Air suspension distribution valve fault diagnosis method, device and equipment and storage medium

Abstract

According to the air suspension distribution valve fault diagnosis method, device and equipment and the storage medium, firstly, multi-dimensional working condition data in the operation process of an air suspension system are collected through corresponding sensing equipment, and the multi-dimensional working condition data can comprehensively represent the actual states of the air suspension system under different operation working conditions; various system operation state parameters strongly associated with the working state of the air suspension distribution valve can be covered; the multi-dimensional working condition data collected in real time are input into a preset fault diagnosis neural network model, and the fault diagnosis neural network model is trained in advance based on historical multi-dimensional working condition data of the air suspension system and can adapt to parameter fluctuation characteristics under different working conditions; fault state recognition can be completed according to internal characteristics of input real-time working condition data, and a corresponding diagnosis result is output. According to the method, the limitation of a fixed judgment threshold in a traditional fault diagnosis scheme is broken through, and the one-sidedness of single parameter monitoring can be eliminated through comprehensive acquisition of multi-dimensional working condition data.

Description

Technical Field

[0001] This application relates to the field of air suspension, specifically to a method, apparatus, equipment, and storage medium for diagnosing faults in an air suspension distribution valve. Background Technology

[0002] As a key hydraulic / pneumatic control component of the air suspension system, the air suspension distribution valve's working state directly determines the inflation and deflation efficiency and accuracy of the air springs, thus affecting the response speed and stability of vehicle height adjustment. Therefore, efficient and accurate fault diagnosis of the air suspension distribution valve is a core requirement for ensuring the reliable operation of the air suspension system, reducing overall vehicle maintenance costs, and improving driving safety. It is also a key research and development direction in the field of automotive chassis electronic control technology.

[0003] In related technologies, the presence of abnormal malfunctions in the distribution valve is determined by collecting the suspension height change value and the distribution valve pipeline pressure value during the air spring inflation and deflation processes. Specifically, during the air spring inflation phase, if the suspension height change value is consistently less than a preset value and the distribution valve pipeline pressure value is consistently greater than the upper pressure limit setting value, the distribution valve is determined to be faulty; during the air spring deflation phase, if the suspension height change value is consistently less than a preset value and the distribution valve pipeline pressure value is consistently less than the upper pressure limit setting value, the distribution valve is determined to be faulty.

[0004] However, using a fixed fault judgment threshold cannot adapt to the parameter fluctuation characteristics under different operating conditions, and is prone to false or missed fault reports, making it difficult to meet the accuracy and reliability requirements for fault diagnosis of the distribution valve in complex vehicle operating scenarios. Summary of the Invention

[0005] This application provides a method, apparatus, equipment, and storage medium for diagnosing faults in an air suspension distribution valve. It can solve the technical problem in related technologies where a fixed fault judgment threshold is used, which cannot adapt to the parameter fluctuation characteristics under different operating conditions and is prone to false or missed fault reports.

[0006] In a first aspect, embodiments of this application provide a method for diagnosing faults in an air suspension distribution valve, the method comprising:

[0007] Collect multi-dimensional operating condition data during the operation of the air suspension system; The real-time collected multi-dimensional operating condition data is input into a preset fault diagnosis neural network model, and the fault diagnosis result is output through the fault diagnosis neural network model.

[0008] In conjunction with the first aspect, in one embodiment, before collecting multi-dimensional operating condition data during the operation of the air suspension system, the method further includes: Acquire historical multi-dimensional operating condition data of the air suspension system, preprocess the historical multi-dimensional operating condition data, and map the historical operating condition data of each dimension to the same numerical range. The historical multi-dimensional operating condition data is used to determine fault conditions and label the operating condition types to obtain a training dataset. The fault diagnosis neural network model is trained based on the training dataset.

[0009] In conjunction with the first aspect, in one implementation, the step of determining fault conditions and labeling the condition type based on the historical multi-dimensional operating condition data includes: If the air pump temperature is stable and the air spring adjustment function is completed, and the absolute value of the deviation between the height of a single air spring and the average height of the remaining air springs exceeds the preset deviation range, then this set of historical multi-dimensional operating condition data will be marked as fault operating condition data.

