Self-adaptive optimization control method for gearbox

Abstract

The invention provides a self-adaptive optimization control method for a gearbox, and relates to the technical field of gearbox control, and the method comprises the steps: carrying out the operation feature extraction of a gearbox operation data flow through an electronic control unit, carrying out the driving prediction based on a vehicle operation state feature set, a steering wheel rotation angle and a brake pedal position, obtaining driving intention prediction information; the driving intention prediction information and the driving condition information serve as constraint parameters to be matched and analyzed with a gearbox control strategy space; and performing control feedback evaluation on the target gearbox based on the target gearbox control strategy parameter. By means of the gearbox control method and device, the technical problems that according to an existing gearbox control method, due to the fact that the accuracy of gearbox control parameter adjustment and the response timeliness are low, the phenomena of gear shifting delay and gear shifting inaccuracy frequently exist, the vehicle driving experience is poor, and the vehicle performance is limited are solved; more comfortable and safer driving experience can be provided for a driver, and meanwhile, the vehicle driving performance is optimized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of gearbox control, in particular to an adaptive optimization control method for a gearbox. BACKGROUND

[0002] Intelligent control of a gearbox is a key component of modern automobile technology. By integrating advanced electronic technology and complex control algorithms, an intelligent gearbox control system can significantly improve the performance of the gearbox, achieving higher driving performance, fuel efficiency, and ride comfort, thereby better meeting the driving needs of users.

[0003] Currently, existing gearbox control methods have low accuracy and response timeliness in adjusting gearbox control parameters, resulting in frequent shifting delays and inaccurate shifting, which causes poor vehicle driving experience and limited vehicle performance. SUMMARY

[0004] The purpose of the present application is to provide an adaptive optimization control method for a gearbox to solve the technical problem of poor vehicle driving experience and limited vehicle performance caused by frequent shifting delays and inaccurate shifting due to the low accuracy and response timeliness of existing gearbox control parameter adjustment.

[0005] In view of the above problems, the present application provides an adaptive optimization control method for a gearbox, which comprises: installing a monitoring and sensing module on a target gearbox, the monitoring and sensing module comprising a speed sensor, a torque sensor, and an oil temperature sensor; collecting and obtaining gearbox operation data streams through the monitoring and sensing module, and transmitting the gearbox operation data streams to an electronic control unit; extracting operation characteristics of the gearbox operation data streams through the electronic control unit, obtaining a set of vehicle operation state characteristics, and simultaneously obtaining a steering wheel angle and a brake pedal position through a vehicle control system; predicting driving based on the set of vehicle operation state characteristics and the steering wheel angle and brake pedal position, obtaining driving intention prediction information; identifying and obtaining driving condition information, constructing a gearbox control strategy space, matching and analyzing the driving intention prediction information and the driving condition information as constraint parameters with the gearbox control strategy space to obtain target gearbox control strategy parameters; based on the target gearbox control strategy parameters, performing control feedback evaluation on the target gearbox, obtaining control feedback effect data, and optimizing and recording the target gearbox control strategy parameters through the control feedback effect data.

[0006] One or more technical solutions provided in the present application have at least the following technical effects or advantages: The monitoring sensing module is installed on the target gearbox, the monitoring sensing module includes a rotating speed sensor, a torque sensor and an oil temperature sensor, gearbox operation data flow is collected through the monitoring sensing module, and the gearbox operation data flow is transmitted to an electronic control unit; the electronic control unit extracts operation characteristics of the gearbox operation data flow, obtains a vehicle operation state characteristic set, simultaneously obtains a steering wheel angle and a brake pedal position through a vehicle control system, performs driving prediction based on the vehicle operation state characteristic set and the steering wheel angle and the brake pedal position, obtains driving intention prediction information, identifies and obtains driving condition information, constructs a gearbox control strategy space, matches and analyzes the driving intention prediction information and the driving condition information as constraint parameters and the gearbox control strategy space, obtains target gearbox control strategy parameters, performs control feedback evaluation on the target gearbox based on the target gearbox control strategy parameters, obtains control feedback effect data, and optimizes the target gearbox control strategy parameters through the control feedback effect data. That is, the rotating speed, torque and oil temperature data of the gearbox are collected through the monitoring sensing module installed on the target gearbox to obtain the gearbox operation data flow; then, operation characteristics are extracted based on the gearbox operation data flow to obtain a vehicle operation state characteristic set, driving prediction is further performed according to the steering wheel angle and the brake pedal position of the vehicle in combination with the vehicle operation state characteristic set to obtain driving intention prediction information; on the other hand, driving condition information is obtained, and the driving intention prediction information and the driving condition information are input into the gearbox control strategy space for matching analysis to obtain gearbox control strategy parameters; finally, the target gearbox is controlled according to the gearbox control strategy parameters, the control effect is feedback evaluated, and control feedback effect data is recorded to optimize subsequent gearbox control strategy parameters; through real-time monitoring and intelligent decision-making, the accuracy and timeliness of gearbox control parameter adjustment can be improved, the gearbox shifting is more smooth and stable, and thus more comfortable and safe driving experience is provided for the driver, and vehicle driving performance can be optimized.

[0007] The above description is only a summary of the technical solutions of the present application, in order to more clearly understand the technical means of the present application, the specific embodiments of the present application can be implemented according to the content of the specification, and in order to make the above and other purposes, characteristics and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described. It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become apparent through the following description. BRIEF DESCRIPTION OF DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the application or the prior art, the accompanying drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the accompanying drawings in the following description are only exemplary, and for those skilled in the art, other drawings can be obtained without creative labor on the basis of the provided drawings.

