Fault positioning method of automobile air conditioner, diagnosis equipment, system and storage medium

Abstract

The embodiment of the invention relates to the technical field of air conditioner fault positioning, in particular to an automobile air conditioner fault positioning method, diagnosis equipment, a system and a storage medium. The method comprises the steps that automobile diagnosis data collected by diagnosis equipment and air conditioner measurement data collected by an air conditioner special tool are obtained, air conditioner associated data corresponding to an automobile air conditioner are analyzed from the automobile diagnosis data, and fault positioning information of the automobile air conditioner is generated based on the air conditioner associated data and the air conditioner measurement data. According to the embodiment of the invention, a multi-source data fusion mode is adopted, the air conditioner associated data corresponding to the automobile air conditioner is analyzed from the automobile diagnosis data by utilizing the automobile diagnosis data acquired by the diagnosis equipment, and then the air conditioner measurement data is acquired by utilizing the air conditioner special tool, so that manual participation is not needed; according to the method, the air conditioner associated data and the air conditioner measurement data can be automatically combined to generate the fault positioning information of the automobile air conditioner, so that interference of human factors is eliminated, and the accuracy and efficiency of fault positioning are improved.

Description

Technical Field

[0001] This application relates to the field of air conditioning fault location technology, and in particular to a fault location method, diagnostic equipment, system and storage medium for automotive air conditioning. Background Technology

[0002] When a car's air conditioning malfunctions, repair personnel manually diagnose the problem based on customer feedback and the air conditioning's symptoms. This method relies heavily on human experience, making it prone to errors. It can lead to technicians mistakenly disassembling or repairing the air conditioning unit, increasing the difficulty and cost of repairs. Summary of the Invention

[0003] One objective of this application is to provide a fault location method, diagnostic device, system, and storage medium for automotive air conditioning systems, thereby improving the accuracy of fault causes provided by related technologies.

[0004] In a first aspect, an embodiment of this application provides a method for locating faults in an automotive air conditioner, comprising: acquiring automotive diagnostic data collected by a diagnostic device and air conditioner measurement data collected by an air conditioner-specific tool; parsing air conditioner-related data corresponding to the automotive air conditioner from the automotive diagnostic data; and generating fault location information for the automotive air conditioner based on the air conditioner-related data and the air conditioner measurement data.

[0005] Optionally, generating fault location information for the vehicle air conditioner based on the air conditioner association data and the air conditioner measurement data includes: in response to the air conditioner association data not containing an air conditioner fault code corresponding to the vehicle air conditioner, obtaining a pre-trained air conditioner diagnostic model; and analyzing the air conditioner association data and the air conditioner measurement data based on the air conditioner diagnostic model to obtain the fault location information for the vehicle air conditioner.

[0006] Optionally, the step of analyzing the air conditioning-related data and the air conditioning measurement data based on the air conditioning diagnostic model to obtain the fault type of the vehicle air conditioning includes: fusing the air conditioning-related data and the air conditioning measurement data to obtain fused data; and inputting the fused data into the air conditioning diagnostic model to obtain fault location information of the vehicle air conditioning.

[0007] Optionally, the step of fusing the air conditioner-related data and the air conditioner measurement data to obtain fused data includes: normalizing the air conditioner-related data and the air conditioner measurement data to obtain normalized air conditioner-related data and normalized air conditioner measurement data, respectively; extracting air conditioner diagnostic features based on the normalized air conditioner-related data; extracting air conditioner measured features based on the normalized air conditioner measurement data; and fusing the air conditioner diagnostic features and the air conditioner measured features to obtain fused data.

[0008] Optionally, inputting the fused data into the air conditioning diagnostic model to obtain the fault location information of the vehicle air conditioning includes: acquiring historical fault data of the vehicle air conditioning; inputting the historical fault data and the fused data into the air conditioning diagnostic model to obtain the fault location information of the vehicle air conditioning.

[0009] Optionally, the air conditioning diagnostic model is a finely tuned large language model or a multimodal large model. The step of analyzing the air conditioning-related data and the air conditioning measurement data based on the air conditioning diagnostic model to obtain the fault location information of the vehicle air conditioning includes: obtaining diagnostic prompt words; inputting the air conditioning-related data, the air conditioning measurement data, and the diagnostic prompt words into the air conditioning diagnostic model so that the air conditioning diagnostic model returns the fault location information of the vehicle air conditioning under the prompt of the diagnostic prompt words.

[0010] Optionally, the air conditioning diagnostic model is configured with a fault location table. The step of analyzing the air conditioning-related data and the air conditioning measurement data based on the air conditioning diagnostic model to obtain the fault location information of the vehicle air conditioning includes: analyzing the air conditioning-related data and the air conditioning measurement data based on the fault location table to obtain the fault location information of the vehicle air conditioning.

[0011] Optionally, the fault location table includes a normal state table and a fault state table. The step of analyzing the air conditioning-related data and the air conditioning measurement data based on the fault location table to obtain the fault location information of the vehicle air conditioning includes: performing semantic transformation processing on the air conditioning-related data and the air conditioning measurement data based on the normal state table to obtain a semantic parameter set, wherein the semantic parameter set is configured to describe the working state of the vehicle air conditioning in different performance dimensions using semantic expressions; and performing matching processing on the semantic parameter set based on the fault state table to obtain the fault location information of the vehicle air conditioning.

[0012] Optionally, both the air conditioning associated data and the air conditioning measurement data include multiple performance parameters. The normal state table includes normal performance sets for different types of automotive air conditioners, and the normal performance sets include normal performance ranges for different performance dimensions of the automotive air conditioners. The step of performing semantic transformation processing on the air conditioning associated data and the air conditioning measurement data based on the normal state table to obtain a semantic parameter set includes: obtaining the target type information of the automotive air conditioner; searching the normal state table for the normal performance set corresponding to the target type information, and the normal performance set corresponding to the target type information is the target normal performance set; and performing semantic transformation processing on each performance parameter in the target parameter set based on the multiple normal performance ranges included in the target normal performance set to obtain the semantic parameter set.

[0013] Optionally, the fault status table includes multiple fault semantic sets, fault causes and fault objects corresponding to each fault semantic set, and the step of matching the semantic parameter set based on the fault status table to obtain the fault location information of the vehicle air conditioner includes: in response to the existence of a fault semantic set in the fault status table that matches the semantic parameter set, determining the fault semantic set that matches the semantic parameter set as the target fault semantic set; and generating the fault location information of the vehicle air conditioner based on the fault causes and fault objects of the target fault semantic set.

[0014] Optionally, the fault location information includes at least one fault object and at least one suspected fault point under the fault object, and the method further includes: acquiring at least one test data corresponding to the suspected fault point; determining a fault detection result for the suspected fault point based on the test data; and determining a target fault point based on the fault detection results of all suspected fault points, wherein the target fault point is a fault point marked as having a fault.

[0015] Optionally, the suspected fault point includes at least one individual test item, and obtaining at least one test data corresponding to the suspected fault point includes: obtaining the project data generated when each individual test item in the suspected fault point is executed; and obtaining the test data corresponding to the suspected fault point based on the project data of each individual test item.

[0016] Optionally, obtaining the project data generated when each individual test item in the suspected fault point is executed includes: obtaining historical fault data; determining the suspected fault point to be determined corresponding to the historical fault data from the at least one suspected fault point; and obtaining the project data generated when each individual test item in the suspected fault point to be determined is executed.

[0017] Optionally, obtaining the project data generated when each individual test item in the suspected fault point is executed includes: determining a recommended test order, wherein the recommended test order represents the execution order of each individual test item in the suspected fault point; and obtaining the project data generated when each individual test item in the suspected fault point is executed in sequence according to the recommended test order.

[0018] Optionally, determining the recommended test order includes: obtaining test evaluation information for each individual test item; and determining the recommended test order based on the test evaluation information for each individual test item.

[0019] Optionally, generating fault location information for the vehicle air conditioner based on the air conditioner association data and the air conditioner measurement data further includes: generating fault location information for the vehicle air conditioner based on the air conditioner fault code corresponding to the vehicle air conditioner in response to the air conditioner association data containing an air conditioner fault code corresponding to the vehicle air conditioner.

[0020] In a second aspect, embodiments of this application provide a diagnostic device, including a memory and a processor. The memory is connected to the processor, and the processor is configured to execute one or more computer programs stored in the memory. When the processor executes the one or more computer programs, it enables the diagnostic device to implement the aforementioned method for locating faults in automotive air conditioning systems.

[0021] In a third aspect, embodiments of this application provide a fault location system for an automotive air conditioning system, including an air conditioning-specific tool and the aforementioned diagnostic equipment. The air conditioning-specific tool is configured to collect air conditioning measurement data; the diagnostic equipment is communicatively connected to the air conditioning-specific tool.

[0022] In a fourth aspect, embodiments of this application provide a computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program, the computer program including program instructions, which, when executed by a processor, cause the processor to perform the above-described fault location method for automotive air conditioning.

[0023] The embodiments of this application can achieve the following technical effects: The embodiments of this application adopt a multi-source data fusion method. Using the vehicle diagnostic data collected by the diagnostic equipment, the air conditioning related data corresponding to the vehicle air conditioning is parsed from the vehicle diagnostic data. Then, the air conditioning measurement data collected by the air conditioning special tool is used. Without human intervention, the air conditioning related data and the air conditioning measurement data can be automatically combined to generate the fault location information of the vehicle air conditioning, thereby eliminating the interference of human factors and improving the accuracy and efficiency of fault location. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 A schematic diagram of the system architecture of the fault location system for automotive air conditioning provided in this application embodiment; Figure 2 for Figure 1 A schematic diagram of a circuit structure for the diagnostic device shown. Figure 3 for Figure 1 A schematic diagram of another circuit structure for the diagnostic device shown; Figure 4 A flowchart illustrating a method for locating faults in an automotive air conditioner, provided as another embodiment of this application; Figure 5 This is a schematic diagram illustrating the comparison of the AC value and OBD value of the target faulty object, as provided in an embodiment of this application. Figure 6 This is a schematic diagram illustrating high pressure and low pressure obtained through publicly available channels in an embodiment of this application. Figure 7 This is a schematic diagram illustrating the acquisition of various performance parameters based on a testing process, according to another embodiment of this application. Figure 8 This is a schematic diagram of the structure of an air conditioning diagnostic model provided in an embodiment of this application; Figure 9 This is a schematic diagram of the structure of an air conditioning diagnostic model provided in another embodiment of this application; Figure 10 A schematic diagram illustrating how the air conditioning diagnostic model provided in this application outputs fault location information under the prompt of diagnostic prompt words; Figure 11 This is a schematic diagram of a normal state table according to an embodiment of this application; Figure 12 A schematic diagram of a fault status table provided in an embodiment of this application; Figure 13 A flowchart illustrating a method for locating faults in an automotive air conditioner, provided in yet another embodiment of this application; Figure 14 A schematic diagram of the interface for collecting high pressure, high pressure temperature, subcooling, low pressure, low pressure temperature and superheat provided in the embodiments of this application; Figure 15 A schematic diagram of the interface for collecting outlet air temperature and evaporator temperature provided in an embodiment of this application; Figure 16 This application provides an embodiment of an interface diagram containing fault location information. Figure 17 A flowchart illustrating a method for locating faults in an automotive air conditioner, provided in yet another embodiment of this application; Figure 18 A schematic diagram illustrating the relationship between various fault points and individual test items provided in the embodiments of this application; Figure 19 A flowchart illustrating a method for locating faults in an automotive air conditioner, provided in yet another embodiment of this application; Figure 20 A schematic diagram of the structure of a fault location device for an automotive air conditioner provided in an embodiment of this application; Figure 21 This is a schematic diagram of the structure of a diagnostic device provided in an embodiment of this application. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments in this application without inventive effort are within the scope of protection of this application.

[0027] It should be noted that, unless there is a conflict, the various features in the embodiments of this application can be combined with each other, all of which are within the protection scope of this application. Furthermore, although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than the module division in the device or the order in the flowchart. Moreover, the terms "first," "second," and "third" used in this application do not limit the data or execution order, but only distinguish identical or similar items with essentially the same function and effect.

[0028] Before introducing the working principle of the embodiments of this application, the performance parameters of the automotive air conditioner are explained as follows: The performance parameters of the automotive air conditioner include high pressure, low pressure, high pressure temperature, low pressure temperature, subcooling, superheating, evaporator temperature, and air outlet temperature.

[0029] High-side pressure is the refrigerant pressure between the condenser inlet and the expansion valve inlet in the air conditioning refrigeration cycle. High-side pressure is mainly generated by the compressor compression and is significantly affected by the condenser heat dissipation efficiency.

[0030] Low-side pressure is the refrigerant pressure between the outlet of the expansion valve and the inlet of the compressor in the air conditioning refrigeration cycle. Low-side pressure depends on the heat absorption efficiency of the evaporator and the throttling effect of the expansion valve.

[0031] High-side temperature is the refrigerant temperature in the high-pressure range, and it is usually measured at two key points: the condenser inlet temperature and the condenser outlet temperature.

[0032] Low-side temperature is the refrigerant temperature in the low-pressure range, usually taken as the evaporator outlet temperature, which directly reflects the heat absorption effect of the evaporator.

[0033] Subcooling degree is the difference between the temperature of the liquid refrigerant at the condenser outlet and the saturation temperature of the refrigerant under the current high pressure. It is an indicator of the "purity of liquid" of the liquid refrigerant.