[0010] In conjunction with the first aspect, in one implementation, the step of determining fault conditions and labeling the condition type based on the historical multi-dimensional operating condition data includes: If the air pump temperature is stable and the air spring adjustment function is completed, after a preset settling time, if the deviation between the target height and the actual height of the air spring exceeds the preset height deviation range, then this set of historical multi-dimensional operating condition data will be marked as fault operating condition data.

[0011] In conjunction with the first aspect, in one implementation, the step of determining fault conditions and labeling the condition type based on the historical multi-dimensional operating condition data includes: If the air pump temperature is stable, the air spring adjustment time exceeds the preset time threshold, and the deviation between the actual height of the air spring and the target height exceeds the preset height deviation range, then this set of historical multi-dimensional operating condition data will be marked as fault operating condition data.

[0012] In conjunction with the first aspect, in one implementation, the multi-dimensional operating condition data includes external ambient temperature data, air pump operating temperature data, height change data at different positions of the vehicle body, air tank pressure data, and air pump operating time data.

[0013] In conjunction with the first aspect, in one implementation, the step of inputting real-time collected multi-dimensional operating condition data into a preset fault diagnosis neural network model includes: The multi-dimensional working condition data is preprocessed to map the working condition data of each dimension to the same numerical range in order to eliminate the impact of different data volumes on subsequent diagnosis. The preprocessed multi-dimensional operating condition data is input into a preset fault diagnosis neural network model.

[0014] Secondly, this application provides an air suspension distribution valve fault diagnosis device, which includes: a data acquisition module for acquiring multi-dimensional operating condition data during the operation of the air suspension system; and a diagnosis output module for inputting the real-time acquired multi-dimensional operating condition data into a preset fault diagnosis neural network model and outputting fault diagnosis results through the fault diagnosis neural network model.

[0015] Thirdly, embodiments of this application provide an air suspension distribution valve fault diagnosis device, which includes a processor, a memory, and an air suspension distribution valve fault diagnosis program stored in the memory and executable by the processor. When the air suspension distribution valve fault diagnosis program is executed by the processor, it implements the steps of the air suspension distribution valve fault diagnosis method as described in some of the above embodiments.

[0016] Fourthly, embodiments of this application provide a computer-readable storage medium storing an air suspension distribution valve fault diagnosis program, wherein when the air suspension distribution valve fault diagnosis program is executed by a processor, it implements the steps of the air suspension distribution valve fault diagnosis method as described in some of the above embodiments.

[0017] The beneficial effects of the technical solutions provided in this application include: First, multi-dimensional operating condition data of the air suspension system is collected using corresponding sensing devices. This data comprehensively characterizes the actual state of the air suspension system under different operating conditions and covers various system operating parameters strongly correlated with the working state of the air suspension distribution valve. Then, the real-time collected multi-dimensional operating condition data is input into a pre-set fault diagnosis neural network model. This model is pre-trained based on historical multi-dimensional operating condition data of the air suspension system, adaptable to parameter fluctuations under different operating conditions, and can identify fault states and output corresponding diagnostic results based on the inherent characteristics of the input real-time operating condition data. This method breaks through the limitations of fixed judgment thresholds in traditional fault diagnosis schemes. By comprehensively collecting multi-dimensional operating condition data, the one-sidedness of single-parameter monitoring can be eliminated. Combined with a fault diagnosis neural network model with operating condition adaptability, it can effectively avoid the problems of missed or false alarms caused by parameter fluctuations under different operating conditions. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating an embodiment of the air suspension distribution valve fault diagnosis method of this application; Figure 2 This is a schematic diagram of the hardware structure of the air suspension distribution valve fault diagnosis device involved in the embodiments of this application. Detailed Implementation

[0019] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0020] In the field of automotive chassis electronic control technology, with consumers' increasing demands for vehicle driving comfort, passability, and handling stability, air suspension systems have become one of the core configurations for mid-to-high-end passenger and commercial vehicles. As a key hydraulic / pneumatic control component of the air suspension system, the air suspension distribution valve's working state directly determines the inflation and deflation efficiency and accuracy of the air springs, thus affecting the response speed and stability of vehicle height adjustment. Therefore, efficient and accurate fault diagnosis of the air suspension distribution valve is a core requirement for ensuring the reliable operation of the air suspension system, reducing overall vehicle after-sales maintenance costs, and improving driving safety. It is also a key research and development direction in the current field of automotive chassis electronic control technology.

