Diagnostic method for pressure sensor of double-clutch automatic transmission

By collecting and preprocessing the pressure sensor data and environmental data of the dual-clutch automatic transmission, dynamically adjusting the threshold and building an adaptive correction factor model, and combining machine learning algorithms to build a fault prediction model, the problem of high false alarm rate and missed alarm rate in the existing technology is solved, and more accurate and reliable diagnostic results and sensor reading accuracy are achieved.

CN120008804AInactive Publication Date: 2025-05-16SMART WORKSHOP ENTERPRISE MANAGEMENT (WUXI) CO LTD
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202510305088.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-03-14
Publication Date
2025-05-16
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing dual-clutch automatic transmission pressure sensor diagnosis method cannot adapt to pressure changes under different operating conditions, resulting in high false alarm rates and missed alarm rates, and the temperature compensation algorithm cannot be dynamically adjusted, resulting in large sensor reading errors.

Method used

By collecting pressure sensor data and environmental data of the dual-clutch automatic transmission, preprocessing and dynamic threshold adjustment, an adaptive correction factor model is built, the reading deviation caused by temperature changes is corrected, and a machine learning algorithm is used to build a fault prediction model to output the fault prediction probability.

Benefits of technology

It effectively reduces the false alarm rate and missed alarm rate, improves the accuracy and reliability of diagnostic results, significantly improves the accuracy of sensor readings, and ensures normal operation under different temperature conditions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120008804A_ABST
    Figure CN120008804A_ABST
Patent Text Reader

Abstract

The invention discloses a double-clutch automatic transmission pressure sensor diagnosis method, and relates to the technical field of pressure sensor fault prediction, and the method comprises the steps: collecting data of a double-clutch automatic transmission pressure sensor under different working conditions of a double-clutch automatic transmission, and synchronously recording environment data; the method comprises the following steps: preprocessing collected sensor data and environmental data, dynamically adjusting pressure thresholds under different working conditions according to historical sensor data and environmental data, constructing a self-adaptive correction factor model based on a current environmental parameter in combination with a dynamic coefficient, and obtaining a correction factor; correcting reading deviation, caused by temperature change, of a pressure sensor in the double-clutch automatic transmission to obtain a corrected deviation value; according to the method, the difference between the current environment temperature and the reference temperature is calculated, the correction factor is calculated by using the temperature coefficient, and the real-time measurement data is corrected, so that the reading accuracy of the sensor is remarkably improved. The method is very important for ensuring normal operation under different temperature conditions.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of pressure sensor fault prediction, in particular to a pressure sensor diagnosis method for a dual-clutch automatic transmission. Background Art

[0002] As an efficient and energy-saving transmission system, dual-clutch automatic transmission has been widely used in the automotive industry in recent years. Its core lies in controlling odd and even gears respectively through two independent clutches, thereby achieving fast and smooth shifting operations. In order to ensure the stability and reliability of the DCT system, pressure sensors are widely used to monitor the internal oil pressure of the transmission to ensure the normal operation of various key components.

[0003] The current DCT pressure sensor diagnostic method has many shortcomings, especially when dealing with complex working conditions and environmental changes. First, the traditional fixed threshold method cannot adapt to the pressure changes under different working conditions, resulting in high false alarm and missed alarm rates. Second, the existing temperature compensation algorithms are mostly static compensations, and fail to make dynamic adjustments based on real-time environmental data, resulting in large errors in the corrected sensor readings. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a dual-clutch automatic transmission pressure sensor diagnosis method to solve the problem that the traditional fixed threshold method cannot adapt to the pressure changes under different working conditions, resulting in high false alarm rate and missed alarm rate.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a dual clutch automatic transmission pressure sensor diagnosis method, which comprises:

[0008] Under different working conditions of the dual-clutch automatic transmission, the dual-clutch automatic transmission pressure sensor data is collected, and the environmental data is recorded synchronously, and the collected sensor data and environmental data are pre-processed;

[0009] Dynamically adjust pressure thresholds under different working conditions based on historical sensor data and environmental data;

[0010] Based on the current environmental parameters combined with the dynamic coefficient and the multi-dimensional environmental parameters, an adaptive correction factor model is constructed to obtain the correction factor, and the reading deviation of the pressure sensor in the dual clutch automatic transmission caused by the temperature change is corrected to obtain the correction deviation value;

[0011] A fault prediction model is constructed using a machine learning algorithm, the pressure threshold and correction deviation value are input into the fault prediction model, and the fault prediction probability is output;

[0012] Set an abnormal threshold and compare the fault prediction probability with the abnormal threshold. When the fault prediction probability exceeds the abnormal threshold, an alarm signal is triggered and specific maintenance suggestions and operation instructions are given.

