term recurring failures, mainly to ensure the applicability and reference value of the selected preferred maintenance cases. When matching specific models, the system needs to compare the equipment model, equipment series, control system version and other information in the maintenance records to ensure that the equipment corresponding to the preferred maintenance case has the same hardware foundation as the target charging equipment. When matching operating conditions, it is necessary to consider the external environment, load conditions, power grid fluctuations, etc. of the equipment operation to ensure that the maintenance conditions of the reference case are similar to the actual conditions of the target equipment.

[0090] The preferred maintenance cases are selected based on the fact that they come from engineers with the lowest short-term recurring failure rate, and the corresponding charging equipment has no short-term recurring failures after repair. The purpose of this selection is to provide a reliable maintenance reference template, serving as a standard for optimizing the maintenance strategy of the target maintenance engineer. By analyzing the maintenance methods, welding process parameters, and repair steps used in the preferred maintenance cases, the target maintenance engineer can be guided to adjust the maintenance operation, reduce the probability of short-term recurring failures of the target charging equipment, and improve maintenance quality and equipment operational stability.

[0091] Furthermore, the charging pile fault backtracking method based on AI analysis also includes the following steps:

[0092] Step S400 : extracting the first performance change trend and the second performance change trend of the charging equipment after the maintenance time point in the specific maintenance case and the preferred maintenance case respectively, and generating a correction factor based on the degree of deviation between the two.

[0093] Specifically, Figure 4 A flow chart for generating correction factors is shown.

[0094] The steps of extracting the first performance change trend and the second performance change trend of the charging equipment after the maintenance time point in the specific maintenance case and the preferred maintenance case, and generating a correction factor based on the degree of deviation between the two specifically include the following steps:

[0095] Step S401: Based on the data storage system and a specific maintenance case, relay on / off voltage data and relay coil pull-in current data of the target charging device within a preset time period after the maintenance time point are obtained, and the obtained data are integrated, arranged in time series, and plotted to obtain a first on / off voltage change curve and a first pull-in current change curve;

[0096] Step S402: Based on the data storage system and the preferred maintenance case, relay on / off voltage data and relay coil pull-in current data of the charging device corresponding to the preferred maintenance case within a preset time period after the maintenance time point are obtained, and the obtained data are integrated, arranged in time series, and plotted to obtain a second on / off voltage change curve and a second pull-in current change curve;

[0097] Step S403 : generating a correction factor based on the degree of deviation between the first on-off voltage variation curve and the second on-off voltage variation curve, and the degree of deviation between the first pickup current variation curve and the second pickup current variation curve.

[0098] In the embodiment of the present invention, the reason for selecting relay on-off voltage data and relay coil pull-in current data as reference indicators is that these two data can directly reflect the working status of the relay and its performance changes. The relay on-off voltage data is mainly used to measure the voltage change of the relay when the switch state is switched, and can characterize problems such as contact adhesion, abnormal contact resistance, and arcing. The relay coil pull-in current data is used to analyze the current response of the coil during the power-on and power-off process of the relay, reflecting problems such as coil aging, hysteresis effect, and excitation instability. The combination of these two data can more comprehensively evaluate the health status of the relay and analyze whether the performance of the repaired relay meets the standards through historical comparison.

[0099] In terms of data quantification, the relay on-off voltage data and relay coil pull-in current data at each designated time point can be recorded by the device's built-in sensors or monitoring system. Specifically, the charging device's control system selects representative data at specific time points during the relay switching process to ensure the accuracy and simplicity of data quantification. For relay on-off voltage data, the system can extract the voltage value at the moment when the relay contacts are fully closed or open as representative data to avoid interference from signal jitter or transition states. For relay coil pull-in current data, the system can extract the steady-state current value at the critical time point when the relay is fully closed or released to reflect the working status of the coil.