[0021] In related technologies, the presence of abnormal malfunctions in the distribution valve is determined by collecting the suspension height change value and the distribution valve pipeline pressure value during the air spring inflation and deflation processes. Specifically, during the air spring inflation phase, if the suspension height change value is consistently less than a preset value and the distribution valve pipeline pressure value is consistently greater than the upper pressure limit setting value, the distribution valve is determined to be faulty; during the air spring deflation phase, if the suspension height change value is consistently less than a preset value and the distribution valve pipeline pressure value is consistently less than the upper pressure limit setting value, the distribution valve is determined to be faulty.

[0022] However, using a fixed fault judgment threshold cannot adapt to the parameter fluctuation characteristics under different operating conditions, and is prone to false or missed fault reports, making it difficult to meet the accuracy and reliability requirements for fault diagnosis of the distribution valve in complex vehicle operating scenarios.

[0023] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0024] In a first aspect, embodiments of this application provide a method for diagnosing faults in an air suspension distribution valve.

[0025] In one embodiment, reference is made to Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the air suspension distribution valve fault diagnosis method of this application. Figure 1As shown, the air suspension distribution valve fault diagnosis method includes: S100: Collects multi-dimensional operating condition data during the operation of the air suspension system; S200: Input the real-time collected multi-dimensional operating condition data into the preset fault diagnosis neural network model, and output the fault diagnosis result through the fault diagnosis neural network model.

[0026] In this embodiment, multi-dimensional operating condition data of the air suspension system during operation is first collected using corresponding sensing devices. This multi-dimensional operating condition data comprehensively characterizes the actual state of the air suspension system under different operating conditions and covers various system operating state parameters strongly correlated with the working state of the air suspension distribution valve. Then, the real-time collected multi-dimensional operating condition data is input into a preset fault diagnosis neural network model. This model is pre-trained based on historical multi-dimensional operating condition data of the air suspension system and can adapt to parameter fluctuations under different operating conditions. It can identify fault states and output corresponding diagnostic results based on the inherent characteristics of the input real-time operating condition data. This method breaks through the limitations of fixed judgment thresholds in traditional fault diagnosis schemes. By comprehensively collecting multi-dimensional operating condition data, the one-sidedness of single-parameter monitoring can be eliminated. Combined with a fault diagnosis neural network model with operating condition adaptability, it can effectively avoid the problems of missed or false alarms caused by parameter fluctuations under different operating conditions, improve the accuracy and reliability of air suspension distribution valve fault diagnosis, and thus ensure the stable operation of the air suspension system, reducing the manpower and material costs of vehicle after-sales maintenance.

[0027] Furthermore, in one embodiment, before S100, there is also S000, which includes the following steps: S001: Obtain historical multi-dimensional operating condition data of the air suspension system, preprocess the historical multi-dimensional operating condition data, and map the historical operating condition data of each dimension to the same numerical range. S002: Determine the fault conditions and label the condition types on the historical multi-dimensional operating data to obtain a training dataset; S003: The fault diagnosis neural network model is trained based on the training dataset.

[0028] In this embodiment, before performing the multi-dimensional operating condition data acquisition step, historical multi-dimensional operating condition data of the air suspension system is first acquired. Preprocessing is then performed on the historical multi-dimensional operating condition data to uniformly map the historical operating condition data of each dimension to the same numerical range, thereby eliminating the adverse interference caused by the difference in magnitude between different dimensions of data on subsequent operating condition judgment and model training. Then, fault operating condition judgment is carried out on the preprocessed historical multi-dimensional operating condition data, and the data is labeled with operating condition type according to the judgment results to form a dataset that can be used for model training. Finally, the fault diagnosis neural network model is trained based on this training dataset, so that the model can establish the corresponding fault identification logic based on the inherent state correlation of historical operating condition data. This preliminary model training preparation and process enables the subsequent fault diagnosis neural network model used for real-time diagnosis to adapt to parameter fluctuations under different operating conditions. When the model receives real-time operating data, it can make accurate judgments based on the pre-established fault identification logic. This effectively eliminates the diagnostic limitations caused by the lack of adaptive diagnostic logic based on historical operating data and the use of fixed thresholds in traditional diagnostic schemes. It provides a reliable model foundation for the entire fault diagnosis method, further enhancing the accuracy and stability of air suspension distribution valve fault diagnosis and providing stronger technical support for the reliable operation of air suspension systems.