[0034] Superheating degree is the difference between the temperature of the gaseous refrigerant at the evaporator outlet and the saturation temperature of the refrigerant at the current low pressure. It is an indicator of the "pure gaseousness" of the gaseous refrigerant.

[0035] Evaporator temperature is the average temperature of the evaporator core surface, which directly determines the "cooling potential" of the cabin air.

[0036] Air outlet temperature is the air temperature at the air outlet of the air conditioning duct. It is a core indicator that users can directly perceive and is affected by evaporator temperature, air volume, cabin set temperature, and ambient temperature.

[0037] The inventors discovered that refrigerant circulation failures in automotive air conditioning systems involve factors such as refrigerant leaks, pipe blockages, and component malfunctions. Repair personnel primarily rely on manual experience to locate faults in automotive air conditioning systems, or they may rely solely on high or low pressure to pinpoint the problem, resulting in low accuracy in fault location.

[0038] The inventors also discovered that many related automotive air conditioning fault diagnosis methods rely on a single data source. For example, diagnosing an automotive air conditioning system solely based on diagnostic data read by diagnostic equipment can easily lead to misdiagnosis when the vehicle's sensors malfunction while the air conditioning components function normally. Furthermore, diagnosing an automotive air conditioning system using only data such as high-pressure, low-pressure, or high-pressure temperature measured by tools like manifolds and temperature sensors fails to integrate the vehicle's various sensors and effectively utilize diagnostic data, resulting in low fault location efficiency and accuracy.

[0039] This application embodiment adopts a multi-source data fusion method. It uses vehicle diagnostic data collected by diagnostic equipment to parse the air conditioning-related data corresponding to the vehicle's air conditioning from the vehicle diagnostic data, and then uses air conditioning measurement data collected by air conditioning-specific tools. Without human intervention, it can automatically combine the air conditioning-related data and the air conditioning measurement data to generate fault location information for the vehicle's air conditioning, thereby eliminating the interference of human factors and improving the accuracy and efficiency of fault location.

[0040] The following embodiments of this application provide a fault location system for an automotive air conditioning system. Please refer to... Figure 1 The fault location system 100 for automotive air conditioning includes an air conditioning tool 200 and a diagnostic device 300, which are connected in communication with the air conditioning tool 200.

[0041] The air conditioner-specific tool 200 is used to collect air conditioner measurement data, including various performance parameters such as high pressure, low pressure, high pressure temperature, low pressure temperature, evaporator temperature, and outlet temperature. The air conditioner-specific tool 200 supports multiple data acquisition functions; for example, it integrates pressure and temperature acquisition functions.

[0042] In some embodiments, the air conditioning tool 200 is equipped with multiple sensors capable of collecting high-pressure, low-pressure, high-pressure, low-pressure, evaporator, and outlet air temperatures. During the data collection process, the sensors can be positioned at corresponding locations to obtain the performance parameters corresponding to those locations. Then, the diagnostic device 300 derives the subcooling and superheating based on these parameters.

[0043] In other embodiments, the air conditioning tool 200 includes a manifold and a temperature measuring instrument. The manifold is used to collect high-pressure, low-pressure, high-pressure, and low-pressure temperature data. The temperature measuring instrument can detect the evaporator temperature and the air outlet temperature. The manifold is equipped with a pressure probe and a thermocouple clamp. Maintenance personnel can place the pressure probe on the high-pressure or low-pressure access port of the vehicle's air conditioning system to obtain the high-pressure or low-pressure data. Maintenance personnel can clamp the thermocouple clamp onto the high-pressure or low-pressure pipe of the vehicle's air conditioning system to obtain the high-pressure and low-pressure temperatures. The temperature measuring instrument is equipped with a temperature probe, which can be placed near the evaporator or at the air outlet of the vehicle's air conditioning system to obtain the evaporator temperature or the air outlet temperature.

[0044] The diagnostic device 300 is used to diagnose vehicle faults. During diagnosis, the repair personnel use the diagnostic device 300 to connect to the vehicle's OBD (On-Board Diagnostics) interface and obtain the vehicle's diagnostic data based on the OBD interface.

[0045] Please see Figure 2 The diagnostic device 300 includes a diagnostic host 31, a vehicle communication interface 32 (VCI), and a connecting cable 33. The diagnostic host 31 is communicatively connected to the vehicle communication interface 32, and the vehicle communication interface 32 is connected to one end of the connecting cable 33. The other end of the connecting cable 33 is an OBD interface, used for adapter connection with the vehicle's OBD interface.

[0046] During diagnosis, the vehicle communication interface 32 connects to the vehicle's OBD interface via the connecting cable 33 to read the vehicle's OBD data and transmit the OBD data to the vehicle communication interface 32. The vehicle communication interface 32 converts the OBD data into vehicle diagnostic data according to the corresponding communication protocol and finally transmits the vehicle diagnostic data to the diagnostic host 31. The corresponding communication protocols include CAN protocol, KWP2000, or J1850, etc.

[0047] In some embodiments, such as Figure 2 As shown, the diagnostic host 31, vehicle communication interface 32 and connecting cable 33 are integrated together to form a diagnostic device 300.

[0048] In other embodiments, such as Figure 3As shown, the vehicle communication interface 32 and the connecting cable 33 are integrated together, while the diagnostic device 300 is a separate component. The diagnostic device 300 and the vehicle communication interface 32 can communicate wirelessly. For example, after the vehicle communication interface 32 converts the OBD data into automotive diagnostic data according to the corresponding communication protocol, it transmits the automotive diagnostic data to the diagnostic host 31 wirelessly. Wireless communication methods include Bluetooth, Wi-Fi, 3G, 4G, 5G, and 6G connections.

[0049] The diagnostic device 300 can combine vehicle diagnostic data and air conditioning measurement data to generate fault location information for the vehicle's air conditioning system. Before using the diagnostic device 300 to perform automated diagnostics on the vehicle's air conditioning system, repair personnel need to perform manual diagnostic operations. Manual diagnostic operations include confirming the customer's vehicle's symptoms and visually inspecting the vehicle's air conditioning system.

[0050] The customer drives their car to the repair shop and reports issues related to the car's air conditioning. After receiving the feedback, the repair personnel perform a symptom assessment. If the customer's reported problem is not related to the air conditioning, the personnel explain the situation to the customer and do not proceed to the next repair step. If the customer's reported problem is related to the air conditioning, the personnel need to categorize the specific issue to provide a starting point for subsequent inspections. Common air conditioning malfunctions include: poor cooling performance, no cooling at all, abnormal airflow from the vents, temperature differences at the vents, unusual odors from the vents, loud noises / sounds when the air conditioning is on, condensation / humid interior, abnormal airflow direction adjustment, and inability to switch between internal and external air circulation.

[0051] After confirming the fault, the repair personnel begin a visual inspection of the car's air conditioning system according to the given checklist. The visual inspection focuses on exposed and easily accessible components of the air conditioning system, such as the compressor, accessory belts, high and low pressure lines, expansion valve, fan, condenser, cabin air filter, and air vent blades. Through visual inspection, the repair personnel determine whether the relevant components of the car's air conditioning system are damaged, loose, detached, or leaking. The purpose of this step is to quickly diagnose and eliminate easily detectable faults through a short visual inspection, conforming to the logic of troubleshooting from simple to complex.

[0052] After symptom confirmation and visual inspection, the repair personnel confirmed a fault in the vehicle's air conditioning system. The diagnostic equipment 300 then began automated diagnostics. This automated diagnostic process includes diagnosing the vehicle's air conditioning system based on vehicle diagnostic data and diagnosing the system using both vehicle diagnostic data and air conditioning measurement data. For example, the diagnostic equipment reads the vehicle's diagnostic data and checks for a corresponding diagnostic code for the air conditioning system. Based on this diagnostic code, the fault in the air conditioning system is located.

[0053] When no fault codes corresponding to the car's air conditioning system are found in the vehicle diagnostic data, the diagnostic equipment can diagnose the car's air conditioning system based on both the vehicle diagnostic data and the air conditioning measurement data. This application's embodiment employs the above diagnostic process, locating the car's air conditioning fault from simple to complex. This approach is highly efficient and comprehensive in fault location, reliably and accurately pinpointing the location of the car's air conditioning system fault.

[0054] To gain a deeper understanding of the fault location method for automotive air conditioning according to the embodiments of this application, the following embodiments of this application are combined with... Figure 4 The fault location method for automotive air conditioning is explained in detail through steps S41 to S43. The main body for executing the fault location method for automotive air conditioning is the diagnostic equipment, as shown below: Step S41: Obtain vehicle diagnostic data collected by diagnostic equipment and air conditioning measurement data collected by air conditioning-specific tools.

[0055] Automotive diagnostic data is a collection of performance parameters of various components in a vehicle, gathered by diagnostic equipment. This data includes various Diagnostic Trouble Codes (DTCs), freeze frame data, and real-time data streams. DTCs indicate the operational status of the corresponding component, freeze frame data records parameters when a component malfunctions, and real-time data streams contain data currently detected by various sensors in the vehicle. For example, automotive diagnostic data may include performance parameters of the engine, transmission, and air conditioning system.

[0056] The timestamps of vehicle diagnostic data and air conditioning measurement data can be aligned or not. When the timestamps of vehicle diagnostic data and air conditioning measurement data are aligned, this embodiment can directly analyze the vehicle diagnostic data and air conditioning measurement data. When the timestamps of vehicle diagnostic data and air conditioning measurement data are not aligned, this embodiment performs time alignment processing on the vehicle diagnostic data and air conditioning measurement data to obtain aligned vehicle diagnostic data and aligned air conditioning measurement data, and then performs analysis based on the aligned vehicle diagnostic data and aligned air conditioning measurement data.

[0057] Air conditioning measurement data is a collection of performance parameters of the car's air conditioning system collected by specialized air conditioning tools. This data includes measurements of high pressure, low pressure, high pressure temperature, low pressure temperature, evaporator temperature, and air outlet temperature.

[0058] Acquiring air conditioning measurement data collected by a dedicated air conditioning tool includes the following steps: establishing a communication connection between the control diagnostic device and the dedicated air conditioning tool, and acquiring the air conditioning measurement data transmitted by the dedicated air conditioning tool based on the communication connection. For example, during the acquisition process, the dedicated air conditioning tool establishes a Bluetooth connection with the diagnostic device, and the dedicated air conditioning tool sends the air conditioning measurement data to the diagnostic device for storage based on the Bluetooth connection. As another example, the dedicated air conditioning tool establishes a wired connection with the diagnostic device via a wired communication cable, and the dedicated air conditioning tool sends the air conditioning measurement data to the diagnostic device for storage based on the wired communication cable.

[0059] Step S42: Extract the air conditioning-related data corresponding to the car's air conditioning from the vehicle diagnostic data.

[0060] Air conditioning-related data refers to the diagnostic data associated with the vehicle's air conditioning system within the vehicle diagnostic data. It's important to understand that vehicle diagnostic data includes not only diagnostic data related to the air conditioning system but also diagnostic data for other components in the vehicle.

[0061] In some embodiments, vehicle diagnostic data includes various fault codes, freeze frame data, and real-time data streams. Extracting air conditioning-related data from the vehicle diagnostic data includes the following steps: filtering fault codes belonging to the vehicle air conditioning system from the vehicle diagnostic data as air conditioning fault codes; parsing freeze frame data to extract key operating parameters when the vehicle air conditioning system malfunctions; extracting parameters associated with the vehicle air conditioning system from the real-time data stream as air conditioning performance parameters; and packaging the air conditioning fault codes, key operating parameters, and air conditioning performance parameters into air conditioning-related data. Air conditioning performance parameters include high-pressure, low-pressure, high-pressure temperature, low-pressure temperature, evaporator temperature, and outlet temperature.

[0062] In other embodiments, parsing air conditioning-related data corresponding to the vehicle's air conditioning from vehicle diagnostic data includes the following steps: obtaining a pre-trained data cleaning model, inputting the vehicle diagnostic data into the data cleaning model, so that the data cleaning model extracts the air conditioning-related data associated with the vehicle's air conditioning from the vehicle diagnostic data.

[0063] In some embodiments, the data cleaning model can be a finely tuned large language model or a multimodal large model. In this application embodiment, an air conditioning prompt word is constructed. This prompt word, along with vehicle diagnostic data, is input into the data cleaning model. Under the prompt of the air conditioning prompt word, the data cleaning model extracts air conditioning-related data from the vehicle diagnostic data. For example, the air conditioning prompt word could be: "Please extract parameters directly related to the vehicle's air conditioning from the following vehicle diagnostic data, including but not limited to air conditioning fault codes, refrigerant pressure sensor values, ambient temperature sensor values, compressor clutch status and corresponding timestamps, and remove irrelevant parameters such as engine speed and transmission gear." Under such a prompt word, the data cleaning model can extract air conditioning-related data from the vehicle diagnostic data.

[0064] In other embodiments, the data cleaning model is a pre-trained deep learning model. In this application embodiment, vehicle diagnostic data is input into the data cleaning model, which automatically labels each data point in the vehicle diagnostic data and automatically outputs data labeled with air conditioning information, thereby obtaining vehicle-related data. The deep learning model includes BiLSTM-CRF models, CNN-MLP models, or lightweight Transformer models, etc.

[0065] Step S43: Based on the air conditioning correlation data and air conditioning measurement data, generate fault location information for the vehicle air conditioning.