[0100] The selection of representative data can be based on preset rules, such as selecting the stable on-off voltage value after the relay is powered on, or selecting the maximum current value reaching the steady state during the relay attraction process. In this way, the data at each time point can be represented as: , where V is the on-off voltage value at this time point, is the attraction current value at this time point, only the representative data at this time point is recorded, rather than the continuous sampling data of the entire on-off process. This way can reduce redundant data and improve computing efficiency, while ensuring that the voltage and current values used for analysis can accurately reflect the operating state of the relay.

[0101] The generation of the change curve depends on the historical operation data provided by the data storage system, and combines signal processing technology to smooth the original data. Specifically, data interpolation, filtering methods (such as Kalman filtering, wavelet transform, etc.) are used to remove noise, so that the discrete data points collected can form a smoother change curve. Finally, time series analysis methods are used to draw the relay on-off voltage change curve and the attraction current change curve. For example, visualization curves can be generated by tools such as Matplotlib or Matlab, or machine learning algorithms can be used to analyze the curve features, thereby extracting key change patterns.

[0102] The deviation between the first on-off voltage change curve and the second on-off voltage change curve, and the deviation between the first attraction current change curve and the second attraction current change curve are used as the reason for the correction factor, because these two groups of data can quantify the gap between the state of the repaired relay and the standard repair case. The calculation method of the deviation degree can be based on the slope change rate or the error sum of squares, etc. statistical methods to quantify the performance deviation of the repaired relay. If the deviation is large, it means that the operating state of the repaired relay is significantly different from the best repair case, and the repair process parameters need to be corrected. By setting the correction factor in this way, it can ensure that the repair optimization process has data support, and avoid relying on experience to judge the repair quality, thereby improving the scientificity and reliability of the repair strategy.

[0103] Further, the charging pile fault backtracking method based on AI analysis further includes the following steps:

[0104] Step S500, based on the historical maintenance record, determine the welding process parameters of the target maintenance engineer when dealing with the target relay contact adhesion anomaly, and correct the welding process parameters through the correction factor.

[0105] Specifically, Figure 5 A flowchart of correcting the welding process parameters is shown.

[0106] The following steps are used to determine the welding process parameters of the target maintenance engineer when handling the target relay contact adhesion anomaly based on historical maintenance records and to correct the welding process parameters using correction factors:

[0107] Step S501: parsing historical maintenance records to extract welding process parameters used by a target maintenance engineer to handle contact adhesion anomalies of a target relay;

[0108] Step S502: calling a preset correction formula and correcting the welding process parameters based on the correction factor;

[0109] Step S503 : applying the corrected welding process parameters to a target maintenance engineer to perform a target relay contact adhesion abnormality repair operation on a target charging device.

[0110] The preset correction formula is: ;

[0111] in Refers to the corrected welding process parameters, Refers to the uncorrected welding process parameters, Refers to the average slope of the first on-off voltage change curve, Refers to the average slope of the second on-off voltage change curve, Refers to the degree of deviation between the first on-off voltage change curve and the second on-off voltage change curve, Refers to the adjustment weight of the deviation degree between the first on-off voltage change curve and the second on-off voltage change curve, Refers to the average slope of the first pickup current change curve, Refers to the average slope of the second pickup current change curve, Refers to the degree of deviation between the first pickup current change curve and the second pickup current change curve. Refers to the adjustment weight of the deviation between the first pickup current variation curve and the second pickup current variation curve.

[0112] The above formula uses an intuitive and simple calculation method. It calculates the deviation of the relay's on-off voltage curve and the pickup current curve after repair from the standard repair case, then performs a weighted combination to adjust the welding process parameters. This calculation method clearly quantifies the difference in relay performance before and after repair and uses a mathematical model to optimize and adjust welding process parameters, thereby improving the stability and consistency of repairs.