[0029] Furthermore, in one embodiment, step S002 includes the following steps: S002-1: If the air pump temperature is stable and the air spring adjustment function is completed, and the absolute value of the deviation between the height of a single air spring and the average height of the remaining air springs exceeds the preset deviation range, then the set of historical multi-dimensional operating condition data will be marked as fault operating condition data.

[0030] In this embodiment, during the fault condition determination process of preprocessed historical multi-dimensional operating condition data, for operating scenarios where the air pump temperature remains stable and the air spring adjustment function is completed, the absolute value of the deviation between the height of a single air spring and the average height of the remaining air springs is used as the determination criterion. If the absolute value of this deviation exceeds a preset deviation range, the historical multi-dimensional operating condition data of the corresponding group is marked as fault condition data, thus defining a clear fault data category for the training dataset. This determination logic can accurately identify the operating condition data corresponding to vehicle height imbalance caused by a faulty distribution valve after the air spring adjustment is completed. This provides sample data with clear fault characteristics for the training of the fault diagnosis neural network model, enabling the model to learn the parameter correlation rules under this type of fault condition. Consequently, the fault diagnosis neural network model subsequently deployed can quickly identify the fault state when encountering similar operating conditions, effectively compensating for the diagnostic loopholes in traditional diagnostic solutions due to the lack of targeted identification of this type of fault characteristic. This lays a data foundation for improving the comprehensiveness and accuracy of overall fault diagnosis, further ensuring the stability of the air suspension system operation.

[0031] Furthermore, in one embodiment, step S002 includes the following steps: S002-2: If the air pump temperature is stable and the air spring adjustment function is completed, after a preset set time of stillness, the deviation between the target height and the actual height of the air spring exceeds the preset height deviation range, then this set of historical multi-dimensional working condition data will be marked as fault working condition data.

[0032] In this embodiment, during the process of determining fault conditions based on the preprocessed historical multi-dimensional operating condition data, a judgment logic for the static time dimension is added for the operating condition scenario where the air pump temperature remains stable and the air spring adjustment function has been completed. That is, after the air spring adjustment function is completed and a preset static time has elapsed, the deviation between the target height and the actual height of the air spring is used as the judgment basis. If the deviation exceeds the preset height deviation range, the historical multi-dimensional operating condition data of the corresponding group is marked as fault condition data, thereby enriching the fault data category dimension of the training dataset. This judgment logic can accurately identify the operating condition data corresponding to height deviation faults caused by distribution valve sealing failures after the air spring adjustment is completed and the system has been left to stand. This provides the training of the fault diagnosis neural network model with sample data containing fault characteristics after standing, enabling the model to learn the parameter correlation rules under this type of fault condition. As a result, the fault diagnosis neural network model that is subsequently put into use can quickly identify the fault state when encountering similar height deviation conditions after standing. This effectively fills the diagnostic gap in traditional diagnostic solutions that do not pay attention to fault characteristics during the standing stage, improves the data sample system for enhancing the comprehensiveness and accuracy of overall fault diagnosis, and further consolidates the technical foundation for the stable operation of the air suspension system.

[0033] Furthermore, in one embodiment, step S002 includes the following steps: S002-3: If the air pump temperature is stable, the air spring adjustment time exceeds the preset time threshold, and the deviation between the actual height of the air spring and the target height exceeds the preset height deviation range, then the set of historical multi-dimensional operating condition data will be marked as fault operating condition data.