[0013] As a preferred solution of the dual clutch automatic transmission pressure sensor diagnosis method of the present invention, wherein: under different working conditions of the dual clutch automatic transmission, the dual clutch automatic transmission pressure sensor data is collected, specifically including the following steps:

[0014] The different working conditions of the dual clutch automatic transmission refer to an open working condition, a pre-flushing oil working condition, and an intermittent working condition;

[0015] The sensor data includes a raw pressure reading, a corrected pressure reading, and a correction offset value.

[0016] As a preferred solution of the dual clutch automatic transmission pressure sensor diagnosis method of the present invention, wherein: the simultaneous recording of environmental data and preprocessing of the collected sensor data and environmental data specifically include the following steps:

[0017] The environmental data includes temperature and humidity;

[0018] The preprocessing includes preliminary cleaning, standardization, and removal of noise and outliers.

[0019] As a preferred solution of the dual clutch automatic transmission pressure sensor diagnosis method of the present invention, wherein: according to historical sensor data and environmental data, the pressure threshold under different working conditions is dynamically adjusted, specifically including the following steps:

[0020] The historical sensor data includes pressure sensor data, time series feature data, historical fault data, operating condition identification, and maintenance records;

[0021] extracting raw features associated with sensor pressure readings;

[0022] The original features include temperature, humidity, and engine speed;

[0023] Select the random forest model and set the parameters to default;

[0024] Randomly extract historical data of pressure sensors to build a decision tree. When each decision tree node is split, select a part of the original features for splitting, and use the average value of the prediction results of all decision trees for integrated training;

[0025] The original features are input into the trained random forest model to obtain the recommended pressure threshold under the current working condition, and the pressure threshold under the current working condition is updated according to the prediction results of the random forest model.

[0026] As a preferred solution of the dual clutch automatic transmission pressure sensor diagnosis method of the present invention, wherein: based on the current environmental parameters combined with the dynamic coefficient, an adaptive correction factor model is constructed to obtain the correction factor, which specifically includes the following steps:

[0027] The dynamic coefficient refers to a temperature coefficient that is dynamically updated using a recursive algorithm combined with historical temperature data and real-time feedback;

[0028] Determine the temperature coefficient according to the equipment manual and determine the nonlinear characteristics of the temperature coefficient as it changes with temperature;

[0029] Based on the influence of environmental parameters on the performance of pressure sensors, weights of environmental parameters are assigned, and environmental parameters with large weights are extracted using principal component analysis and converted into a comprehensive environmental index as an input parameter for obtaining the correction factor;

[0030] Based on nonlinear characteristics and comprehensive environmental index, a correction factor linear regression mathematical model is constructed;

[0031] The current environmental data and the dynamic coefficient are input into the correction factor linear regression mathematical model to obtain the correction factor. As a preferred solution of the dual clutch automatic transmission pressure sensor diagnosis method of the present invention, wherein: the correction of the reading deviation of the dual clutch automatic transmission pressure sensor caused by temperature change to obtain the correction deviation value specifically includes the following steps:

[0032] Calculate the correction factor and the real-time measurement data of the pressure sensor to obtain the actual pressure reading;

[0033] The correction deviation value is obtained according to the difference between the actual pressure reading and the real-time measurement data of the pressure sensor.