[0113] In addition to this formula, other calculation methods can also be used. For example, methods based on time series analysis can be used to calculate the slope change rate through curve fitting to determine trend deviations in different time periods; machine learning regression models such as multivariate linear regression, random forest regression, or neural networks can be used to train models using historical maintenance data to predict the optimal welding time; and methods based on Bayesian reasoning can be used to introduce probability analysis and dynamically adjust welding parameters based on the success rate of different maintenance cases to improve the reliability of maintenance results. In addition, fuzzy logic control can be combined to consider the fuzzy weight relationship of multiple maintenance parameters to construct a dynamic optimization algorithm that is more in line with engineering practice, making the correction of welding process parameters more accurate and intelligent.

[0114] In this embodiment of the present invention, historical maintenance records are analyzed to extract the welding process parameters used by the target maintenance engineer when addressing the target relay contact adhesion anomaly. Welding time is preferentially selected as the correction parameter. Welding time refers to the duration that the welding equipment maintains a molten state while applying heat during the welding operation. This parameter directly determines the quality of the solder joint, the degree of solder diffusion, and the stability of the welding contact surface.

[0115] Welding time was chosen as the parameter to be corrected because it has the most direct and significant impact on welding quality. Welding time that is too short may prevent the solder joint from fully melting, resulting in a cold joint or poor contact. This can easily lead to high contact resistance or abnormal heating of the relay contacts during operation, increasing the risk of failure. Welding time that is too long may cause excessive solder flow, resulting in a short circuit and even affecting the stability of the relay's internal mechanical structure. Furthermore, welding time is easily quantifiable and can be statistically analyzed through historical maintenance records. Combined with correction factor optimization, this allows for standardized maintenance operations, improving the reliability and service life of the relay after repair.

[0116] The welding time obtained through correction factors can be dynamically optimized based on the operating status of the target relay, ensuring that it meets the optimal welding process parameters for standard repair cases while also adapting to the specific operating conditions of different equipment. Compared to welding times set using fixed empirical values, the corrected welding time can reduce the probability of short-term recurrence of failures and bring the operating status of the repaired relay closer to the stable level of an excellent repair case. Furthermore, the optimized welding time can reduce repair defects caused by excessive or insufficient welding, improve maintenance consistency, and reduce the equipment return rate due to welding quality issues, thereby improving the overall operational stability and reliability of the charging equipment.

[0117] Further, Figure 6 The application architecture diagram of the system provided by the embodiment of the present invention is shown.

[0118] Among them, in another preferred embodiment provided by the present invention, a charging pile fault backtracking system based on AI analysis includes:

[0119] The data acquisition module 100 is used to determine the specific model and operating condition of the target charging device and obtain the historical maintenance record of the target maintenance engineer when the target maintenance engineer intends to repair the target relay contact adhesion abnormality of the target charging device.

[0120] Furthermore, the charging pile fault backtracking system based on AI analysis also includes:

[0121] The specific maintenance case acquisition module 200 is used to determine whether it is necessary to perform maintenance behavior comparison analysis for the target maintenance engineer based on historical maintenance records. If it is determined to be necessary, specific maintenance cases that match the specific model and operating conditions of the target charging equipment are screened out.

[0122] Specifically, Figure 7 It shows a structural block diagram of the specific maintenance case acquisition module 200 in the system provided by an embodiment of the present invention.

[0123] In a preferred embodiment of the present invention, the specific maintenance case acquisition module 200 specifically includes:

[0124] The record parsing unit 201 is used to parse historical maintenance records and extract several relay contact adhesion abnormality maintenance records of the same type as the target charging equipment handled by the target maintenance engineer;

[0125] The proportion judgment unit 202 is used to call the data storage system, count the proportion of maintenance records with short-term recurring faults in the extracted maintenance records, and judge whether the proportion exceeds a preset threshold;

[0126] The specific maintenance case determination unit 203 is used to perform a comparative analysis of the maintenance behavior of the target maintenance engineer if it is determined that the proportion exceeds a preset threshold, and to screen out specific maintenance cases that match the specific model and operating conditions of the target charging equipment.