[0034] In this embodiment, during the process of determining fault conditions on the preprocessed historical multi-dimensional operating condition data, a dual determination dimension of adjustment time and height deviation is added for the operating condition scenario where the air pump temperature remains stable. That is, when the air spring adjustment time exceeds the preset time threshold and the deviation between the actual height of the air spring and the target height exceeds the preset height deviation range, the historical multi-dimensional operating condition data of the corresponding group is marked as fault condition data, thereby further enriching the fault data feature dimensions of the training dataset. This judgment logic can accurately identify the operating condition data corresponding to faults caused by abnormal adjustment efficiency of the distribution valve, resulting in adjustment timeout and failure to meet the height standard. It supplements the training of the fault diagnosis neural network model with sample data that has the fault characteristics associated with adjustment time, enabling the model to learn the multi-parameter coupling correlation law under this type of fault condition. In turn, the fault diagnosis neural network model that is subsequently put into use can accurately identify the fault state when encountering similar operating conditions of adjustment timeout and excessive height deviation. It effectively fills the diagnostic blind spot in the traditional diagnostic scheme due to the lack of consideration of the adjustment time dimension, and provides key data support for improving the overall fault diagnosis sample system, enhancing the comprehensiveness and accuracy of diagnosis, and further strengthening the reliability of air suspension system operation status monitoring.

[0035] Furthermore, in one embodiment, the multi-dimensional operating condition data includes external ambient temperature data, air pump operating temperature data, height change data at different positions of the vehicle body, air tank pressure data, and air pump operating time data.

[0036] In this embodiment, the specific scope of the multi-dimensional operating condition data of the air suspension system is clearly defined. It specifically covers external ambient temperature data, air pump operating temperature data, height change data at different positions on the vehicle body, air tank pressure data, and air pump operating time data. Through the collaborative acquisition of this type of multi-dimensional data, comprehensive coverage and accurate capture of multi-level status information such as environmental conditions, core component operating status, vehicle posture parameters, air pressure system pressure level, and core component operating time during the operation of the air suspension system can be achieved. This clear definition of the data scope allows the collected real-time operating condition data to completely map the actual operating status of the air suspension system and distribution valve from multiple dimensions. This provides comprehensive and relevant input data for the preset fault diagnosis neural network model, avoiding the diagnostic limitations of traditional diagnostic schemes due to single data dimensions and incomplete information. The model can rely on the inherent correlation between multi-dimensional data to accurately identify fault states during analysis, providing crucial data input guarantees for improving the accuracy and reliability of overall fault diagnosis, and further strengthening the comprehensive monitoring capability of the air suspension system's operating status.

[0037] Furthermore, in one embodiment, step S200 includes the following steps: S201: Preprocess the multi-dimensional working condition data to map the working condition data of each dimension to the same numerical range in order to eliminate the impact of different data volumes on subsequent diagnosis. S202: Input the preprocessed multi-dimensional operating condition data into the preset fault diagnosis neural network model.

[0038] In this embodiment, during the process of inputting real-time collected multi-dimensional operating condition data into a preset fault diagnosis neural network model to complete fault diagnosis, a preprocessing operation is first performed on the multi-dimensional operating condition data to uniformly map the operating condition data of each dimension to the same numerical range. This eliminates the interference caused by the difference in magnitude between different dimensions of data on subsequent fault diagnosis. Then, the preprocessed multi-dimensional operating condition data is input into the preset fault diagnosis neural network model, and the model completes fault state identification and outputs diagnostic results based on the preprocessed data. This preprocessing process can normalize the magnitude of real-time operating condition data, ensuring that the data input into the fault diagnosis neural network model has a unified numerical benchmark. This avoids deviations in the model's identification of key operating condition features due to significant differences in the magnitude of different dimensions of data, providing the model with a consistent data input foundation. This enables the model to more accurately capture the inherent correlation and fault characteristics between multi-dimensional operating condition data, effectively compensating for the potential deficiency in diagnostic accuracy when no data preprocessing is performed. This further enhances the reliability and stability of the entire fault diagnosis method, building a solid technical barrier at the data input level for the accurate identification of air suspension distribution valve faults.

[0039] In summary, the embodiments of this application can use a self-organizing map (SOM) neural network (a typical classification neural network model) to classify the input data. The training of this model depends only on the characteristics of the input data itself. It can summarize the potential patterns between data through internal iteration and achieve classification output, thereby realizing accurate fault diagnosis of air suspension distribution valve and solving the threshold adaptability defects of traditional solutions.