[0034] As a preferred solution of the dual clutch automatic transmission pressure sensor diagnosis method of the present invention, wherein: a fault prediction model is constructed by using a machine learning algorithm, a pressure threshold and a correction deviation value are input into the fault prediction model, and a fault prediction probability is output, which specifically includes the following steps:

[0035] Selecting an LSTM-CNN hybrid model as the basic model of the fault prediction model, the LSTM-CNN hybrid model includes an input layer, a CNN layer, an LSTM layer, a merging layer and an output layer;

[0036] Use categorical cross entropy to measure the difference between the predicted results and the actual results, and use the Adam optimizer for optimization;

[0037] Use the dataset of the LSTM-CNN hybrid model, trained at the specified epochs and batch size;

[0038] The data set of the LSTM-CNN hybrid model includes pressure sensor readings, temperature sensor readings, and humidity sensor readings;

[0039] The pressure threshold, correction deviation value, temperature, humidity, and engine speed are input into the trained fault prediction model, and the fault prediction probability is output.

[0040] As a preferred solution of the dual clutch automatic transmission pressure sensor diagnosis method of the present invention, wherein: an abnormal threshold is set, the prediction result is compared with the abnormal probability, and when the prediction result exceeds the abnormal threshold, an alarm signal is triggered, and specific maintenance suggestions and operation instructions are given, which specifically includes the following steps:

[0041] Collect the fault prediction probability after the correction deviation value calculation and the fault prediction model output;

[0042] According to the historical pressure fault data and real-time pressure data of dual clutch automatic transmission, the initial abnormal threshold is calculated using descriptive statistics method;

[0043] Obtain the fault prediction probability at the current time point from the fault prediction model, and calculate the pressure threshold at the current time point using the dynamic adjustment method of the random forest model based on the real-time pressure data and historical pressure fault data. When the fault prediction probability exceeds the dynamically adjusted pressure threshold, it indicates that a fault exists.

[0044] Set up multi-level alarms to trigger alarm signals when a fault occurs.

[0045] In a second aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the dual-clutch automatic transmission pressure sensor diagnostic method as described in the first aspect of the present invention is implemented.

[0046] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the dual-clutch automatic transmission pressure sensor diagnostic method as described in the first aspect of the present invention is implemented.

[0047] The beneficial effects of the present invention are as follows: by extracting the original features related to the sensor pressure reading (such as temperature, humidity, engine speed), and using the random forest model for integrated training, the pressure threshold is dynamically adjusted, so that the diagnosis result is more accurate and reliable. This method effectively reduces the false alarm rate and the missed alarm rate, and improves stability and reliability. In addition, by calculating the difference between the current ambient temperature and the reference temperature, and using the temperature coefficient to calculate the correction factor, the real-time measurement data is corrected, which significantly improves the accuracy of the sensor reading. It is essential to ensure normal operation under different temperature conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0049] Figure 1 This is a flow chart of the dual clutch automatic transmission pressure sensor diagnostic method in Example 1.

[0050] Figure 2 The dynamic threshold adjustment and temperature compensation model diagram in Example 1

[0051] Figure 3 LSTM-CNN hybrid fault prediction model diagram in Example 1

[0052] Figure 4 This is a flowchart of the alarm mechanism and maintenance suggestion in Example 1 DETAILED DESCRIPTION

[0053] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0054] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0055] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0056] Example 1, reference Figure 1 , Figure 2 , Figure 3 and Figure 4 , which is the first embodiment of the present invention, provides a dual clutch automatic transmission pressure sensor diagnosis method, comprising the following steps:

[0057] S1. Under different working conditions of the dual-clutch automatic transmission, the dual-clutch automatic transmission pressure sensor data is collected, and the environmental data is recorded synchronously, and the collected sensor data and environmental data are preprocessed.

[0058] The specific steps include:

[0059] S1.1. To ensure the effectiveness and pertinence of data collection, it is necessary to clarify the different types of working conditions and their characteristics. The specific working status of the working condition refers to; Opening condition: the state when the vehicle is started, usually accompanied by the initialization and self-test process of the hydraulic system. Pre-filling oil condition: the state when the transmission is ready to shift gears, at this time the hydraulic system will pre-fill oil to ensure smooth shifting. Intermittent condition: the state when the transmission switches between different gears, accompanied by higher dynamic changes. By clarifying different working conditions, it ensures that the collected data is pertinent, which is helpful for subsequent analysis and model training. And according to the characteristics of different working conditions, it can better adapt to various operating conditions and improve the accuracy of fault detection.