[0127] Furthermore, the charging pile fault backtracking system based on AI analysis also includes:

[0128] The preferred maintenance case acquisition module 300 is used to call the data storage system to obtain the preferred maintenance case that matches the specific model and operating conditions of the target charging equipment.

[0129] Specifically, Figure 8 FIG. 4 shows a structural block diagram of the correction factor generation module 300 in the system provided by an embodiment of the present invention.

[0130] In a preferred embodiment of the present invention, the correction factor generation module 300 specifically includes:

[0131] The proportion statistics unit 301 is used to call the data storage system to retrieve the historical maintenance records of other maintenance engineers and calculate the proportion of short-term recurring faults in the maintenance tasks of the same type as the target charging equipment handled by each maintenance engineer;

[0132] The excellent maintenance engineer determination unit 302 is used to select excellent maintenance engineers with the lowest proportion of short-term recurring faults;

[0133] The preferred maintenance case determination unit 303 is used to search the historical maintenance records of the selected excellent maintenance engineers for maintenance cases that match the specific model and operating conditions of the target charging equipment and have no short-term recurring faults, and use them as preferred maintenance cases.

[0134] Furthermore, the charging pile fault backtracking system based on AI analysis also includes:

[0135] The correction factor generation module 400 is used to extract the first performance change trend and the second performance change trend of the charging equipment after the maintenance time point in the specific maintenance case and the preferred maintenance case, respectively, and generate a correction factor based on the degree of deviation between the two.

[0136] Specifically, Figure 9 FIG. 4 is a block diagram showing a structure of a correction factor generation module 400 in a system provided by an embodiment of the present invention.

[0137] In a preferred embodiment of the present invention, the correction factor generation module 400 specifically includes:

[0138] A first curve drawing unit 401 is configured to obtain relay on / off voltage data and relay coil pull-in current data of a target charging device within a preset time period after a maintenance time point based on a data storage system and a specific maintenance case, integrate the obtained data, arrange them in a time series, and draw a first on / off voltage change curve and a first pull-in current change curve;

[0139] The second curve drawing unit 402 is configured to obtain, based on the data storage system and the preferred maintenance case, relay on / off voltage data and relay coil pull-in current data of the charging device corresponding to the preferred maintenance case within a preset time period after the maintenance time point, integrate the obtained data, arrange them in time series, and draw a second on / off voltage change curve and a second pull-in current change curve;

[0140] The correction factor generating unit 403 is configured to generate a correction factor based on the deviation between the first on-off voltage variation curve and the second on-off voltage variation curve, and the deviation between the first pickup current variation curve and the second pickup current variation curve.

[0141] Furthermore, the charging pile fault backtracking system based on AI analysis also includes:

[0142] The parameter correction module 500 is used to determine the welding process parameters of the target maintenance engineer when handling the target relay contact adhesion abnormality based on the historical maintenance records, and to correct the welding process parameters through the correction factor.

[0143] It should be understood that, although the various steps in the flow chart of each embodiment of the present invention are shown in sequence according to the indication of the arrows, these steps are not necessarily performed in sequence according to the order indicated by the arrows. Unless otherwise specified herein, the execution of these steps is not strictly limited in order, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