[0066] Fault location information is used to locate faults in automotive air conditioning. In this application embodiment, various analysis methods can be used to process air conditioning-related data and air conditioning measurement data to obtain fault location information for automotive air conditioning. Therefore, by combining multiple data sources, this application embodiment can automatically generate fault location information for automotive air conditioning, which is beneficial to improving the accuracy and efficiency of fault location.

[0067] In some embodiments, the air conditioning associated data and air conditioning measurement data are input into a pre-trained artificial intelligence model, so that the artificial intelligence model performs fault location analysis on the air conditioning associated data and air conditioning measurement data to obtain fault location information of the car air conditioning.

[0068] In some embodiments, this application can perform fault location analysis by combining the presence of air conditioning fault codes in the air conditioning associated data. Specifically, this application generates fault location information for the automotive air conditioning system based on the air conditioning associated data and air conditioning measurement data through steps S431 and S432, as shown below: Step S431: In response to the fact that the air conditioning associated data does not contain an air conditioning fault code corresponding to the car air conditioning, obtain the pre-trained air conditioning diagnostic model.

[0069] Step S432: Analyze the air conditioning related data and air conditioning measurement data based on the air conditioning diagnostic model to obtain the fault location information of the car air conditioning.

[0070] The air conditioning fault code is a fault code corresponding to the car's air conditioning system. This application embodiment, based on air conditioning associated data, detects whether the associated data contains an air conditioning fault code corresponding to the car's air conditioning system. If the associated data contains such a fault code, fault location information for the car's air conditioning system is generated based on the fault code. If the associated data does not contain such a fault code, a pre-trained air conditioning diagnostic model is obtained.

[0071] Based on air conditioning associated data, the process of detecting whether the air conditioning associated data contains air conditioning fault codes corresponding to the car's air conditioning system includes the following steps: inputting the air conditioning associated data into a preset fault finding model, so that the fault finding model can detect whether the air conditioning associated data contains air conditioning fault codes corresponding to the car's air conditioning system.

[0072] In some embodiments, the fault finding model is a finely tuned large language model or a multimodal large model. This application embodiment constructs fault code lookup prompts, inputs these prompts along with air conditioning association data into the fault finding model, and enables the model to extract air conditioning fault codes associated with the vehicle's air conditioning from the associated data, guided by the fault code lookup prompts. For example, the fault code lookup prompts are: "Please extract all fault codes belonging to the vehicle's air conditioning system from the following associated data. Air conditioning fault code judgment criteria: 1. Prefixed with B (body system) and related to HVAC (Heating, Ventilation and Air Conditioning); 2. Prefixed with P and related to refrigerant pressure and compressor clutch; 3. Custom fault codes contain keywords such as air conditioning, refrigeration, pressure, and fan. Non-air conditioning fault codes should be ignored. Air conditioning association data: {input air conditioning association data}". Guided by these prompts, the fault finding model extracts air conditioning fault codes associated with the vehicle's air conditioning from the associated data.

[0073] If the fault finding model does not output an air conditioning fault code corresponding to the vehicle's air conditioning system, this embodiment of the application obtains a pre-trained air conditioning diagnostic model for subsequent diagnosis of the vehicle's air conditioning system. If the fault finding model outputs an air conditioning fault code corresponding to the vehicle's air conditioning system, this embodiment of the application generates fault location information for the vehicle's air conditioning system based on the corresponding fault code.

[0074] In some other embodiments, please refer to Table 1. The fault finding model is configured with an air conditioning fault table, which records various fault codes of car air conditioning systems of various brands.

[0075] Table 1

[0076] In some embodiments, the process of generating the air conditioning fault table includes: as described in Table 1, designers can collect various types of automotive air conditioning data, then manually clean the fault codes of each component from the automotive air conditioning data, and record the fault codes of each component in the air conditioning fault table, thereby making the air conditioning fault table a collection of various fault codes of automotive air conditioning from various brands.

[0077] In other embodiments, the process of generating the air conditioning fault table includes: inputting various types of collected automotive air conditioning data into a data cleaning model, so that the data cleaning model records various fault codes of each automotive air conditioning system in the air conditioning fault table.

[0078] If the air conditioning fault table does not contain a fault code corresponding to the vehicle's air conditioning system, this embodiment of the application obtains a pre-trained air conditioning diagnostic model so that the vehicle's air conditioning system can be diagnosed using the model subsequently. If the air conditioning fault table contains a fault code corresponding to the vehicle's air conditioning system, this embodiment of the application generates fault location information for the vehicle's air conditioning system based on the fault code.

[0079] When the diagnostic equipment detects an air conditioning fault code corresponding to the car's air conditioning through air conditioning-related data, it can further use the first diagnostic path corresponding to the air conditioning diagnostic code to perform a more granular diagnosis of the car's air conditioning.

[0080] In some embodiments, this application can generate diagnostic guidance information, which is used to prompt maintenance personnel to take corresponding measures to diagnose the fault object corresponding to the air conditioning diagnostic code according to the content of the diagnostic guidance information. For example, the diagnostic guidance information includes one or more of the following diagnostic methods: ① Obtaining the air conditioning OBD value corresponding to the air conditioning fault code, obtaining the measurement value of the target fault object measured by an air conditioning-specific tool, and comparing the air conditioning OBD value with the measurement value, wherein the target fault object is the fault object corresponding to the air conditioning fault code; ② Controlling the target fault object to perform an action, and obtaining the air conditioning OBD value when the target fault object performs the action, and checking whether the air conditioning OBD value is abnormal; ③ Controlling the target fault object to perform an action, and observing whether the actual action of the target fault object is abnormal. The air conditioning OBD value is the OBD data corresponding to the car's air conditioning that the diagnostic equipment filters from the car diagnostic data.

[0081] For another example, please refer to Table 2: Table 2

[0082] As shown in Table 2, the diagnostic guidance information is the information in the first row of Table 2. That is, in the diagnostic guidance information in the first row, the air conditioner fault code is P0530, and the diagnostic guidance information is: compare the OBD value of the refrigerant pressure sensor with the pressure value of the air conditioner-specific tool. The OBD value of the refrigerant pressure sensor is the pressure value detected by the refrigerant pressure sensor. In some embodiments, the air conditioner-specific tool includes a diagnostic instrument, a lower-level computer, and measuring tools, including a manifold gauge, temperature clamp, etc. In other embodiments, the air conditioner-specific tool is a manifold gauge, temperature clamp, etc.

[0083] Please refer to the diagram. Figure 5 The repair technician compares the pressure value detected by the refrigerant pressure sensor with the pressure value measured by a specialized air conditioning tool. If they match or the difference is within a preset threshold, a confirmation message indicating a refrigerant pressure sensor malfunction is generated. If they do not match or the difference is outside the preset threshold, a malfunction message is generated.

[0084] In other embodiments, this application embodiment can directly determine the target fault object based on the air conditioner fault code without generating diagnostic guidance information, obtain the measurement value of the target fault object measured by air conditioner special tools, compare the measurement value of the target fault object with the OBD value, and when the difference between the measurement value of the target fault object and the OBD value is within a preset threshold range, generate confirmation information about the refrigerant pressure sensor malfunction and output first DTC repair guidance information. The first DTC repair guidance information is used to prompt the repair personnel for the next inspection steps. When the difference between the measurement value of the target fault object and the OBD value is within a preset threshold range, generate information about the refrigerant pressure sensor malfunction and output second DTC repair guidance information. The second DTC repair guidance information is used to prompt the repair personnel for the next inspection steps.

[0085] This application embodiment directly identifies the faulty object through air conditioning fault codes, significantly shortening the diagnosis time and thus improving diagnosis efficiency. Furthermore, it eliminates the need to rely on the experience of maintenance personnel to determine the faulty object, which helps improve the accuracy of fault location.

[0086] When no air conditioning fault code is found from the air conditioning associated data in this application embodiment, the location of the fault cannot be identified due to the lack of information. Therefore, this application embodiment needs to obtain a pre-trained air conditioning diagnostic model and use it to diagnose the vehicle's air conditioning system. The air conditioning diagnostic model is a pre-trained model capable of outputting fault location information about the vehicle's air conditioning system. The air conditioning diagnostic model includes an artificial intelligence model trained based on deep learning algorithms, or a model obtained by fine-tuning and training a commercial large language model or multimodal large model as a base.

[0087] This application embodiment, in accordance with the requirements for air conditioner diagnosis, pre-trains and generates an air conditioner diagnostic model. Specifically, this application embodiment requires collecting training sample data, labeling the training sample data, and processing the labeled training sample data based on deep learning algorithms or fine-tuning rules to obtain the air conditioner diagnostic model.

[0088] The training sample data includes performance parameters and fault diagnosis data of various automotive air conditioners. For example, the training sample data includes fault symptom description information, maintenance data, OBD data stream, data measured by air conditioning special tools, temperature, overheating / overcooling data, etc.

[0089] There are two scenarios: the performance parameters of a car air conditioner can be obtained through public channels, and the performance parameters of a car air conditioner can only be obtained through testing.

[0090] 1) The performance parameters of automotive air conditioning systems are available through public channels. Designers can collect various performance parameters of automotive air conditioning systems under normal operating conditions by analyzing publicly available after-sales service data from different automakers, categorized by brand, model, year, and configuration. Please refer to [link / reference]. Figure 6 By collecting information on the high and low pressure of car air conditioners under normal operating conditions through publicly available after-sales maintenance data from various car manufacturers, based on brand, model, year, and configuration, we were able to obtain this information.

[0091] 2) The performance parameters of automotive air conditioning can only be obtained through testing. Designers can set the air conditioning units of each type of vehicle under normal operating conditions and obtain various performance parameters of this type of air conditioning unit using various testing tools according to a predetermined testing procedure. For example, the testing procedure for various test vehicles in this application embodiment is as follows: 2.1) Record the current temperature and humidity (30℃-35℃ is recommended).

[0092] 2.2) Record the nameplate information of the current vehicle.

[0093] 2.3) Record the current vehicle type information (e.g., sedan, coupe, performance sports car, SUV, single-cab pickup, double-cab pickup, etc.).

[0094] 2.4) Record the current refrigerant type and refrigerant charge amount of the vehicle.

[0095] 2.5) Record the current characteristics of the vehicle's cooling cycle (e.g., coaxial pipe, expansion joint). 2.6) Turn the key, start the engine, keep the engine idling, turn on the air conditioning in cooling mode, maximum airflow, coldest temperature, and recirculation mode, and connect the diagnostic equipment to various testing devices.

[0096] 2.7) Record the time when the engine idles to achieve the best cooling effect. When the cooling cycle is stable and circulating within a certain pressure or temperature range, record the high pressure, low pressure, outlet temperature, and evaporator temperature read by the detection device.

[0097] 2.8) Stabilize the engine speed at 1500 rpm.

[0098] 2.9) When the refrigeration cycle is stable and circulating within a certain pressure or temperature range, record the high pressure of the refrigerant, the low pressure of the refrigerant, the outlet temperature, and the evaporator temperature read by the detection device.

[0099] 2.10) Fill in the record information in the form as required.

[0100] According to the above testing procedure, the high pressure, low pressure, air outlet temperature, and evaporator temperature of the car air conditioner of each model under normal operating conditions can be obtained in this application embodiment.

[0101] Please refer to Table 3: Table 3

[0102] As shown in Table 3, the embodiments of this application can also collect performance parameters of the car air conditioning of vehicles with different structures (such as sedans / SUVs / pickups, with / without coaxial pipes, fixed displacement / variable displacement compressors, etc.) and different conditions (outdoor temperature, humidity, engine idle speed / 1500 rpm, internal circulation / external circulation, etc.). The collection of performance parameters is summarized in Table 3.

[0103] The performance parameters obtained from the above testing process are summarized in this application embodiment, as shown in Table 4: Table 4

[0104] As shown in Table 4, the high-pressure, low-pressure, outlet temperature, and evaporator temperature of the car air conditioning system under normal operating conditions correspond to the normal performance range for each type of vehicle. As shown in Table 4, the normal performance range of the low-pressure for the first vehicle type under normal operating conditions is 1.00 bar - 3.00 bar.

[0105] The inventors discovered that some vehicle air conditioners lack information on the normal performance ranges of high-pressure temperature, low-pressure temperature, subcooling, and superheating, and these parameters are rarely used for fault diagnosis. This application's embodiments can specify a test procedure to test the normal performance ranges of high-pressure temperature, low-pressure temperature, subcooling, and superheating of a vehicle's air conditioner.

[0106] For example, the embodiments of this application provide the following testing process, as shown below: 3.1) The test vehicle should be placed in a shady place, not in direct sunlight.

[0107] 3.2) Close all windows and doors of the test vehicle and open the engine hood.

[0108] 3.3) Set the vent fan to high speed.

[0109] 3.4) Turn on the air conditioning system and set it to recirculation mode.

[0110] 3.5) Set the air outlet temperature to the lowest setting.

[0111] 3.6) The high-pressure temperature clamp is attached to the outlet of the condenser, and the low-pressure temperature clamp is attached to the outlet of the expansion valve.

[0112] 3.7) The vehicle should be idled for 10 minutes before testing.

[0113] Based on the above testing process, the embodiments of this application can obtain the following: Figure 7 The performance parameters shown.

[0114] Furthermore, regarding the normal performance range of superheat and supercooling, the embodiments of this application determine them according to the following rules: 4.1) Superheat range: The lowest value is -0.7℃ and the highest value is 23.3℃. Most data are distributed between 5℃ and 15℃, with few extreme values. The range values ​​for some models are 0.2℃-1.4℃ and 13.1℃-14.2℃. Therefore, the normal performance range of superheat in this application embodiment is 1℃-20℃.