[0060] S1.2. In order to fully monitor the status of the dual-clutch automatic transmission, it is necessary to install pressure sensors, temperature sensors, and humidity sensors in the dual-clutch automatic transmission. Record the pressure value of the hydraulic system inside the transmission, the temperature inside the transmission and the external environment, and the ambient humidity. By installing pressure sensors, temperature sensors, and humidity sensors, the status of the dual-clutch automatic transmission can be more comprehensively monitored, providing rich data support. Multi-dimensional data input provides more accurate information for subsequent fault prediction and diagnosis.

[0061] S1.3. Open the working condition, pre-oiling condition, and intermittent working condition according to the actual operating status of the vehicle, collect the corresponding working condition data and the data of the pressure sensor, temperature sensor, and humidity sensor, and ensure the consistency of the timestamps to facilitate subsequent time series analysis. Store it in a temporary buffer area or database for easy management and query, and improve data processing efficiency. Identify and delete data points that are obviously beyond the normal range to improve the quality of the data and reduce the impact of noise on the analysis results. For missing data points, use linear interpolation to fill in. Convert the collected data into a standard normal distribution with a mean of 0 and a standard deviation of 1. Use a moving average filter to remove high-frequency noise. Convert the data to the frequency domain through the Fourier transform method to remove unnecessary frequency components, and further improve the smoothness and signal-to-noise ratio of the data.

[0062] S1.4. Segment the preprocessed data according to the operating condition type, add the corresponding operating condition label to each data segment and add an accurate timestamp, which is also convenient for subsequent time series analysis.

[0063] S2. Dynamically adjust the pressure threshold under different working conditions based on historical sensor data and environmental data.

[0064] The specific steps include:

[0065] S2.1. Extract features related to sensor pressure readings from historical data: temperature, humidity, and engine speed. The input of multi-dimensional features improves the prediction accuracy of the model, better reflects the pressure changes under actual working conditions, and provides more useful information for subsequent model training.

[0066] S2.2. Use a random forest model with strong generalization and anti-overfitting capabilities to initialize the parameters of the random forest model. Randomly extract some historical data as the training set, and select a part of the original features for splitting when each decision tree node splits. Use the average value of the prediction results of all decision trees for integrated training.

[0067] S2.3, input the original features into the trained random forest model, obtain the recommended pressure threshold under the current working condition, and dynamically adjust the pressure threshold under the current working condition according to the result recommended by the model, which can adapt to different working conditions and environmental changes, and also improve the sensitivity of fault detection. And it is applied to subsequent fault detection and diagnosis.

[0068] S3. Based on the current environmental data and the dynamic coefficient, an adaptive correction factor model is constructed to obtain the correction factor.

[0069] The specific steps include:

[0070] S3.1. Real-time collection of temperature, humidity, and air pressure through the sensor network to normalize environmental parameters of different dimensions for subsequent calculation and analysis. Calculate the temperature coefficient based on the expansion characteristics of the equipment and materials and the temperature characteristics of the resistor. Dynamically update the temperature coefficient through a recursive algorithm to ensure that it adapts to environmental changes. Assign weights (such as higher weights for temperature and lower weights for humidity and air pressure). Use principal component analysis (PCA) to extract temperatures with higher weights to reduce computational complexity, and convert temperatures with higher weights into a comprehensive environmental index as an input parameter to obtain the correction factor.

[0071] Based on the nonlinear characteristics of temperature coefficient and comprehensive environmental index, a linear regression model of correction factor is constructed.

[0072] Independent variables and dependent variables are used to build models. Through a simple linear relationship, that is, one independent variable affects one dependent variable, a straight line is found from this framework to fit the data points as much as possible, thereby building a linear regression model.

[0073] The correction factor is calculated in real time based on the current environmental parameters and dynamic coefficients. The random forest model is trained using historical environmental parameters to optimize the calculation accuracy of the correction factor.

[0074] Further explanation: Real-time calculation of dynamic temperature coefficient can accurately reflect the impact of temperature changes on pressure sensor readings, especially in environments with nonlinear temperature changes. By integrating multi-dimensional environmental parameters, the correction factor can more comprehensively compensate for the errors caused by environmental changes. The use of neural network nonlinear models makes it easier to capture the complex relationship between temperature and environmental parameters, improving compensation accuracy. The introduction of machine learning algorithms makes temperature compensation technology intelligent and can automatically learn and optimize compensation strategies.