Claims

1. A charging pile fault backtracking method based on AI analysis, characterized in that: The method comprises: When a target maintenance engineer intends to repair a target relay contact adhesion abnormality of a target charging device, determine the specific model and operating condition of the target charging device and obtain the target maintenance engineer's historical maintenance records; Based on historical maintenance records, a comparative analysis of maintenance behavior is performed to determine whether a target maintenance engineer needs to perform maintenance. If necessary, specific maintenance cases that match the specific model and operating conditions of the target charging equipment are selected. Call the data storage system to obtain the preferred maintenance case that matches the specific model and operating conditions of the target charging equipment; Extracting the first performance change trend and the second performance change trend of the charging equipment after the maintenance time point in the specific maintenance case and the preferred maintenance case, respectively, and generating a correction factor based on the degree of deviation between the two; The steps of respectively extracting a first performance change trend and a second performance change trend of the charging equipment after the maintenance time point in the specific maintenance case and the preferred maintenance case, and generating a correction factor based on the degree of deviation between the two include: Based on the data storage system and specific maintenance cases, the relay on-off voltage data and relay coil pull-in current data of the target charging device within a preset time period after the maintenance time point are obtained, and the obtained data are integrated, arranged in time series, and plotted to obtain a first on-off voltage change curve and a first pull-in current change curve; Based on the data storage system and the preferred maintenance case, the relay on-off voltage data and the relay coil pull-in current data of the charging device corresponding to the preferred maintenance case within a preset time period after the maintenance time point are obtained, and the obtained data are integrated, arranged in time series, and plotted to obtain a second on-off voltage change curve and a second pull-in current change curve; generating a correction factor based on a degree of deviation between the first on-off voltage variation curve and the second on-off voltage variation curve, and a degree of deviation between the first pickup current variation curve and the second pickup current variation curve; Based on historical maintenance records, the welding process parameters of the target maintenance engineer when dealing with the target relay contact adhesion anomaly are determined, and the welding process parameters are corrected using correction factors.

2. The charging pile fault backtracking method based on AI analysis according to claim 1 is characterized in that: Based on historical maintenance records, it is determined whether a maintenance behavior comparison analysis is required for the target maintenance engineer. If so, the steps for selecting specific maintenance cases that match the specific model and operating conditions of the target charging equipment include: Parse historical maintenance records and extract maintenance records of several relay contact adhesion anomalies handled by target maintenance engineers for the same type of target charging equipment. Calling the data storage system to count the proportion of maintenance records with short-term recurring faults among the extracted maintenance records, and determining whether the proportion exceeds a preset threshold; If it is determined that the proportion exceeds the preset threshold, a comparative analysis of the maintenance behavior of the target maintenance engineers will be performed, and specific maintenance cases that match the specific model and operating conditions of the target charging equipment will be screened out.

3. The charging pile fault backtracking method based on AI analysis according to claim 2 is characterized in that: The steps of calling the data storage system to obtain the preferred maintenance case that matches the specific model and operating conditions of the target charging equipment include: Call the data storage system to retrieve the historical maintenance records of other maintenance engineers and calculate the proportion of short-term recurring faults in each maintenance engineer's maintenance tasks for the same type of target charging equipment; Screen out outstanding maintenance engineers with the lowest proportion of short-term recurring failures; In the historical maintenance records of the selected excellent maintenance engineers, find maintenance cases that match the specific model and operating conditions of the target charging equipment and have no short-term recurring failures, and select them as preferred maintenance cases.

4. The charging pile fault backtracking method based on AI analysis according to claim 3 is characterized in that: The steps of determining the welding process parameters of the target maintenance engineer when handling the target relay contact adhesion abnormality based on the historical maintenance records, and correcting the welding process parameters using the correction factor include: Analyze historical maintenance records to extract the welding process parameters used by target maintenance engineers to address contact adhesion anomalies on target relays; Call the preset correction formula and correct the welding process parameters based on the correction factor; The corrected welding process parameters are applied to the target maintenance engineer to perform the target relay contact adhesion abnormality repair work of the target charging equipment.

5. The charging pile fault backtracking method based on AI analysis according to claim 4 is characterized in that: The preset correction formula is: ; in Refers to the corrected welding process parameters, Refers to the uncorrected welding process parameters, Refers to the average slope of the first on-off voltage change curve, Refers to the average slope of the second on-off voltage change curve, Refers to the degree of deviation between the first on-off voltage change curve and the second on-off voltage change curve, Refers to the adjustment weight of the deviation degree between the first on-off voltage change curve and the second on-off voltage change curve, Refers to the average slope of the first pickup current change curve, Refers to the average slope of the second pickup current change curve, Refers to the degree of deviation between the first pickup current change curve and the second pickup current change curve. Refers to the adjustment weight of the deviation between the first pickup current variation curve and the second pickup current variation curve.