[0115] 4.2) Subcooling range: The lowest value is -3.2℃ and the highest value is 28.6℃. Most data are distributed between 5℃ and 14℃, with few extreme values. The range values ​​for some models are 5.7℃-6.2℃ and 16.7℃-28.6℃. Therefore, the normal performance range of superheat in this application embodiment is 2℃-15℃.

[0116] Following the above approach, performance parameters such as high pressure, low pressure, high pressure temperature, low pressure temperature, subcooling, superheating, evaporator temperature, and air outlet temperature are entered into the database according to the brand, model, year, and configuration, and based on different ambient temperatures and humidity levels.

[0117] It is understood that the embodiments of this application not only need to obtain training sample data of the car air conditioner under normal operating conditions as described above, but also need to obtain training sample data of the car air conditioner under fault operating conditions, so as to combine the training sample data under normal operating conditions and the training sample data under fault operating conditions to train the air conditioner diagnostic model.

[0118] Obtaining training sample data for automotive air conditioning systems under fault conditions involves the following steps: 1) Simulating corresponding fault conditions for different components of the automotive air conditioning system, placing each component in its corresponding failure mode. 2) Conducting tests under these fault conditions to obtain various performance parameters.

[0119] For example, please refer to Table 5: Table 5

[0120] As shown in Table 5, this embodiment of the application simulates a compressor failure condition using the following scenario construction method, causing the compressor to be in a failure mode of "piston or cylinder damage and poor sealing". The scenario construction method involves using a viewport tool to connect the high and low pressure service ports, opening a portion of the passage to allow refrigerant to circulate within the viewport. This embodiment of the application tests the automotive air conditioning system under the aforementioned failure condition to obtain various performance parameters.

[0121] As another example, this application embodiment simulates a compressor failure condition according to the following scenario construction method, causing the compressor to be in the failure mode of "damaged interface seal or gasket leakage". The scenario construction method is: adding less refrigerant, at 20% of the total amount. This application embodiment tests the automotive air conditioner under the above-mentioned failure condition and obtains various performance parameters.

[0122] According to the scenario construction method of various components shown in Table 5, the embodiments of this application simulate fault conditions for the corresponding components, so that the components are in the corresponding failure modes, and test the car air conditioner under the above fault conditions to obtain various performance parameters.

[0123] After collecting various training sample data, this embodiment of the application preprocesses the training sample data to obtain preprocessed training sample data. Preprocessing includes normalization and noise filtering. This embodiment of the application also labels the preprocessed training sample data, including fault labels and normal operating condition labels.

[0124] It is understood that the embodiments of this application can use deep learning algorithms to train an air conditioning diagnostic model based on collected training sample data of various types, or fine-tune a general large model or a multimodal large model to obtain an air conditioning diagnostic model, or even generate an air conditioning diagnostic model based on a fault location table. The specific discussion is as follows: ① An air conditioning diagnostic model was obtained by training based on a deep learning algorithm.

[0125] In this application embodiment, the gradient boosting decision tree model (LightGBM / XGBoost), convolutional recurrent neural network (CNN-LSTM), or fully connected neural network algorithm is used to process the labeled training sample data to obtain an air conditioning diagnostic model.

[0126] Based on the air conditioning diagnostic model, the analysis of air conditioning related data and air conditioning measurement data to obtain the fault type of the car air conditioning includes the following steps: fusing the air conditioning related data and air conditioning measurement data to obtain fused data, and inputting the fused data into the air conditioning diagnostic model to obtain the fault location information of the car air conditioning.

[0127] The fused data is a structured set of air conditioner diagnostic features and measured features. The process of fusing air conditioner correlation data and air conditioner measurement data to obtain fused data includes the following steps: normalizing the air conditioner correlation data and air conditioner measurement data to obtain normalized air conditioner correlation data and normalized air conditioner measurement data respectively; extracting air conditioner diagnostic features based on the normalized air conditioner correlation data; extracting air conditioner measured features based on the normalized air conditioner measurement data; and fusing the air conditioner diagnostic features and air conditioner measured features to obtain fused data.

[0128] Air conditioning diagnostic features are derived from air conditioning-related data and are used to describe the performance status of the car's air conditioning from the perspective of OBD data stream (which belongs to electronic control data), reflecting the electronic control logic status of the car's air conditioning.

[0129] The measured features of the air conditioning system are derived from air conditioning measurement data. They are used to describe the performance status of the automotive air conditioning system from the perspective of air conditioning measurement data (which belongs to physical data), reflecting the actual thermodynamic state of the automotive air conditioning system. The timestamps of the air conditioning diagnostic features are aligned with the timestamps of the measured features.

[0130] Both air conditioning correlation data and air conditioning measurement data include various performance parameters. Different types of performance parameters have different dimensions, necessitating normalization processing to ensure all performance parameters are on the same order of magnitude, thus guaranteeing the fairness of inference in the air conditioning diagnostic model. This application embodiment can employ various normalization functions to normalize the air conditioning correlation data and air conditioning measurement data respectively. For example, one normalization formula is: , This refers to the smallest performance parameter among the same type of performance parameters in air conditioning-related data or air conditioning measurement data. This refers to the highest performance parameter among the same type of performance parameters in air conditioning-related data or air conditioning measurement data. As a performance parameter, These are the normalized performance parameters.

[0131] Air conditioning diagnostic features include pressure change rate, temperature change rate, subcooling change rate, superheat change rate, average pressure, average temperature, average subcooling, average superheat, or pressure variance, temperature variance, subcooling variance, or superheat variance, obtained from OBD data streams. Extracting air conditioning diagnostic features based on normalized air conditioning correlation data involves the following steps: extracting the pressure change rate, temperature change rate, subcooling change rate, superheat change rate, average pressure, average temperature, average subcooling, average superheat, or pressure variance, temperature variance, subcooling variance, or superheat variance from the normalized air conditioning correlation data. Pressure includes high-pressure and low-pressure; temperature includes high-pressure temperature, low-pressure temperature, evaporator temperature, and outlet temperature.

[0132] Air conditioning diagnostic features include pressure change rate, temperature change rate, subcooling change rate, superheat change rate, average pressure, average temperature, average subcooling, average superheat, or pressure variance, temperature variance, subcooling variance, or superheat variance, etc., obtained from air conditioning measurement data. Extracting measured features of air conditioning based on normalized air conditioning measurement data includes the following steps: extracting pressure change rate, temperature change rate, subcooling change rate, superheat change rate, average pressure, average temperature, average subcooling, average superheat, or pressure variance, temperature variance, subcooling variance, or superheat variance from normalized air conditioning measurement data.

[0133] Please see Figure 8 The air conditioning diagnostic model 80 includes an input layer 81, a processing layer 82, and an output layer 83. The input layer 81 receives inputs of air conditioning diagnostic features and measured air conditioning features. The processing layer 82, which can be a fully connected layer, processes and analyzes these features. The diagnostic features are extracted from associated air conditioning data, while the measured features are extracted from air conditioning measurement data. The output layer 83 is configured with various fault location tags, each corresponding to a specific fault location information. The output layer 83 maps the output of the processing layer 82 to the corresponding fault location tag to output the corresponding fault location information.

[0134] This application embodiment can fuse multi-dimensional air conditioning diagnostic features with multi-dimensional air conditioning measured features to obtain multi-dimensional fused data. In some embodiments, before fusion, this application embodiment uses a dimensionality reduction algorithm to reduce the dimensionality of the air conditioning diagnostic features and the air conditioning measured features, obtaining dimensionality-reduced air conditioning diagnostic features and dimensionality-reduced air conditioning measured features. Then, the dimensionality-reduced air conditioning diagnostic features and dimensionality-reduced air conditioning measured features are fused to obtain multi-dimensional fused data. Dimensionality reduction algorithms include principal component analysis algorithms or linear discriminant analysis algorithms, etc. By performing dimensionality reduction on the air conditioning diagnostic features and air conditioning measured features, this application embodiment can maximize the preservation of original information, while facilitating the reduction of computational complexity of the air conditioning diagnostic model and improving inference speed.

[0135] This application embodiment fuses air conditioning diagnostic features with air conditioning measured features. The resulting fused data can reflect the performance status of the vehicle air conditioning from both the OBD data stream dimension and the real-time performance status of the vehicle air conditioning from the air conditioning measurement data dimension. This enables the air conditioning diagnostic model to output reliable and accurate fault location information.

[0136] Typically, when a car's air conditioning system experiences a current malfunction, it is often closely related to previous malfunctions. For example, the current malfunction and a previous malfunction may belong to the same category, such as both being caused by a refrigerant pressure sensor failure. Alternatively, the current malfunction may be derived from a previous malfunction.

[0137] In order to quickly and accurately output fault location information, in some embodiments, inputting fused data into the air conditioning diagnostic model to obtain fault location information of the vehicle air conditioning includes the following steps: obtaining historical fault data of the vehicle air conditioning, inputting the historical fault data and fused data into the air conditioning diagnostic model to obtain fault location information of the vehicle air conditioning.

[0138] Historical fault data includes past fault records, maintenance records, and common fault data for the vehicle's air conditioning system. Past fault records include the time of occurrence, fault type, faulty component, repair plan, and the effectiveness of the solution. Maintenance records include the refrigerant type and dosage, replacement or maintenance dates for core components such as the compressor / condenser / expansion valve, and leak detection history in the piping. Common fault data includes frequently occurring, latent fault types for the same vehicle model.

[0139] The process of inputting historical fault data and fused data into the air conditioning diagnostic model to obtain fault location information for the automotive air conditioning system includes the following steps: preprocessing the historical fault data to obtain preprocessed historical fault data; performing feature extraction processing on the preprocessed historical fault data to obtain historical correlation features; splicing the historical correlation features, air conditioning diagnostic features, and air conditioning measured features to obtain target spliced ​​features; and inputting the target spliced ​​features into the air conditioning diagnostic model to obtain fault location information for the automotive air conditioning system.

[0140] The air conditioning diagnostic model is equipped with an attention mechanism module, which concatenates historical correlation features, air conditioning diagnostic features, and air conditioning measured features to obtain the target concatenated features. The steps include: based on the attention mechanism module, weighting the historical correlation features, air conditioning diagnostic features, and air conditioning measured features to obtain weighted historical correlation features, weighted air conditioning diagnostic features, and weighted air conditioning measured features, respectively; and concatenating the weighted historical correlation features, weighted air conditioning diagnostic features, and weighted air conditioning measured features to obtain the target concatenated features.

[0141] Please see Figure 9 The input layer 81 includes an attention mechanism module 84. The input layer 81 outputs historical correlation features, air conditioning diagnostic features, and air conditioning measured features to the attention mechanism module 84. The attention mechanism module 84 performs weighted processing on these features to obtain weighted historical correlation features, weighted air conditioning diagnostic features, and weighted air conditioning measured features, respectively. The attention mechanism module 84 then concatenates these weighted features to obtain the target concatenated features, which are then transmitted to the processing layer 82 for processing and analysis. The output layer 83 is configured with various fault location tags, each corresponding to a specific fault location information. The output layer 83 can map the output of the processing layer 82 to the corresponding fault location tag to output the corresponding fault location information.

[0142] This application embodiment uses historical fault data and fused data as multi-source data input into the air conditioning diagnostic model, enabling the air conditioning diagnostic model to perform multi-dimensional reasoning, solving the diagnostic blind spot of recurring faults in the same vehicle, which is conducive to quickly narrowing the diagnostic scope and generating fault location information efficiently and accurately.

[0143] ② Fine-tune the training of the general large model or the multimodal large model to obtain the air conditioning diagnostic model.

[0144] In some embodiments, the air conditioning diagnostic model is a large language model or a multimodal large model that has been fine-tuned and trained. In this application, the embodiments use training sample data to fine-tune a general large model (such as BERT-Tiny, DistilGPT2, Llama-2-7B) or a multimodal large model with small samples to obtain the air conditioning diagnostic model. This enables the air conditioning diagnostic model to not only understand structured data (such as pressure values ​​and temperature values) and unstructured descriptions (such as fault phenomena), but also to support cross-modal analysis.

[0145] The analysis of air conditioning-related data and air conditioning measurement data based on the air conditioning diagnostic model to obtain fault location information of the vehicle air conditioning includes the following steps: obtaining diagnostic prompt words, inputting air conditioning-related data, air conditioning measurement data and diagnostic prompt words into the air conditioning diagnostic model, so that the air conditioning diagnostic model returns fault location information of the vehicle air conditioning under the prompt of the diagnostic prompt words.

[0146] Diagnostic prompts are used to instruct the air conditioning diagnostic model to analyze air conditioning-related data and measurement data to output fault location information. For example, a diagnostic prompt might be: "As an automotive air conditioning fault diagnosis expert, please analyze the fault type, faulty component, and troubleshooting suggestions for the automotive air conditioning system based on the provided air conditioning-related data and measurement data."

[0147] Please see Figure 10 In this embodiment, air conditioning associated data and air conditioning measurement data are converted into key-value pair structured text. Diagnostic prompt words, key-value pair structured text of air conditioning associated data, and key-value pair structured text of air conditioning measurement data are concatenated in sequence to obtain the model input sequence. The model input sequence is then input into the air conditioning diagnostic model, so that the air conditioning diagnostic model matches the model input sequence with preset fault judgment rules under the prompt of diagnostic prompt words, so as to generate fault location information.