[0075] S4. Correcting a reading deviation of a dual clutch automatic transmission pressure sensor caused by temperature change to obtain a correction deviation value.

[0076] The specific steps include:

[0077] S4.1. Apply the correction factor to the original pressure reading and calculate the corrected pressure reading. This eliminates the effect of temperature changes on the pressure reading, improves the consistency and reliability of the data and makes the reading closer to the true value. The corrected pressure reading provides more accurate input for the subsequent fault prediction model and improves the accuracy of fault prediction.

[0078] S4.2. Calculate the correction deviation value based on the corrected pressure reading and the original pressure reading. It can quantify the impact of temperature changes on pressure readings, so that the final pressure reading is closer to the true value. For example, in a low temperature environment, the uncorrected reading may be low, and a more accurate pressure value is obtained after correction. It enables engineers and technicians to clearly see the errors caused by temperature changes and make further analysis and adjustments accordingly.

[0079] S5. Use machine learning algorithms to build a fault prediction model, input the pressure threshold and correction deviation value into the fault prediction model, and output the fault prediction probability.

[0080] The specific steps include:

[0081] S5.1. Build an LSTM-CNN hybrid model, define the input layer, and receive multiple time series features and environmental features. Use the convolutional neural network (CNN) to extract local features, and the long short-term memory network (LSTM) to extract time series features. Combine the outputs of the CNN layer and the LSTM layer to form a comprehensive feature vector, which helps improve the expressiveness of the model. Define the output layer to output the fault prediction probability. Use the activation function to map the output to the range of [0,1] to capture complex nonlinear relationships and time series features, so as to more accurately identify potential fault modes and facilitate subsequent fault prediction and alarm.

[0082] Further explanation: the fault prediction model can provide early warning before a fault occurs, helping technicians take preventive measures in a timely manner, thereby reducing false positives (incorrectly identified as faults) and missed positives (failure to identify actual faults), improving the reliability and stability of pressure sensors, and avoiding equipment damage or fault expansion.

[0083] S5.2. Divide the dataset into training set, validation set and test set in a ratio of 7:2:1. Set the initial parameters of the model. Use categorical cross entropy as the loss function to measure the difference between the predicted results and the actual results. Use the Adam optimizer for optimization, which can find the optimal solution in a short time and improve the model performance. At the same time, train the model under the specified rounds and batch size, and use the validation set to evaluate the model performance. Common evaluation indicators such as accuracy, precision and recall can comprehensively evaluate the performance of the model.

[0084] S5.3, input the pressure threshold, correction deviation value, temperature, humidity, and engine speed into the trained fault prediction model, and the model will analyze the data in real time to output the fault prediction probability. By analyzing the data and outputting the fault prediction probability, potential faults can be warned in advance to reduce equipment damage and accidents.

[0085] S6. Set an abnormal threshold and compare the fault prediction result with the abnormal probability. When the prediction result exceeds the abnormal threshold, an alarm signal is triggered and specific maintenance suggestions and operation instructions are given.

[0086] The specific steps include:

[0087] S6.1. Based on historical fault data and real-time pressure data, descriptive statistical methods are used to calculate the abnormal threshold. In order to adapt to different working conditions and environmental changes, the abnormal threshold is dynamically adjusted using a machine learning model. Temperature, humidity, and engine speed are extracted from historical fault data, and these features are trained using a random forest model to predict the abnormal threshold under the current working condition. The abnormal threshold under the current working condition is dynamically updated according to the results of the model prediction to improve the sensitivity and accuracy of fault detection.

[0088] S6.2. Obtain the fault prediction probability at the current time point from the fault prediction model. This probability reflects the possibility of the pressure sensor failing in a certain period of time in the future. Compare the fault prediction probability with the set abnormal threshold. When the fault prediction probability exceeds the abnormal threshold, an alarm signal is triggered. Timely detection of potential faults reduces the time of unexpected downtime, improves equipment availability and production efficiency, and avoids equipment damage and accidents.