[0040] To train the SOM neural network, its input parameters are defined as current room temperature, air pump temperature, changes in left front vehicle height, changes in right front vehicle height, changes in left rear vehicle height, changes in right rear vehicle height, and air tank pressure. Multiple sets of sample data under different operating conditions need to be collected as the training set. Examples of these operating conditions are as follows: Ambient temperature 20℃, air pump temperature 20℃, air pump working time T1, left front height rises A1mm, right front height rises B1mm, left rear height rises C1mm, right rear height rises D1mm, air tank pressure value is E1. With a room temperature of 20℃, an air pump temperature of 25℃, an air pump operating time of T2, a left front height increase of A2mm, a right front height increase of B2mm, a left rear height increase of C2mm, a right rear height increase of D2mm, and an air tank pressure of E2, and so on, n sets of data under different operating conditions are collected to complete the construction of the training set and model training.

[0041] The overall technical solution is divided into three core parts: sensor data processing, SOM neural network training, and fault output. The specific technical contents are as follows: (a) Sensor data processing This section includes two sub-steps: data acquisition and normalization, and data condition determination.

[0042] 1. Data Acquisition and Normalization Data acquisition: The system uses a temperature sensor to collect the outside temperature and the air pump operating temperature, a height sensor to measure the height change at four positions on the vehicle body (left front, right front, left rear, and right rear, with height increase being positive and decrease being negative), a pressure sensor to measure the air tank pressure, and software to record the air pump operating time. Normalization: To eliminate biases caused by differences in data volume between different datasets, it is necessary to perform a normalization operation on the collected data. The normalization formula is as follows:

[0043] Where n is the raw value currently collected by the sensor. It is the minimum value in this set of data. This represents the maximum value in the set of data. After normalization, the data will be mapped to the interval [0,1].

[0044] 2. Data operating condition determination Calibration Condition Data Acquisition: Parameter calibration and data acquisition were performed for eight types of operating conditions, including different air pump temperatures and different vehicle height adjustment directions. The specific operating conditions are as follows: ① With the air pump temperature at 120℃, the vehicle's altitude rating is raised from "medium" to "very high"; ② When the air pump temperature reaches 60℃, the vehicle's altitude rating is raised from "medium" to "very high"; ③ When the air pump temperature is 0℃, the vehicle's altitude rating is raised from "medium" to "very high"; ④ When the air pump temperature reaches -20℃, the vehicle's altitude rating increases from "medium" to "very high"; ⑤ With the air pump temperature at 120℃, the vehicle's altitude rating is lowered from "medium" to "very low"; ⑥ When the air pump temperature reaches 60℃, the vehicle's altitude rating drops from "medium" to "very low"; ⑦ When the air pump temperature is 0℃, the vehicle's altitude rating drops from "medium" to "very low"; ⑧ When the air pump temperature reaches -20℃, the vehicle's altitude rating drops from "medium" to "very low"; For the above operating conditions, the controller needs to simultaneously record the air pump temperature, room temperature, altitude values ​​collected by the four altitude sensors, air tank pressure value, and distribution valve operating time.

[0045] Fault Condition Determination: Data sets that meet any of the following conditions are determined to be fault condition data; the rest are normal condition data. Normal condition data is marked as 0, and fault condition data is marked as 1. The specific determination conditions are as follows: ① After the air pump temperature stabilizes and the air spring adjustment function is completed, the absolute value of the deviation between the height X of a certain air spring and the average value of the heights X1, X2, and X3 of the other three air springs is greater than 5mm. ②After the air pump temperature stabilizes and the air spring adjustment function is completed, and the air spring is left to stand for 12 hours, the deviation between the target height and the actual height of the air spring is greater than 5mm. ③ When the air pump temperature is stable, the air spring adjustment time T exceeds the threshold t, and the absolute value of the deviation between the actual height of the air spring and the target height is greater than 5mm.

[0046] (II) Training of SOM Neural Network Self-organizing map (SOM) neural networks can achieve classification by mapping input data to different spatial regions. They rely solely on the intrinsic features of the input data, revealing internal patterns through multiple iterations, and outperform other neural network models in terms of timeliness. This network employs a two-layer structure: an input layer and a competition layer. The input layer has n neurons, while the competition layer neurons are typically distributed in an m×n matrix, for a total of m×n neurons.