[0148] This application embodiment guides the air conditioning diagnostic model through diagnostic prompts, making full use of the semantic understanding and powerful analysis capabilities of the air conditioning diagnostic model as a large language model or multimodal large model. This enables the air conditioning diagnostic model to perform semantic analysis on air conditioning related data and air conditioning measurement data in order to output accurate fault location information.

[0149] ③ Generate an air conditioning diagnostic model based on the fault location table.

[0150] The air conditioning diagnostic model is equipped with a fault location table, which is a structured judgment rule table built based on the thermodynamic principles of automotive air conditioning, maintenance experience, and historical fault data. The fault location table is used to locate faults in automotive air conditioning by combining air conditioning related data and air conditioning measurement data.

[0151] In some embodiments, the fault location table is a structured judgment rule table. For example, the fault location table contains multiple entries, including feature threshold range, data matching conditions, corresponding fault type, fault priority, and troubleshooting suggestions.

[0152] In other embodiments, the fault location table includes a normal state table and a fault state table. The normal state table is obtained by the designer based on various performance parameters of the air conditioning of different types of cars under normal operating conditions. The normal state table is used to reflect the normal performance range of various performance parameters of the air conditioning of different types of cars under normal operating conditions.

[0153] Please see Figure 11 In this embodiment of the application, under normal operating conditions of the automotive air conditioning system, performance parameters such as high pressure, low pressure, high pressure temperature, low pressure temperature, subcooling, superheating, evaporator temperature, and air outlet temperature are stored in a database according to different ambient temperatures and humidity levels, based on brand, model, year, and configuration, to obtain the following results: Figure 11 The table shown is the normal status table.

[0154] The fault status table is obtained by the designer based on tests of the air conditioning systems of different types of cars under fault conditions. The fault status table is used to reflect the deviation of various performance parameters of the air conditioning systems of different types of cars from the normal performance range under fault conditions.

[0155] Please see Figure 12 In this embodiment of the application, when the automotive air conditioning is in a faulty operating condition, performance parameters such as high pressure, low pressure, high pressure temperature, low pressure temperature, subcooling, superheating, evaporator temperature, and outlet temperature are obtained as follows: Figure 12 The fault status table is shown.

[0156] The steps for analyzing air conditioning-related data and air conditioning measurement data based on the air conditioning diagnostic model to obtain fault location information for automotive air conditioning include: analyzing air conditioning-related data and air conditioning measurement data based on the fault location table to obtain fault location information for automotive air conditioning.

[0157] This application embodiment achieves rapid matching of "feature-fault" through a rule-structured fault location table, eliminating the need for complex deep learning reasoning in the model, significantly improving diagnostic efficiency and the interpretability of fault location, and also allowing for flexible updates or additions to the rules or data in the fault location table.

[0158] Please see Figure 13 In this embodiment of the application, through steps S433 and S434, the air conditioning related data and air conditioning measurement data are analyzed based on the fault location table to obtain the fault location information of the automotive air conditioning, as shown below: Step S433: Based on the normal state table, perform semantic transformation processing on the air conditioner associated data and air conditioner measurement data to obtain a semantic parameter set.

[0159] The semantic parameter set is configured to describe the operating status of the automotive air conditioner across different performance dimensions using semantic expressions. These semantic expressions can include Boolean expressions or other methods. The semantic parameter set includes multiple performance semantic parameters corresponding to different performance dimensions, where each performance semantic parameter represents a deviation from the normal performance range.

[0160] Step S434: Based on the fault status table, the semantic parameter set is matched to obtain the fault location information of the car air conditioner.

[0161] The fault status table includes multiple fault semantic sets and the corresponding fault causes and fault objects for each fault semantic set. Each fault semantic set includes multiple fault semantic parameters corresponding to different performance dimensions. These parameters represent situations where performance parameters deviate from the normal performance range using semantic representation.

[0162] For example, such as Figure 12 As shown, when the high pressure is less than the normal high pressure range, the fault semantic parameter corresponding to the high pressure is "low". When the high pressure is within the normal high pressure range, the fault semantic parameter corresponding to the high pressure is "normal" or "unchanged". When the high pressure is greater than the normal high pressure range, the fault semantic parameter corresponding to the high pressure is "high".

[0163] For another example, when the superheat is less than the normal superheat range, the fault semantic parameter corresponding to the superheat is "low"; when the superheat is within the normal superheat range, the fault semantic parameter corresponding to the superheat is "normal" or "unchanged"; and when the superheat is greater than the normal superheat range, the fault semantic parameter corresponding to the superheat is "high".

[0164] The cause of the malfunction is the reason why the car's air conditioning is failing. Please refer to... Figure 12 ,exist Figure 12 In this context, since there is a positive correlation between evaporator temperature and outlet air temperature—that is, a high evaporator temperature leads to a high outlet air temperature, and a low evaporator temperature leads to a low outlet air temperature—for simplicity, Figure 12 Only the fault semantic parameters for evaporator temperature are shown; the fault semantic parameters for outlet air temperature are not shown. Additionally, Figure 12 The provided "failure modes" can be used to understand the causes of failures.

[0165] When the fault semantic set of "high pressure - high pressure temperature - low pressure - low pressure temperature - superheat - subcooling - evaporator temperature" is {high, high, high, high, low, high, high}, the failure mode (i.e. the cause of the failure) is overfilling of refrigerant.

[0166] When the fault semantic set of "high pressure pressure - high pressure temperature - low pressure pressure - low pressure temperature - superheat - subcooling - evaporator temperature" is {low, low, low, high, high, low, high}, the failure mode (i.e. the cause of the failure) is insufficient or inadequate refrigerant charge.

[0167] The fault object is the component that causes the car air conditioning to malfunction. When the fault semantic set of "high pressure pressure - high pressure temperature - low pressure pressure - low pressure temperature - superheat - subcooling - evaporator temperature" is {high, high, high, high, low, high, high}, the failure mode (i.e. the cause of the fault) is overfilling of refrigerant, and the fault object is the refrigerant.

[0168] When the fault semantic set of "high pressure - high pressure temperature - low pressure - low pressure temperature - superheat - subcooling - evaporator temperature" is {low, low, low, high, high, low, high}, the failure mode (i.e. the cause of the failure) is leakage due to damage to the interface seal or gasket, and the fault object is the compressor.

[0169] Fault location information includes the fault object and the fault cause. For example, a fault could be caused by insufficient or inadequate refrigerant charging. Another example could be a compressor fault caused by damage to the interface seal or gasket, leading to leakage. This application embodiment performs semantic transformation processing on the target parameter set based on a normal state table, obtaining a semantic parameter set that can describe the operating state of the automotive air conditioner from different performance dimensions. This achieves the transformation from a specific target parameter set to a simple and highly generalized semantic parameter set. Finally, the semantic parameter set is matched with the fault state table to automatically generate fault location information without relying on manual intervention, thus improving the accuracy and efficiency of fault location.

[0170] Both air conditioning related data and air conditioning measurement data include multiple performance parameters, each corresponding to a specific performance dimension. One performance parameter is high-pressure, which describes the pressure conditions of the car's air conditioning system in the high-pressure range. Repair personnel can connect a manifold gauge set to the high-pressure interface of the car's air conditioning system to measure the high-pressure. The manifold gauge sends the high-pressure reading to diagnostic equipment, which then obtains the high-pressure reading. Another performance parameter is low-pressure, which describes the pressure conditions of the car's air conditioning system in the low-pressure range. Repair personnel can connect a manifold gauge set to the low-pressure interface of the car's air conditioning system to measure the low-pressure reading. The manifold gauge sends the low-pressure reading to diagnostic equipment, which then obtains the low-pressure reading.

[0171] The performance parameter is high-pressure temperature, which describes the temperature conditions of the car's air conditioning system in the high-pressure range. Repair personnel can use a temperature detection device to measure the compressor's exhaust port temperature and the condenser's outlet temperature. Based on these temperatures, the high-pressure temperature is obtained and sent to diagnostic equipment, which then receives the high-pressure temperature. The performance parameter is low-pressure temperature, which describes the temperature conditions of the car's air conditioning system in the low-pressure range. Repair personnel can use a temperature detection device to measure the compressor's suction pipe to obtain the low-pressure temperature.

[0172] The performance parameter is subcooling, which describes the condition of the liquid refrigerant in the automotive air conditioning system. Repair personnel can use a temperature detection device to measure the temperature of the liquid refrigerant at the condenser outlet and compare it to the refrigerant saturation temperature under the current high pressure. The subcooling is determined based on the difference between the liquid refrigerant temperature and the refrigerant saturation temperature.

[0173] The performance parameter is superheat, which describes the condition of the gaseous refrigerant in the automotive air conditioning system. Repair personnel can use a temperature detection device to measure the temperature of the gaseous refrigerant at the evaporator outlet and the refrigerant saturation temperature at the current low-pressure setting. The superheat is determined based on the difference between the gaseous refrigerant temperature and the refrigerant saturation temperature.

[0174] The performance parameter is evaporator temperature, which describes the surface temperature of the evaporator in the car's air conditioning system. Repair personnel can obtain the evaporator temperature by measuring the average surface temperature of the evaporator core using a temperature detection device. The performance parameter is also called vent temperature, which describes the temperature of the air vents in the car's air conditioning system. Repair personnel can obtain the vent temperature by measuring the air temperature at the air outlets in the air conditioning ducts using a temperature detection device.

[0175] The normal state table includes the normal performance set of different types of automotive air conditioners. The normal performance set is a collection of the normal performance ranges of various performance dimensions of the automotive air conditioner under normal operating conditions. The normal state table includes the normal performance set of the automotive air conditioner for the first model, which is {outlet temperature, high pressure, low pressure, high pressure temperature, low pressure temperature, subcooling, superheating, evaporator temperature} = {(6.0℃, 8.0℃), (1204kPa, 1471kPa), (171kPa, 209kPa), (60℃, 80℃), (1.0℃, 10℃), (1.0℃, 15℃), (2.0℃, 20℃), (3.0℃, 5.0℃)}.

[0176] The normal performance set includes the normal performance range of different performance dimensions of the automotive air conditioner. As mentioned earlier, the normal performance range of the air outlet temperature is (6.0℃, 8.0℃), the normal performance range of the high pressure is (1204kPa, 1471kPa), and so on.

[0177] The semantic parameter set is obtained by performing semantic transformation processing on air conditioning related data and air conditioning measurement data based on the normal state table. The steps include: obtaining the target type information of the car air conditioner; searching for the normal performance set corresponding to the target type information in the normal state table; the normal performance set corresponding to the target type information is the target normal performance set; and performing semantic transformation processing on each performance parameter in the target parameter set based on the multiple normal performance ranges contained in the target normal performance set to obtain the semantic parameter set.

[0178] Target type information is used to identify the vehicle's air conditioning system. This information includes the vehicle model number, brand, and year. Repair personnel can operate diagnostic equipment, manually inputting the target type information into the equipment, which then retrieves this information.

[0179] This application embodiment uses target type information as an index to search the normal performance set corresponding to the target type information in the normal state table, and uses the normal performance set corresponding to the target type information as the target normal performance set. For example, when the target type information is a first vehicle model, this application embodiment selects the following normal performance set Q as the target normal performance set: Q={(6.0℃,8.0℃),(1204kpa,1471kpa),(171kpa,209kpa),(60℃,80℃),(1.0℃,10℃),(1.0℃,15℃),(2.0℃,20℃),(3.0℃,5.0℃)}.

[0180] In this embodiment, the normal performance range corresponding to the performance parameters is selected as the target performance range. Semantic parameters are generated based on the target performance range and the performance parameters. The semantic parameters corresponding to all performance parameters are combined to obtain a semantic parameter set. For example, assuming the target type information is a first vehicle model, when the performance parameter is the air outlet temperature, the target performance range is (6.0℃, 8.0℃). When the performance parameter is high pressure, the target performance range is (1204kPa, 1471kPa). When the performance parameter is low pressure temperature, the target performance range is (1.0℃, 10℃), and so on.

[0181] Performance semantic parameters include normal semantic parameters, increased semantic parameters, and decreased semantic parameters. Normal semantic parameters are those where the performance parameter is within the normal performance range. For example, when the high pressure is within the normal high pressure range, the semantic parameter corresponding to the high pressure is a normal semantic parameter.

[0182] An elevated semantic parameter is a semantic parameter whose performance parameter is greater than the normal performance range. For example, when the high pressure is greater than the normal high pressure range, the semantic parameter corresponding to the high pressure is the elevated semantic parameter.

[0183] A reduced semantic parameter is a semantic parameter whose performance parameter is less than the normal performance range. For example, when the high pressure is less than the normal high pressure range, the semantic parameter corresponding to the high pressure is the reduced semantic parameter.

[0184] When the performance parameter is within the target performance range, this embodiment generates normal semantic parameters; when the performance parameter is greater than the maximum endpoint value of the target performance range, this embodiment generates elevated semantic parameters; when the performance parameter is less than the minimum endpoint value of the target performance range, this embodiment generates degraded semantic parameters.

[0185] For example, when the performance parameter is the outlet temperature and is 7°C, the performance parameter is within the target performance range (6.0°C, 8.0°C), and this embodiment generates a normal semantic parameter. When the performance parameter is the outlet temperature and is 9°C, the performance parameter is greater than the maximum endpoint value of 8.0°C within the target performance range (6.0°C, 8.0°C), and this embodiment generates an increased semantic parameter. When the performance parameter is the outlet temperature and is 5°C, the performance parameter is less than the minimum endpoint value of 6.0°C within the target performance range (6.0°C, 8.0°C), and this embodiment generates a decreased semantic parameter.