[0089] S6.3. In order to better manage fault warnings, a multi-level alarm mechanism is usually set up. Common alarm levels include: Level 1 alarm: low risk, reminding users to pay attention to potential problems. A warning light will be displayed and a slight sound prompt will be issued. Level 2 alarm: medium risk, it is recommended that users check the equipment as soon as possible, a stronger sound prompt will be issued, and detailed fault information will be displayed on the dashboard. Level 3 alarm: high risk, immediately triggering an emergency shutdown and notifying maintenance personnel, automatically stopping the vehicle operation, and sending an emergency notification to the driver or maintenance personnel through vehicle communication.

[0090] Different response measures are taken according to the severity of the fault based on the multi-level alarm mechanism to avoid overreaction or neglect of important problems. Further explanation of the fault prediction probability provides specific maintenance suggestions and operation guidance, which can help users better manage and maintain equipment and improve the user experience.

[0091] S6.4. Providing specific maintenance suggestions based on the fault type output by the fault prediction model can help users quickly locate the cause of the fault and reduce troubleshooting time. Common fault types and their corresponding maintenance suggestions include calibrating the pressure sensor, checking the hydraulic oil quality and pipeline connection, and checking the electronic control unit ECU and related circuits.

[0092] This embodiment also provides a computer device suitable for the dual-clutch automatic transmission pressure sensor diagnostic method, including: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the dual-clutch automatic transmission pressure sensor diagnostic method proposed in the above embodiment.

[0093] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0094] The present embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the dual-clutch automatic transmission pressure sensor diagnostic method proposed in the above embodiment is implemented; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic storage, flash memory, disk or optical disk.

[0095] In summary, the present invention extracts the original features related to the sensor pressure reading (such as temperature, humidity, engine speed), and uses the random forest model for integrated training to dynamically adjust the pressure threshold, so that the diagnosis result is more accurate and reliable. This method effectively reduces the false alarm rate and the missed alarm rate, and improves stability and reliability. In addition, by calculating the difference between the current ambient temperature and the reference temperature, and using the temperature coefficient to calculate the correction factor, the real-time measurement data is corrected, which significantly improves the accuracy of the sensor reading. It is essential to ensure normal operation under different temperature conditions.

[0096] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A dual clutch automatic transmission pressure sensor diagnosis method, characterized in that: include: Under different working conditions of the dual-clutch automatic transmission, the dual-clutch automatic transmission pressure sensor data is collected, and the environmental data is recorded synchronously, and the collected sensor data and environmental data are pre-processed; Dynamically adjust pressure thresholds under different working conditions based on historical sensor data and environmental data; Based on the current environmental parameters combined with the dynamic coefficient, an adaptive correction factor model is constructed to obtain the correction factor, and the reading deviation of the pressure sensor in the dual-clutch automatic transmission caused by temperature change is corrected to obtain the correction deviation value; A fault prediction model is constructed using a machine learning algorithm, the pressure threshold and correction deviation value are input into the fault prediction model, and the fault prediction probability is output; Set an abnormal threshold and compare the fault prediction probability with the abnormal threshold. When the fault prediction probability exceeds the abnormal threshold, an alarm signal is triggered and specific maintenance suggestions and operation instructions are given.

2. The dual clutch automatic transmission pressure sensor diagnosis method according to claim 1, characterized in that: Under different working conditions of the dual-clutch automatic transmission, the pressure sensor data of the dual-clutch automatic transmission is collected, which specifically includes the following steps: The different working conditions of the dual clutch automatic transmission refer to an open working condition, a pre-flushing oil working condition, and an intermittent working condition; The sensor data includes a raw pressure reading, a corrected pressure reading, and a correction offset value.

3. The dual clutch automatic transmission pressure sensor diagnosis method according to claim 2, characterized in that: The method of synchronously recording environmental data and preprocessing the collected sensor data and environmental data specifically includes the following steps: The environmental data includes temperature and humidity; The preprocessing includes preliminary cleaning, standardization, and removal of noise and outliers.