[0047] The specific steps for model training are as follows: 1. Weight and learning rate initialization: Randomly assign connection weights to the input vector and each neuron within the interval [0,1]; simultaneously set the initial learning rate to a constant within the interval (0,1); 2. Input Vector Normalization: A backpropagation (BP) neural network normalization process, similar to the sensor data normalization described earlier, is used to normalize the input vector. The input is fed into the input layer of the SOM neural network; 3. Euclidean distance calculation: Calculate the input vector and neuron weights. The Euclidean distance is given by the following formula:

[0048] 4. Winning Neuron Determination: By comparing the various Euclidean distances, the neuron with the smallest distance is selected as the winning neuron, satisfying the following equation:

[0049] 5. Neighborhood Weight Adjustment: The weights of the winning neuron and its neighboring neurons are adjusted using the following formula:

[0050] in, Let t be the learning rate at time t, which will decrease linearly to 0 over time. After the weight adjustment is completed, it is necessary to determine whether the output result meets the preset requirements. If it does, the training is terminated. If it does not, the process returns to step 2 to continue iterating until the training requirements are met.

[0051] (III) Fault Output The parameters of the trained SOM neural network are frozen, and the operating condition data collected and processed during the actual vehicle operation is input into the model. The model can output fault diagnosis results in real time (0 for non-fault results and 1 for fault results) and display the results on the vehicle instrument panel in real time. Users can intuitively judge whether there is a fault in the air suspension distribution valve based on the information displayed on the instrument panel.

[0052] Secondly, this application also provides an air suspension distribution valve fault diagnosis device, which includes: a data acquisition module for acquiring multi-dimensional operating condition data during the operation of the air suspension system; and a diagnosis output module for inputting the real-time acquired multi-dimensional operating condition data into a preset fault diagnosis neural network model, and outputting fault diagnosis results through the fault diagnosis neural network model. The functions of each module in the above-mentioned air suspension distribution valve fault diagnosis device correspond to the steps in the above-mentioned air suspension distribution valve fault diagnosis method embodiment, and their functions and implementation processes will not be described in detail here.

[0053] Thirdly, embodiments of this application provide an air suspension distribution valve fault diagnosis device, which can be a personal computer (PC), laptop computer, server, or other device with data processing capabilities.

[0054] Reference Figure 2 , Figure 2 This is a schematic diagram of the hardware structure of the air suspension distribution valve fault diagnosis device involved in the embodiments of this application. In the embodiments of this application, the air suspension distribution valve fault diagnosis device may include a processor, a memory, a communication interface, and a communication bus.

[0055] The communication bus can be of any type and is used to interconnect the processor, memory, and communication interface.

[0056] The communication interface includes input / output (I / O) interfaces, physical interfaces, and logical interfaces for interconnecting components within the air suspension distribution valve fault diagnosis equipment, as well as interfaces for interconnecting the air suspension distribution valve fault diagnosis equipment with other devices (such as other computing devices or user equipment). Physical interfaces can be Ethernet interfaces, fiber optic interfaces, ATM interfaces, etc.; user equipment can be displays, keyboards, etc.

[0057] Memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0058] The processor can be a general-purpose processor, which can call the air suspension distribution valve fault diagnosis program stored in the memory and execute the air suspension distribution valve fault diagnosis method provided in the embodiments of this application. For example, the general-purpose processor can be a central processing unit (CPU). The method executed when the air suspension distribution valve fault diagnosis program is called can be referred to the various embodiments of the air suspension distribution valve fault diagnosis method of this application, and will not be repeated here.

[0059] Those skilled in the art will understand that Figure 2 The hardware structure shown does not constitute a limitation of this application and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0060] Fourthly, embodiments of this application also provide a readable storage medium.

[0061] This application has a readable storage medium storing an air suspension distribution valve fault diagnosis program, wherein when the air suspension distribution valve fault diagnosis program is executed by a processor, it implements the steps of the air suspension distribution valve fault diagnosis method as described above.

[0062] The method implemented when the air suspension distribution valve fault diagnosis procedure is executed can be referred to in various embodiments of the air suspension distribution valve fault diagnosis method of this application, and will not be repeated here.