[0186] For example, when the high pressure is greater than the normal high pressure range, the high pressure temperature is less than the normal high pressure range, the low pressure is greater than the normal low pressure range, the low pressure temperature is greater than the normal low pressure range, the superheat is greater than the normal superheat range, the subcooling is less than the normal subcooling range, the evaporator temperature is greater than the normal evaporator temperature range, and the outlet temperature is greater than the normal outlet temperature range, the semantic parameter set is {high, low, high, low, high, low, high, high}.

[0187] The fault status table includes multiple fault semantic sets, fault causes and fault objects corresponding to each fault semantic set. The process of matching the semantic parameter set based on the fault status table to obtain the fault location information of the car air conditioner includes the following steps: in response to the existence of a fault semantic set that matches the semantic parameter set in the fault status table, the fault semantic set that matches the semantic parameter set is determined as the target fault semantic set, and the fault location information of the car air conditioner is generated based on the fault causes and fault objects of the target fault semantic set.

[0188] In this embodiment, each performance semantic parameter in the semantic parameter set is traversed and matched in the fault status table. When the semantic parameter set completely matches the fault semantic set, the fault semantic set that matches the semantic parameter set is determined as the target fault semantic set.

[0189] For example, the semantic parameter set B1 is {high, low, high, low, high, low, high, high}, and the fault status table includes a fault semantic set G1 and a fault semantic set G2, wherein the fault semantic set G1 is {low, low, high, low, high, low, high, high}, and the fault semantic set G2 is {high, low, high, low, high, low, high, high}.

[0190] Since the first fault semantic parameter "low" in the fault semantic set G1 does not match the first performance semantic parameter "high" in the semantic parameter set B1, the fault semantic set G1 is not the target fault semantic set of the semantic parameter set B1. Since the fault semantic set G2 is a perfect match with the semantic parameter set B1, the fault semantic set G2 is the target fault semantic set of the semantic parameter set B1.

[0191] This application embodiment generates a fault description statement by matching the fault cause and the fault object, and outputs the fault description statement as fault location information. This application embodiment matches the semantic parameter set in the fault status table to obtain a target fault semantic set that can reliably and accurately match the semantic parameter set. Then, fault location information is generated based on the fault cause and fault object of the target fault semantic set. The entire process can be completed automatically without manual intervention. Furthermore, automotive air conditioning faults often exhibit global or multi-faceted symptoms. The semantic parameter set consists of performance semantic parameters from multiple performance dimensions, assisting in locating automotive air conditioning faults from multiple performance dimensions, thus improving the accuracy of fault location.

[0192] In some embodiments, the fault status table further includes fault feedback information, which describes the fault condition of the car air conditioner from the user's perspective. For example, the fault feedback information may be "poor cooling effect or no cooling at all", or "air conditioner has an odor".

[0193] Fault feedback information corresponds to one or more fault semantic sets. The fault feedback information "poor cooling effect or no cooling at all" corresponds to multiple fault semantic sets. It can be understood that the same fault feedback information can correspond to different fault causes (i.e., failure modes) and fault objects.

[0194] To improve the efficiency of determining the target fault semantic set and narrow the search range in the fault status table, this embodiment of the application can combine customer feedback information to narrow the search range in the fault status table. Specifically, in response to the existence of a fault semantic set in the fault status table that matches the semantic parameter set, determining the fault semantic set that matches the semantic parameter set as the target fault semantic set includes the following steps: obtaining customer feedback information; determining the fault feedback information that matches the customer feedback information as the target fault feedback information in the fault status table; forming a target search range by combining one or more fault semantic sets corresponding to the target fault feedback information; and in response to the existence of a fault semantic set that matches the semantic parameter set in the target search range, determining the fault semantic set that matches the semantic parameter set as the target fault semantic set.

[0195] Customer feedback information is used to describe the general situation of a car air conditioning malfunction. For example, if user A reports to a repair technician that "the air conditioning has an odor," then "the air conditioning has an odor" is the customer feedback information, providing an overall description of the malfunction. The repair technician can input the customer feedback information into the diagnostic equipment based on the user's description, and the equipment will then receive the feedback. Alternatively, the diagnostic equipment can collect the user's voice description of the malfunction and generate customer feedback information based on that voice.

[0196] In some embodiments, the diagnostic device can directly match customer feedback information with each fault feedback information in the fault status table. If the two are completely matched, the fault feedback information that matches the customer feedback information is determined to be the target fault feedback information. If the two are not completely matched, the fault feedback information is determined not to be the target fault feedback information.

[0197] In some embodiments, the diagnostic device collects the user's voice description of the fault, converts the voice into descriptive text, calculates the similarity between each fault feedback message and the descriptive text in the fault status table to obtain content similarity, and selects the fault feedback message with the highest content similarity among the various content similarity values ​​as the target fault feedback message. For example, the target fault feedback message is "the air conditioner is noisy." In this embodiment, the fault semantic set corresponding to the corresponding sequence number and the fault semantic set corresponding to the corresponding sequence number form the target search range.

[0198] This application embodiment matches the semantic parameter set with each fault semantic set within the target search range, determining the fault semantic set that matches the semantic parameter set as the target fault semantic set. By narrowing the search range for the target fault semantic set, this application embodiment improves the efficiency of finding the target fault semantic set, thereby improving the efficiency of fault location.

[0199] When the detection device that collects various performance parameters malfunctions, the collected performance parameters may not match the performance changes caused by the actual fault. For example, a customer reports a strong odor from the car's air conditioning. According to the fault condition corresponding to "strong odor from the air conditioning," the fault semantic parameter for high pressure should be "low." However, the performance semantic parameter obtained based on the actual collected performance parameters would be "high."

[0200] In this situation, the customer's perception or feeling is the most objective and accurate, and this application embodiment chooses to prioritize trusting the customer's perception or feeling. Therefore, in some embodiments, the method further includes: in response to the absence of a fault semantic set matching the semantic parameter set in the fault status table, generating fault location information based on all fault causes within the target search range.

[0201] In this embodiment of the application, if the semantic parameter set does not match the semantic set of all faults in the fault status table, all fault causes and fault objects within the target search range are presented to the customer or maintenance personnel. In this way, even if there is a mismatch, the customer or maintenance personnel can still be provided with reference fault location information, so as to facilitate the customer or maintenance personnel to select the corresponding maintenance strategy or avoid unnecessary maintenance operations.

[0202] The embodiments of this application locate the fault of the car air conditioner by using multiple performance dimensions. The probability of multiple performance parameter detection devices malfunctioning at the same time is relatively low. For semantic parameter sets that contain a small number of inaccurate performance semantic parameters, the fault can still be located accurately and reliably based on such semantic parameter sets.

[0203] Specifically, in response to the absence of a fault semantic set matching the semantic parameter set in the fault status table, presenting all fault causes and fault objects within the target search range includes the following steps: in response to the absence of a fault semantic set matching the semantic parameter set in the fault status table, calculating the semantic similarity between the semantic parameter set and each fault semantic set in the target search range, finding the maximum semantic similarity among the various semantic similarities, and presenting the fault causes and fault objects of the fault semantic set corresponding to the maximum semantic similarity.

[0204] For example, such as Figure 12 As shown, suppose the actual fault of the car air conditioner is that the electromagnetic clutch of the compressor has slipped, resulting in poor cooling effect. The fault semantic set for this situation is {low, low, high, high, high, low, high}. However, the high-pressure detection device malfunctions. Based on the collected target parameter set, the semantic parameter set E1 is obtained as {high, low, high, high, high, low, high}.

[0205] The target search scope includes fault semantic sets G3, G4, G5, and G6. The semantic similarity between semantic parameter set E1 and fault semantic set G3 is 0.4, between E1 and G4 is 0.6, between E1 and G5 is 0.9, and between E1 and G6 is 0.4. The maximum semantic similarity is 0.9. This embodiment presents the fault cause and fault object of fault semantic set G5. Thus, even when the inspection device malfunctions and causes performance parameters to deviate from reality, this embodiment can still present accurate and reliable fault causes and fault objects to customers or maintenance personnel.

[0206] To illustrate the specific working process of the embodiments of this application in detail, the embodiments of this application are combined with... Figure 14 , Figure 15 and Figure 16 This will be explained in detail below: like Figure 14 As shown, the diagnostic equipment obtained the following data for the car's air conditioning system: high pressure 60 psi, high pressure temperature 12°C, subcooling 6.1°C, low pressure 60 psi, low pressure temperature 12°C, and superheat 3.2°C.

[0207] like Figure 15 As shown, the diagnostic equipment obtained the air outlet temperature of the car's air conditioner as 19°C and the evaporator temperature as 3.2°C.

[0208] like Figure 16 As shown, the diagnostic device, based on the above eight performance parameters and combined with the method provided in the embodiments of this application, generates the following... Figure 16 The fault location information shown includes excessive refrigerant.

[0209] Understandably, once the diagnostic equipment outputs fault location information, it can perform fault detection by comprehensively analyzing this information. Specifically, the fault location information includes at least one faulty object and at least one suspected fault point under the faulty object. For example, the faulty object could be a compressor, an internal or external circulation damper motor, or a temperature sensor, etc.

[0210] A suspected fault point is a location where a fault is suspected, and the cause of the fault is listed below. For example, the suspected fault point might be the compressor's refrigerant, and the cause might be excessive refrigerant. Another example is the compressor's motor, and the cause might be motor jamming.

[0211] Please see Figure 17 The method for locating faults in automotive air conditioning systems also includes the following steps: Step S435: Obtain at least one test data point corresponding to the suspected fault point.

[0212] Step S436: Determine the fault detection results for suspected fault points based on the test data.

[0213] Step S437: Determine the target fault point based on the fault detection results of all suspected fault points, wherein the target fault point is the fault point that has been marked as having a fault.

[0214] In step S435, the test data is the data obtained by testing the suspected fault point with one or more individual test items, which is used to reflect whether the suspected fault point actually has a fault.

[0215] Suspected fault points include at least one individual test item, which is a specific test item performed on the suspected fault point. Based on engineering experience, this application summarizes eight main categories of individual test items for automotive air conditioning systems. These eight categories are: sensor detection, refrigerant identification and visual inspection, compressor testing, system leak judgment and leak point detection, system blockage testing, abnormal airflow detection at the air outlet, odor detection at the air outlet, and abnormal noise detection during air conditioning operation.

[0216] Please see Figure 18 In this embodiment of the application, based on different fault points and causes, a single test item to be executed is configured for each fault point, and the individual test items are arranged in order from simple to complex, such as... Figure 18 As shown.

[0217] To better understand this example, this application embodiment, in conjunction with Table 6, introduces multiple suspected fault points and their corresponding individual test items, as shown in Table 6: Table 6

[0218] As shown in Table 6, when the suspected fault point is the refrigerant, and the cause of the fault is insufficient or under-charged refrigerant, the single test item for the refrigerant configuration in this embodiment includes system leak judgment and leak detection. When the suspected fault point is the expansion valve, and the cause of the fault is valve blockage or jamming due to insufficient opening, the single test item for the expansion valve configuration in this embodiment is system blockage test. When the suspected fault point is the evaporator temperature sensor, and the cause of the fault is surface dirt covering the temperature sensor, the single test item for the temperature sensor configuration in this embodiment is sensor detection.

[0219] Obtaining at least one test data point corresponding to a suspected fault point includes the following steps: obtaining the project data generated when each individual test item in the suspected fault point is executed, and obtaining the test data corresponding to the suspected fault point based on the project data of each individual test item.

[0220] The project data is obtained by testing suspected fault points according to individual test items.

[0221] For each individual test item at a suspected fault location, the repair personnel use the corresponding testing tool, equipping it at the location and establishing a communication connection between the tool and the diagnostic equipment. The testing tool performs the test according to the requirements of each individual test item, obtaining the test data, which is then transmitted to the diagnostic equipment. Alternatively, the repair personnel can transfer the data obtained by the testing tool to a storage device, such as a USB flash drive. The USB flash drive is then inserted into the diagnostic equipment, which automatically reads the data from it. Alternatively, the testing tool can upload the data to a cloud server, and the diagnostic equipment can request the cloud server to return the data, or the cloud server can automatically send the data to the diagnostic equipment. Once the diagnostic equipment receives the data for each individual test item, it packages this data into a single test dataset.

[0222] In some embodiments, obtaining the project data generated when each individual test item in the suspected fault point is executed includes the following steps: obtaining historical fault data, determining the suspected fault point to be determined corresponding to the historical fault data from at least one suspected fault point, and obtaining the project data generated when each individual test item in the suspected fault point to be determined is executed.

[0223] Historical fault data includes past fault records, maintenance records, and common fault data for the vehicle's air conditioning system. Past fault records include the time of occurrence, fault type, faulty component, repair plan, and the effectiveness of the solution. Maintenance records include the refrigerant type and dosage, replacement or maintenance dates for core components such as the compressor / condenser / expansion valve, and leak detection history in the piping. Common fault data includes frequently occurring, latent fault types for the same vehicle model.

[0224] This application embodiment can request a cloud server to return pre-stored historical fault data, or extract historical fault data from the OBD fault log, or access the vehicle maintenance system's database to obtain historical fault data.