4. The dual clutch automatic transmission pressure sensor diagnosis method according to claim 3, characterized in that: According to historical sensor data and environmental data, the pressure threshold under different working conditions is dynamically adjusted, which specifically includes the following steps: The pressure sensor data includes pressure sensor data, time series characteristic data, historical fault data, working condition identification, and maintenance records; Extracting raw features associated with sensor pressure readings; The original characteristics include temperature, humidity and engine speed; Select the random forest model and set the parameters to default; Randomly extract historical data of pressure sensors to build a decision tree. When each decision tree node is split, select a part of the original features for splitting, and use the average value of the prediction results of all decision trees for integrated training; The original features are input into the trained random forest model to obtain the recommended pressure threshold under the current working condition, and the pressure threshold under the current working condition is updated according to the prediction results of the random forest model.

5. The dual clutch automatic transmission pressure sensor diagnosis method according to claim 4, characterized in that: Based on the current environmental parameters combined with the dynamic coefficient, an adaptive correction factor model is constructed to obtain the correction factor, which specifically includes the following steps: The dynamic coefficient refers to a temperature coefficient that is dynamically updated using a recursive algorithm combined with historical temperature data and real-time feedback; Determine the temperature coefficient according to the equipment manual and determine the nonlinear characteristics of the temperature coefficient as it changes with temperature; Based on the influence of environmental parameters on the performance of pressure sensors, weights of environmental parameters are assigned, and environmental parameters with large weights are extracted using principal component analysis and converted into a comprehensive environmental index as an input parameter for obtaining the correction factor; Based on nonlinear characteristics and comprehensive environmental index, a correction factor linear regression mathematical model is constructed; The current environmental data and dynamic coefficients are input into the correction factor linear regression mathematical model to obtain the correction factor.

6. The dual clutch automatic transmission pressure sensor diagnosis method according to claim 5, characterized in that: The correction of the reading deviation of the dual clutch automatic transmission pressure sensor caused by temperature change to obtain a correction deviation value specifically includes the following steps: Calculate the correction factor and the real-time measurement data of the pressure sensor to obtain the actual pressure reading; The correction deviation value is obtained according to the difference between the actual pressure reading and the real-time measurement data of the pressure sensor.

7. The dual clutch automatic transmission pressure sensor diagnosis method according to claim 6, characterized in that: The fault prediction model is constructed using machine learning algorithms. The pressure threshold and correction deviation value are input into the fault prediction model, and the fault prediction probability is output. The specific steps include the following: Selecting an LSTM-CNN hybrid model as the basic model of the fault prediction model, the LSTM-CNN hybrid model includes an input layer, a CNN layer, an LSTM layer, a merging layer and an output layer; Use categorical cross entropy to measure the difference between the predicted results and the actual results, and use the Adam optimizer for optimization; Use the dataset of the LSTM-CNN hybrid model, trained at the specified epochs and batch size; The data set of the LSTM-CNN hybrid model includes pressure sensor readings, temperature sensor readings, and humidity sensor readings; The pressure threshold, correction deviation value, temperature, humidity, and engine speed are input into the trained fault prediction model, and the fault prediction probability is output.

8. The dual clutch automatic transmission pressure sensor diagnosis method according to claim 7, characterized in that: Set the abnormal threshold and compare the prediction result with the abnormal probability. When the prediction result exceeds the abnormal threshold, an alarm signal is triggered and specific maintenance suggestions and operation instructions are given. The specific steps include the following: Collect the fault prediction probability after the correction deviation value calculation and the fault prediction model output; According to the historical pressure fault data and real-time pressure data of dual clutch automatic transmission, the initial abnormal threshold is calculated using descriptive statistics method; Obtain the fault prediction probability at the current time point from the fault prediction model, and calculate the pressure threshold at the current time point using the dynamic adjustment method of the random forest model based on the real-time pressure data and historical pressure fault data. When the fault prediction probability exceeds the dynamically adjusted pressure threshold, it indicates that a fault exists. Set up multi-level alarms to trigger alarm signals when a fault occurs.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the dual clutch automatic transmission pressure sensor diagnosis method according to any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the dual clutch automatic transmission pressure sensor diagnosis method according to any one of claims 1 to 8 are implemented.

Citation Information

Cited By

  • Pressure sensor real-time fault detection method based on dynamic environment adaptation

    CN120293405A