6. A charging pile fault backtracking system based on AI analysis, characterized in that: The system includes: a data acquisition module, a specific maintenance case acquisition module, a preferred maintenance case acquisition module, a correction factor generation module and a parameter correction module, wherein: A data acquisition module is used to determine the specific model and operating condition of the target charging device and obtain the historical maintenance records of the target maintenance engineer when the target maintenance engineer intends to repair the target relay contact adhesion abnormality of the target charging device; A specific maintenance case acquisition module is used to determine whether a maintenance behavior comparison analysis is needed for the target maintenance engineer based on historical maintenance records. If so, it selects specific maintenance cases that match the specific model and operating conditions of the target charging equipment. The preferred maintenance case acquisition module is used to call the data storage system to obtain the preferred maintenance case that matches the specific model and operating conditions of the target charging equipment; A correction factor generation module is used to extract the first performance change trend and the second performance change trend of the charging equipment after the maintenance time point in the specific maintenance case and the preferred maintenance case, respectively, and generate a correction factor based on the degree of deviation between the two; The correction factor generation module specifically includes: A first curve drawing unit is configured to obtain relay on / off voltage data and relay coil pull-in current data of a target charging device within a preset time period after a maintenance time point based on a data storage system and a specific maintenance case, integrate the obtained data, arrange them in a time series, and draw a first on / off voltage change curve and a first pull-in current change curve; A second curve drawing unit is configured to obtain, based on the data storage system and the preferred maintenance case, relay on / off voltage data and relay coil pull-in current data of the charging device corresponding to the preferred maintenance case within a preset time period after the maintenance time point, integrate the obtained data, arrange them in time series, and draw a second on / off voltage change curve and a second pull-in current change curve; a correction factor generating unit, configured to generate a correction factor based on a degree of deviation between the first on-off voltage variation curve and the second on-off voltage variation curve, and a degree of deviation between the first pickup current variation curve and the second pickup current variation curve; The parameter correction module is used to determine the welding process parameters of the target maintenance engineer when handling the target relay contact adhesion anomaly based on historical maintenance records, and to correct the welding process parameters through correction factors.

7. The charging pile fault backtracking system based on AI analysis according to claim 6 is characterized in that: The specific maintenance case acquisition module specifically includes: A record parsing unit is used to parse historical maintenance records and extract maintenance records of several relay contact adhesion anomalies of the same type as the target charging equipment handled by the target maintenance engineer; A proportion judgment unit is used to call the data storage system, count the proportion of maintenance records with short-term recurring faults in the extracted maintenance records, and judge whether the proportion exceeds a preset threshold; The specific maintenance case determination unit is used to perform a comparative analysis of the maintenance behavior of the target maintenance engineer if it is determined that the proportion exceeds a preset threshold, and to screen out specific maintenance cases that match the specific model and operating conditions of the target charging equipment.

8. The charging pile fault backtracking system based on AI analysis according to claim 7 is characterized in that: The preferred maintenance case acquisition module specifically includes: A percentage statistics unit is used to call the data storage system, retrieve the historical maintenance records of other maintenance engineers, and calculate the percentage of short-term recurring faults handled by each maintenance engineer when handling maintenance tasks of the same type as the target charging equipment; Excellent maintenance engineer identification unit, used to screen out excellent maintenance engineers with the lowest proportion of short-term recurring faults; The preferred maintenance case determination unit is used to search for maintenance cases that match the specific model and operating conditions of the target charging equipment and have no short-term recurring faults in the historical maintenance records of the selected excellent maintenance engineers, and use them as preferred maintenance cases.

Images