[0225] Determining a suspected fault point from at least one suspected fault point that corresponds to historical fault data includes the following steps: detecting whether there is a fault point name in the historical fault data that corresponds to the suspected fault point; if it exists, then determining that the suspected fault point is a suspected fault point to be determined corresponding to the historical fault data; if it does not exist, then determining that the suspected fault point is not a suspected fault point to be determined corresponding to the historical fault data.

[0226] This application embodiment takes into account that fault points are prone to recurrence, and previous fault points are likely to recur and be detected as suspected fault points. In order to narrow down the range of suspected fault points and quickly filter out the actual fault point from multiple suspected fault points, this application embodiment uses historical fault data to filter multiple suspected fault points, selects the suspected fault points to be determined, and then performs one or more individual test items on the suspected fault points to be determined, so as to quickly obtain the test data of the suspected fault points to be determined, so as to quickly filter out the actual fault point.

[0227] The process of detecting whether there is a fault point name corresponding to a suspected fault point in historical fault data includes the following steps: extracting fault data that meets the most recent time condition from historical fault data as the data to be analyzed, and detecting whether there is a fault point name corresponding to a suspected fault point in the data to be analyzed.

[0228] The most recent time condition is customized by the designer based on engineering experience. For example, the constituent elements of the most recent time condition are within the most recent six months, the most recent three months, the most recent one month, the most recent one week, or a stepped time period. For example, a stepped time period includes the most recent one week → the most recent one month → the most recent three months → the most recent three months → the most recent six months, etc.

[0229] The constituent elements of the most recent time condition are a stepped time period. Extracting fault data that meets the most recent time condition from historical fault data as the data to be analyzed includes the following steps: Selecting fault data corresponding to different time periods in sequence according to time priority as the data to be analyzed.

[0230] For example, in this embodiment of the application, the fault data of the most recent week is first extracted from the historical fault data as the data to be analyzed. If there is no fault point name corresponding to the suspected fault point in the data to be analyzed, then the fault data of the most recent month is extracted from the historical fault data as the data to be analyzed. If there is no fault point name corresponding to the suspected fault point in the data to be analyzed, then the fault data of the most recent three months is extracted from the historical fault data as the data to be analyzed, and so on.

[0231] Once a recently malfunctioning fault point is repaired, it has a higher probability of recurring the fault compared to other fault points that took longer to repair. Therefore, based on this pattern, this application extracts fault data from different time periods from historical fault data in sequence according to the time rule from recent to distant, and analyzes the data to be analyzed. This is beneficial for quickly screening out the fault point that actually caused the fault from multiple suspected fault points.

[0232] Obtaining the project data generated when each individual test item in a suspected fault point is executed includes the following steps: determining the recommended test order, which indicates the execution order of each individual test item in the suspected fault point; and sequentially obtaining the project data generated when each individual test item in the suspected fault point is executed according to the recommended test order. This embodiment of the application, by following the recommended test order, can determine the fault detection results of the suspected fault point in the shortest possible time, avoid invalid tests, and facilitate the rapid selection of the target fault point.

[0233] Determining the recommended test order involves the following steps: obtaining test evaluation information for each individual test item, and determining the recommended test order based on the test evaluation information for each individual test item.

[0234] Test evaluation information is used to assess the performance of individual test items.

[0235] In some embodiments, the test evaluation information is the test cost, which is the cost incurred in completing a single test item. Determining the recommended test order based on the test evaluation information of each individual test item includes the following steps: determining the recommended test order according to the test costs of each individual test item from low to high.

[0236] In some embodiments, test evaluation information is test workload, which is the amount of work required to complete a single test item. Determining the recommended test order based on the test evaluation information of each individual test item includes the following steps: determining the recommended test order according to the order of test workload of each individual test item from least to most.

[0237] In some embodiments, test evaluation information is test complexity, which is the number of work elements required to complete a single test item. Determining the recommended test order based on the test evaluation information of each individual test item includes the following steps: determining the recommended test order according to the test complexity of each individual test item from simple to complex.

[0238] In some embodiments, the test evaluation information is the test time, which is the time required to complete a single test item. Determining the recommended test order based on the test evaluation information of each individual test item includes the following steps: determining the recommended test order according to the test time of each individual test item in ascending order.

[0239] In some embodiments, the test evaluation information is the probability value of a suspected fault point being detected as having a fault after completing a single test item. Determining the recommended test order based on the test evaluation information of each individual test item includes the following steps: determining the recommended test order according to the probability values ​​of each individual test item from largest to smallest.

[0240] In some embodiments, the test evaluation information includes two or more of the following: test cost, test workload, test complexity, test time, and probability value. Determining the recommended test order based on the test evaluation information of each individual test item includes the following steps: weighting two or more of the following for each individual test item: test cost, test workload, test complexity, test time, and probability value to obtain a comprehensive evaluation value; and determining the recommended test order according to the comprehensive evaluation values ​​of each individual test item from high to low.

[0241] For example, if the variable displacement solenoid valve of the compressor malfunctions, i.e., the suspected fault point is the variable displacement solenoid valve, and the cause of the fault is a pressure relief leak in the regulating valve, resulting in poor cooling performance, this embodiment of the application, based on the above approach, determines that the unidirectional test items belonging to this suspected fault point include compressor testing, refrigerant recovery weighing judgment, and leak detection. Then, according to the test evaluation information of compressor testing, refrigerant recovery weighing judgment, and leak detection, this embodiment of the application obtains the following recommended test order: compressor testing → refrigerant recovery weighing judgment → leak detection.

[0242] For another example, the suspected fault point is the dryer tank, which is clogged. Based on the above methods, this embodiment of the application determines that the unidirectional test items for this suspected fault point include refrigerant visual inspection and clog testing. Then, according to the test evaluation information of the refrigerant visual inspection and clog testing, this embodiment of the application obtains the following recommended test order: refrigerant visual inspection → clog testing.

[0243] This application follows the principle of "testing costs from low to high, testing workload from small to large, testing complexity from simple to complex, testing time from small to large, and probability value from high to low" to formulate a recommended testing order to ensure accurate and rapid screening of target fault points.

[0244] In step S436, the test data includes multiple test results. The test results are the results of the suspected fault point completing a single test item. The test results include a first result and a second result. The first result is the result that the suspected fault point is identified as having a fault after passing the single test item, and the second result is the result that the suspected fault point is identified as not having a fault after passing the single test item.

[0245] Determining the fault detection result for a suspected fault point based on test data includes the following steps: determining whether the test result contains a first result; if it does, the first result is taken as the fault detection result for the suspected fault point; if it does not, the second result is taken as the fault detection result for the suspected fault point.

[0246] In step S437, determining the target fault point based on the fault detection results of all suspected fault points includes the following steps: searching for the existence of a first result among the fault detection results of all suspected fault points; if it exists, determining the suspected fault point corresponding to the first result as the target fault point; if it does not exist, generating information that the car air conditioner is in normal operating condition and generating information that the car air conditioner sensor is abnormal.

[0247] After initially screening out suspected fault points, this application's embodiments proceed to a specialized testing phase. In this phase, the test data from each suspected fault point are combined to identify the actual fault point, thereby helping maintenance personnel adopt accurate repair measures, reducing misdiagnosis, and improving maintenance efficiency.

[0248] To elaborate on the fault location method for automotive air conditioning provided in this application embodiment, before the diagnostic equipment performs automated diagnosis of the automotive air conditioning, the repair personnel verify the fault symptoms based on customer feedback. If a fault is identified in the automotive air conditioning, a visual inspection is performed. If the visual inspection confirms a fault, the automated diagnosis phase begins. If no fault is identified in the automotive air conditioning, the diagnostic operation ends. This application embodiment combines... Figure 19 The automated diagnostic process is explained in detail below: S191, Perform OBD diagnostic procedure.

[0249] S192, Is an air conditioning diagnostic code detected? If yes, proceed to step S193; otherwise, proceed to step S197.

[0250] S193, execute the diagnostic steps corresponding to the air conditioning diagnostic code.

[0251] S194, diagnosis is made through motion testing or OBD data stream or in combination with actual operation.

[0252] S195, if the diagnosis is normal, perform several more tests to confirm whether it is an intermittent fault, and then end the operation.

[0253] S196, when an abnormality is diagnosed, the fault is detected according to the specified instructions: 1. Check the plug connection; 2. Check the component resistance; 3. Check the component power supply and grounding; 4. Check the control module.

[0254] S197 generates fault location information based on air conditioning associated data and air conditioning measurement data.

[0255] S198, based on the suspected fault point given by the fault location information, enter the test process of the single test item.

[0256] S199, execute each individual test item according to the recommended test order.

[0257] S1910, Select the target fault point, and repair, replace or eliminate the faulty object at the target fault point.

[0258] S1911, after repair and confirmation, perform vehicle diagnostics on the car air conditioning.

[0259] S1912, check if there is an air conditioning diagnostic code or fault symptoms. If there is an air conditioning diagnostic code, proceed to step S194. If there are fault symptoms, proceed to step S197. If there is no air conditioning diagnostic code or fault symptoms, end the operation.

[0260] S1913, Operation terminated.

[0261] In summary, the embodiments of this application have at least the following technical effects: 1) This application embodiment automatically generates fault location information for automotive air conditioning by combining multiple data sources, which helps to improve the accuracy and efficiency of fault location.

[0262] 2) This application embodiment utilizes an air conditioning diagnostic model, taking air conditioning related data and air conditioning measurement data as input, to automatically generate fault location information for automotive air conditioning, which helps to improve the accuracy and efficiency of fault location.

[0263] 3) By utilizing the target parameter set of the automotive air conditioning system, the cause of the fault can be automatically located by combining the normal state table and the fault state table, which helps to improve the accuracy and efficiency of fault location.

[0264] 4) Based on the normal state table, the target parameter set is semantically transformed to obtain a semantic parameter set that can describe the working state of the car air conditioner from different performance dimensions. This realizes the transformation from a specific target parameter set to a simple and generalized semantic parameter set. Finally, the semantic parameter set is matched with the fault state table to automatically generate fault location information without relying on manual intervention. This helps to improve the accuracy and efficiency of fault location.

[0265] 5) The fault manifestations of car air conditioning often present a global or multi-faceted manifestation. The semantic parameter set is composed of performance semantic parameters of multiple performance dimensions. It assists in locating the fault of car air conditioning from multiple performance dimensions, so as to improve the accuracy of fault location.

[0266] 6) Under the premise that the semantic parameter set does not match all fault semantic sets in the fault status table, it can provide customers or maintenance personnel with reference fault location information, so as to facilitate customers or maintenance personnel to select the corresponding maintenance strategy or avoid unnecessary maintenance operations.

[0267] 7) Even if the performance parameters deviate from the actual value due to abnormality of the inspection device, it can still present the customer or maintenance personnel with accurate and reliable fault causes and fault objects.

[0268] It should be noted that in the above embodiments, there is no necessarily a certain order between the steps. Those skilled in the art can understand from the description of the embodiments of this application that the above steps may have different execution orders in different embodiments, that is, they may be executed in parallel or in turn, etc.

[0269] As another aspect of this application, this application provides a fault location device for an automotive air conditioner. The fault location device for the automotive air conditioner can be a software module, which includes several instructions stored in a memory. A processor can access the memory, call the instructions, and execute them to complete the fault location method for the automotive air conditioner described in the above embodiments.

[0270] In some embodiments, the fault location device for automotive air conditioning can also be constructed from hardware components. For example, the fault location device for automotive air conditioning can be constructed from one or more chips, which can work in coordination to complete the fault location method for automotive air conditioning described in the various embodiments above. As another example, the fault location device for automotive air conditioning can also be constructed from various logic devices, such as general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), microcontrollers, ARM (Acorn RISC Machine) or other programmable logic devices, discrete gate or transistor logic, discrete hardware components, or any combination of these components.

[0271] Please see Figure 20 The automotive air conditioning fault location device 2000 includes: a data acquisition module 201, a data parsing module 202, and a fault location module 203. The data acquisition module 201 acquires automotive diagnostic data collected by diagnostic equipment and air conditioning measurement data collected by air conditioning-specific tools. The data parsing module 202 extracts air conditioning-related data corresponding to the automotive air conditioning system from the automotive diagnostic data. The fault location module 203 generates fault location information for the automotive air conditioning system based on the air conditioning-related data and the air conditioning measurement data.

[0272] In some embodiments, the fault location module 203 is specifically used to: in response to the air conditioning associated data not containing an air conditioning fault code corresponding to the vehicle air conditioning, obtain a pre-trained air conditioning diagnostic model, analyze the air conditioning associated data and air conditioning measurement data based on the air conditioning diagnostic model, and obtain fault location information of the vehicle air conditioning.

[0273] In some embodiments, the fault location module 203 is further specifically used to: fuse air conditioning associated data and air conditioning measurement data to obtain fused data, input the fused data into the air conditioning diagnostic model, and obtain fault location information of the vehicle air conditioning.

[0274] In some embodiments, the fault location module 203 is further configured to: normalize the air conditioning associated data and the air conditioning measurement data to obtain normalized air conditioning associated data and normalized air conditioning measurement data respectively; extract air conditioning diagnostic features based on the normalized air conditioning associated data; extract air conditioning measured features based on the normalized air conditioning measurement data; and fuse the air conditioning diagnostic features and the air conditioning measured features to obtain fused data.

[0275] In some embodiments, the fault location module 203 is further specifically used to: acquire historical fault data of the vehicle air conditioner, input the historical fault data and fused data into the air conditioner diagnostic model, and obtain fault location information of the vehicle air conditioner.