[0063] It should be noted that the sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0064] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. The terms "first," "second," and "third," etc., are used to distinguish different objects, etc., and do not indicate a sequence, nor do they limit "first," "second," and "third" to different types.

[0065] In the description of the embodiments of this application, terms such as "exemplary," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a concrete manner.

[0066] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.

[0067] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.

[0068] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to execute the methods described in the various embodiments of this application.

[0069] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for diagnosing faults in an air suspension distribution valve, characterized in that, The air suspension distribution valve fault diagnosis method includes: Collect multi-dimensional operating condition data during the operation of the air suspension system; The real-time collected multi-dimensional operating condition data is input into a preset fault diagnosis neural network model, and the fault diagnosis result is output through the fault diagnosis neural network model.

2. The air suspension distribution valve fault diagnosis method as described in claim 1, characterized in that, Before collecting multi-dimensional operating condition data during the operation of the air suspension system, the following is also included: Acquire historical multi-dimensional operating condition data of the air suspension system, preprocess the historical multi-dimensional operating condition data, and map the historical operating condition data of each dimension to the same numerical range. The historical multi-dimensional operating condition data is used to determine fault conditions and label the operating condition types to obtain a training dataset. The fault diagnosis neural network model is trained based on the training dataset.

3. The air suspension distribution valve fault diagnosis method as described in claim 2, characterized in that, The step of determining fault conditions and labeling the condition type based on the historical multi-dimensional operating condition data includes: If the air pump temperature is stable and the air spring adjustment function is completed, and the absolute value of the deviation between the height of a single air spring and the average height of the remaining air springs exceeds the preset deviation range, then this set of historical multi-dimensional operating condition data will be marked as fault operating condition data.

4. The air suspension distribution valve fault diagnosis method as described in claim 2, characterized in that, The step of determining fault conditions and labeling the condition type based on the historical multi-dimensional operating condition data includes: If the air pump temperature is stable and the air spring adjustment function is completed, after a preset settling time, if the deviation between the target height and the actual height of the air spring exceeds the preset height deviation range, then this set of historical multi-dimensional operating condition data will be marked as fault operating condition data.

5. The air suspension distribution valve fault diagnosis method as described in claim 2, characterized in that, The step of determining fault conditions and labeling the condition type based on the historical multi-dimensional operating condition data includes: If the air pump temperature is stable, the air spring adjustment time exceeds the preset time threshold, and the deviation between the actual height of the air spring and the target height exceeds the preset height deviation range, then this set of historical multi-dimensional operating condition data will be marked as fault operating condition data.

6. The air suspension distribution valve fault diagnosis method as described in claim 1, characterized in that, The multi-dimensional operating condition data includes external ambient temperature data, air pump operating temperature data, height change data at different positions on the vehicle body, air tank pressure data, and air pump operating time data.

7. The air suspension distribution valve fault diagnosis method as described in claim 1, characterized in that, The step of inputting real-time collected multi-dimensional operating condition data into a preset fault diagnosis neural network model includes: The multi-dimensional working condition data is preprocessed to map the working condition data of each dimension to the same numerical range in order to eliminate the impact of different data volumes on subsequent diagnosis. The preprocessed multi-dimensional operating condition data is input into a preset fault diagnosis neural network model.

8. A fault diagnosis device for an air suspension distribution valve, characterized in that, The air suspension distribution valve fault diagnosis device includes: The data acquisition module is used to collect multi-dimensional operating condition data during the operation of the air suspension system; The diagnostic output module is used to input real-time collected multi-dimensional operating condition data into a preset fault diagnosis neural network model, and output fault diagnosis results through the fault diagnosis neural network model.

9. A fault diagnosis device for an air suspension distribution valve, characterized in that, The air suspension distribution valve fault diagnosis device includes a processor, a memory, and an air suspension distribution valve fault diagnosis program stored in the memory and executable by the processor, wherein when the air suspension distribution valve fault diagnosis program is executed by the processor, it implements the steps of the air suspension distribution valve fault diagnosis method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores an air suspension distribution valve fault diagnosis program, wherein when the air suspension distribution valve fault diagnosis program is executed by a processor, it implements the steps of the air suspension distribution valve fault diagnosis method as described in any one of claims 1 to 7.

Images