[0276] In some embodiments, the air conditioning diagnostic model is a large language model or a multimodal large model that has been fine-tuned and trained. The fault location module 203 is specifically used to: obtain diagnostic prompt words, input air conditioning related data, air conditioning measurement data and diagnostic prompt words into the air conditioning diagnostic model, so that the air conditioning diagnostic model returns the fault location information of the car air conditioning under the prompt of the diagnostic prompt words.

[0277] In some embodiments, the air conditioning diagnostic model is configured with a fault location table, and the fault location module 203 is specifically used to: analyze the air conditioning related data and air conditioning measurement data based on the fault location table to obtain the fault location information of the car air conditioning.

[0278] In some embodiments, the fault location table includes a normal state table and a fault state table. The fault location module 203 is further specifically used to: perform semantic conversion processing on the air conditioning associated data and air conditioning measurement data based on the normal state table to obtain a semantic parameter set. The semantic parameter set is configured to describe the working status of the car air conditioning in different performance dimensions using a semantic expression method. The semantic parameter set is matched based on the fault state table to obtain the fault location information of the car air conditioning.

[0279] In some embodiments, both the air conditioning associated data and the air conditioning measurement data include multiple performance parameters. The normal state table includes normal performance sets for different types of automotive air conditioners. The normal performance sets include normal performance ranges for different performance dimensions of the automotive air conditioners. The fault location module 203 is also specifically used to: obtain target type information of the automotive air conditioner; search for the normal performance set corresponding to the target type information in the normal state table; the normal performance set corresponding to the target type information is the target normal performance set; and based on the multiple normal performance ranges contained in the target normal performance set, perform semantic transformation processing on each performance parameter in the target parameter set to obtain a semantic parameter set.

[0280] In some embodiments, the fault status table includes multiple fault semantic sets, fault causes and fault objects corresponding to each fault semantic set, and the fault location module 203 is further specifically used to: in response to the existence of a fault semantic set that matches the semantic parameter set in the fault status table, determine the fault semantic set that matches the semantic parameter set as the target fault semantic set, and generate fault location information of the automotive air conditioner based on the fault causes and fault objects of the target fault semantic set.

[0281] In some embodiments, the fault location information includes at least one fault object and at least one suspected fault point under the fault object. The fault location module 203 is further specifically used to: acquire at least one test data corresponding to the suspected fault point, determine the fault detection result about the suspected fault point based on the test data, and determine the target fault point based on the fault detection results of all suspected fault points, wherein the target fault point is the fault point marked as having a fault.

[0282] In some embodiments, the suspected fault point includes at least one individual test item, and the fault location module 203 is further specifically used to: obtain the project data generated when each individual test item in the suspected fault point is executed, and obtain the test data corresponding to the suspected fault point based on the project data of each individual test item.

[0283] In some embodiments, the fault location module 203 is further specifically used to: acquire historical fault data, determine the suspected fault point to be determined corresponding to the historical fault data from at least one suspected fault point, and acquire the project data generated when each individual test item in the suspected fault point to be determined is executed.

[0284] In some embodiments, the fault location module 203 is further specifically used to: determine the recommended test order, which represents the execution order of each individual test item in the suspected fault point, and sequentially obtain the project data generated when each individual test item in the suspected fault point is executed according to the recommended test order.

[0285] In some embodiments, the fault location module 203 is further configured to: obtain test evaluation information for individual test items, and determine the recommended test order based on the test evaluation information for each individual test item.

[0286] In some embodiments, the fault location module 203 is further specifically used to: in response to the air conditioning associated data containing an air conditioning fault code corresponding to the vehicle air conditioning, generate fault location information for the vehicle air conditioning based on the air conditioning fault code corresponding to the vehicle air conditioning.

[0287] It should be noted that the above-mentioned automotive air conditioning fault location device can execute the automotive air conditioning fault location method provided in the embodiments of this application, and has the corresponding functional modules and beneficial effects of the method. Technical details not described in detail in the embodiments of the automotive air conditioning fault location device can be found in the automotive air conditioning fault location method provided in the embodiments of this application.

[0288] See Figure 21 , Figure 21 This is a schematic diagram of a diagnostic host provided in an embodiment of this application. The diagnostic host 31 includes one or more processors 311 and a memory 312. The memory 312 is connected to one or more processors 311, for example, via a bus.

[0289] Processor 311 is configured to support the diagnostic device in performing the corresponding functions in the methods described in the above method embodiments. The processor may be a central processing unit (CPU), a network processor (NP), a hardware chip, or any combination thereof. The aforementioned hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The aforementioned PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0290] Memory 312 is used to store program code, etc. Memory may include volatile memory (VM), such as random access memory (RAM); memory may also include non-volatile memory (NVM), such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid-state drive (SSD); memory may also include combinations of the above types of memory.

[0291] The memory 312 can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as the program instructions / modules corresponding to the fault location method for automotive air conditioning in the embodiments of this application. The processor executes the non-volatile software programs, instructions, and modules stored in the memory to perform various functional applications and data processing of the fault location method and fault location device for automotive air conditioning, thereby realizing the functions of each module or unit of the fault location method and fault location device for automotive air conditioning provided in the above method embodiments.

[0292] The memory 312 may include a program storage area and a data storage area, wherein the program storage area may store the operating system and application programs required for at least one function. The data storage area may store data created based on the use of the fault location device of the automotive air conditioning system. In some embodiments, the memory may optionally include memory remotely located relative to the processor, which can be connected to the fault location device of the automotive air conditioning system via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0293] The one or more modules are stored in the memory. When executed by the one or more processors, they perform the fault location method for the automotive air conditioner in any of the above method embodiments. For example, they perform the method steps described in the above method embodiments to realize the functions of the modules described in the above device embodiments.

[0294] This application also provides a computer-readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a diagnostic host, cause the diagnostic host to perform the method described in the foregoing embodiments.

[0295] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0296] The above-disclosed embodiments are merely preferred embodiments of this application and should not be construed as limiting the scope of this application. Therefore, any equivalent variations made in accordance with the claims of this application shall still fall within the scope of this application.

Claims

1. A method for locating faults in an automotive air conditioning system, characterized in that, include: Acquire vehicle diagnostic data collected by diagnostic equipment and air conditioning measurement data collected by air conditioning-specific tools; Parse the vehicle diagnostic data to extract the air conditioning-related data corresponding to the vehicle's air conditioning system; Based on the air conditioning association data and the air conditioning measurement data, fault location information of the car air conditioning is generated.

2. The fault location method according to claim 1, characterized in that, The step of generating fault location information for the vehicle's air conditioning system based on the air conditioning correlation data and the air conditioning measurement data includes: In response to the fact that the air conditioning associated data does not contain an air conditioning fault code corresponding to the car air conditioning, a pre-trained air conditioning diagnostic model is obtained. Based on the air conditioning diagnostic model, the associated data and measurement data of the air conditioning are analyzed to obtain the fault location information of the car air conditioning.

3. The fault location method according to claim 2, characterized in that, The analysis of the air conditioning-related data and the air conditioning measurement data based on the air conditioning diagnostic model yields the fault types of the vehicle air conditioning, including: The air conditioner-related data and the air conditioner measurement data are fused together to obtain fused data; The fused data is input into the air conditioning diagnostic model to obtain the fault location information of the car air conditioning.

4. The fault location method according to claim 3, characterized in that, The process of fusing the air conditioner-related data with the air conditioner measurement data to obtain fused data includes: The air conditioner correlation data and the air conditioner measurement data are normalized to obtain normalized air conditioner correlation data and normalized air conditioner measurement data, respectively. Based on the normalized air conditioner correlation data, air conditioner diagnostic features are extracted; Based on the normalized air conditioning measurement data, the measured features of the air conditioning system are extracted. The diagnostic features of the air conditioner are fused with the measured features of the air conditioner to obtain fused data.

5. The fault location method according to claim 3, characterized in that, The step of inputting the fused data into the air conditioning diagnostic model to obtain the fault location information of the vehicle air conditioning includes: Obtain historical fault data of the vehicle air conditioning system; The historical fault data and the fused data are input into the air conditioning diagnostic model to obtain the fault location information of the car air conditioning.

6. The fault location method according to claim 2, characterized in that, The air conditioning diagnostic model is a finely tuned large language model or a multimodal large model. The analysis of the air conditioning correlation data and the air conditioning measurement data based on the air conditioning diagnostic model to obtain the fault location information of the vehicle air conditioning includes: Retrieve diagnostic prompts; The air conditioning associated data, the air conditioning measurement data, and the diagnostic prompts are input into the air conditioning diagnostic model so that the air conditioning diagnostic model returns fault location information of the vehicle air conditioning under the prompts of the diagnostic prompts.

7. The fault location method according to claim 2, characterized in that, The air conditioning diagnostic model is configured with a fault location table. The analysis of the air conditioning-related data and the air conditioning measurement data based on the air conditioning diagnostic model yields the fault location information of the vehicle's air conditioning system, including: Based on the fault location table, the air conditioning related data and the air conditioning measurement data are analyzed to obtain the fault location information of the car air conditioning.

8. The fault location method according to claim 7, characterized in that, The fault location table includes a normal state table and a fault state table. The analysis of the air conditioning-related data and the air conditioning measurement data based on the fault location table to obtain the fault location information of the vehicle air conditioning includes: Based on the normal state table, semantic transformation processing is performed on the air conditioning associated data and the air conditioning measurement data to obtain a semantic parameter set. The semantic parameter set is configured to describe the working state of the car air conditioning in different performance dimensions using semantic expression. Based on the fault status table, the semantic parameter set is matched to obtain the fault location information of the car air conditioner.

9. The fault location method according to claim 8, characterized in that, Both the air conditioning associated data and the air conditioning measurement data include multiple performance parameters. The normal state table includes normal performance sets for different types of automotive air conditioners, and the normal performance sets include normal performance ranges for different performance dimensions of the automotive air conditioners. The semantic transformation processing of the air conditioning associated data and the air conditioning measurement data based on the normal state table yields a semantic parameter set, including: Obtain the target type information of the automotive air conditioner; Search the normal performance set corresponding to the target type information in the normal status table. The normal performance set corresponding to the target type information is the target normal performance set. Based on the multiple normal performance ranges included in the target normal performance set, semantic transformation processing is performed on each performance parameter in the target parameter set to obtain a semantic parameter set.

10. The fault location method according to claim 8, characterized in that, The fault status table includes multiple fault semantic sets, fault causes and fault objects corresponding to each fault semantic set, and the step of matching the semantic parameter sets based on the fault status table to obtain the fault location information of the vehicle air conditioner includes: In response to the existence of a fault semantic set that matches the semantic parameter set in the fault status table, the fault semantic set that matches the semantic parameter set is determined as the target fault semantic set; Based on the fault causes and fault objects in the target fault semantic set, fault location information of the automotive air conditioner is generated.

11. The fault location method according to any one of claims 1 to 10, characterized in that, The fault location information includes at least one faulty object and at least one suspected fault point under the faulty object, and the method further includes: Obtain at least one test data point corresponding to the suspected fault point; Based on the test data, determine the fault detection results for the suspected fault points; The target fault point is determined based on the fault detection results of all suspected fault points, wherein the target fault point is the fault point that has been marked as having a fault.

12. The fault location method according to claim 11, characterized in that, The suspected fault point includes at least one individual test item, and obtaining at least one test data corresponding to the suspected fault point includes: Obtain the project data generated when each individual test item in the suspected fault point is executed; Based on the project data of each individual test item, test data corresponding to the suspected fault point is obtained.

13. The fault location method according to claim 12, characterized in that, The process of obtaining the project data generated when each individual test item in the suspected fault point is executed includes: Obtain historical fault data; From the at least one suspected fault point, determine the suspected fault point to be determined that corresponds to the historical fault data; Obtain the project data generated when each of the individual test items in the suspected fault points to be determined is executed.

14. The fault location method according to claim 12, characterized in that, The process of obtaining the project data generated when each individual test item in the suspected fault point is executed includes: Determine the recommended test order, which indicates the execution order of each individual test item in the suspected fault point; According to the recommended test order, the project data generated when each individual test item in the suspected fault point is executed is obtained in sequence.

15. The fault location method according to claim 14, characterized in that, The determination of the recommended test order includes: Obtain the test evaluation information for the individual test item; The recommended test order is determined based on the test evaluation information of each individual test item.

16. The fault location method according to claim 2, characterized in that, The step of generating fault location information for the vehicle air conditioning system based on the air conditioning correlation data and the air conditioning measurement data further includes: In response to the air conditioning associated data containing an air conditioning fault code corresponding to the vehicle air conditioning, fault location information of the vehicle air conditioning is generated based on the air conditioning fault code corresponding to the vehicle air conditioning.

17. A diagnostic device, characterized in that, The device includes a memory and a processor, the memory being connected to the processor, the processor being configured to execute one or more computer programs stored in the memory, the processor causing the diagnostic device to implement the fault location method for automotive air conditioning as described in any one of claims 1-16 when executing the one or more computer programs.

18. A fault location system for an automotive air conditioner, characterized in that, include: Air conditioner-specific tools are configured to collect air conditioner measurement data; The diagnostic device as described in claim 17 is communicatively connected to the air conditioning-specific tool.

19. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, the computer program including program instructions, which, when executed by a processor, cause the processor to perform the fault location method for an automotive air conditioner as described in any one of claims 1-16.

Images