[0009] Figure 1 Flowchart of the adaptive optimization control method for the gearbox according to the present application; Figure 2 Flowchart of obtaining driving intention prediction information in the adaptive optimization control method for the gearbox according to the present application. DETAILED DESCRIPTION

[0010] The present application provides an adaptive optimization control method for a gearbox, which solves the technical problem that the existing gearbox control method has low accuracy and timeliness in adjusting gearbox control parameters, resulting in frequent shift delay and inaccurate shift, poor driving experience and limited vehicle performance. Through real-time monitoring and intelligent decision-making, the accuracy and timeliness of gearbox control parameter adjustment can be improved, making the gearbox shift smoother and more stable, thereby providing a more comfortable and safe driving experience for the driver, and optimizing the driving performance of the vehicle.

[0011] The technical solutions in the application will be described clearly and completely below with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the application, not all embodiments of the application. It should be understood that the application is not limited by the example embodiments described herein. Based on the embodiments of the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application. In addition, it should be noted that only parts related to the application are shown in the drawings for convenience of description, not all parts.

[0012] Please refer to the accompanying Figure 1 The present application provides an adaptive optimization control method for a gearbox, which specifically comprises the following steps: Step 1: Install a monitoring and sensing module on the target gearbox, the monitoring and sensing module includes a speed sensor, a torque sensor and an oil temperature sensor, acquire gearbox operation data stream through the monitoring and sensing module, and transmit the gearbox operation data stream to an electronic control unit.

[0013] Specifically, first, a monitoring sensing module is installed on the target gearbox, which includes a speed sensor, a torque sensor, and an oil temperature sensor. The speed sensor is used to measure the speed of the gearbox input shaft and output shaft, which is crucial for determining the shift timing and monitoring the response of the gearbox. The torque sensor measures the torque transmitted to the gearbox, which helps to assess the load of the engine and the working efficiency of the gearbox. The oil temperature sensor monitors the temperature of the gearbox oil, as both excessively high or low oil temperature can affect the performance and lifespan of the gearbox. Then, at preset time nodes, the gearbox operation data stream is collected by the monitoring sensing module, which can be set according to actual conditions, such as setting the interval between adjacent time nodes to 1 second. The gearbox operation data stream includes speed data, torque data, and oil temperature data. Further, the gearbox operation data stream is transmitted to the electronic control unit, which is the brain of the vehicle and is responsible for processing these data and making corresponding control decisions.

[0014] Step two: Through the electronic control unit, the running characteristics of the gearbox operation data stream are extracted to obtain the vehicle running state feature set, and the steering wheel angle and brake pedal position are obtained through the vehicle control system. Based on the vehicle running state feature set and the steering wheel angle and brake pedal position, driving prediction is performed to obtain driving intention prediction information.

[0015] Specifically, the electronic control unit extracts the running characteristics of the gearbox operation data stream, which first preprocesses the collected data before feature extraction, including removing noise, correcting abnormal values, filtering, and normalization, etc., to ensure the accuracy and consistency of the data. Then, feature extraction algorithms are used to extract the most useful information from the original data to describe the vehicle running state, including statistical methods, time-frequency analysis, machine learning techniques, etc. For example, speed data features include average speed, speed change rate, and speed fluctuation, etc.; torque data features include average torque, torque peak value, and torque fluctuation, etc.; oil temperature data features include oil temperature trend, oil temperature stability, and oil temperature peak value, etc.; and the vehicle running state feature set is obtained.

[0016] In another aspect, the steering wheel angle and brake pedal position are obtained by the vehicle control system, where the steering wheel angle sensor can accurately measure the rotation angle of the steering wheel, which is crucial for the steering control of the vehicle. By analyzing the steering wheel angle data, the steering intention of the vehicle can be predicted, and the stability control strategy of the vehicle can be adjusted accordingly; the brake pedal position can determine the stroke of the brake pedal to determine the deceleration or parking intention of the driver. By analyzing the brake pedal position data, the deceleration or parking demand of the vehicle can be predicted, and the brake system control strategy of the vehicle can be adjusted accordingly. Then, based on the vehicle operating state feature set and the steering wheel angle and brake pedal position, driving prediction is performed, where driving prediction refers to predicting the driver's intention based on real-time vehicle operating data and driver's operation input, obtaining driving intention prediction information, which includes acceleration, deceleration, turning or parking, etc. Machine learning algorithms or other pattern recognition techniques can be used to analyze the data and predict the future driving intention of the driver, such as using support vector machines to build a prediction model for prediction.

[0017] Step three: identify the driving condition information, build the gearbox control strategy space, and match and analyze the driving intention prediction information and the driving condition information as constraint parameters with the gearbox control strategy space to obtain the target gearbox control strategy parameters.

[0018] Specifically, the current driving conditions are identified and classified, such as road type (expressway, urban road, mountain road, etc.), traffic condition (congestion, smooth, etc.), weather condition (sunny, rainy, snowy, etc.), etc. to obtain driving condition information, such as: the current driving condition can be obtained by navigation software and other means. The gearbox control strategy space is built, where the gearbox control strategy space is the adjustment threshold of multiple gearbox control parameters, including shift parameters, throttle response parameters, clutch control parameters, etc., which are obtained by attribute analysis of the target gearbox.

[0019] Then the driving intention prediction information and the driving condition information are matched and analyzed as constraint parameters with the gearbox control strategy space, where the matching and analysis includes optimization algorithms, machine learning models, etc. to find the best gearbox control parameter setting, and then the target gearbox control strategy parameters are determined based on the matching analysis results, where the target gearbox control strategy parameters are the most suitable gearbox control parameters under the current driving condition. In this way, the vehicle control system can dynamically adjust the control strategy of the gearbox according to the real-time changing driving conditions and the predicted intention of the driver, to provide a safer, more comfortable and efficient driving experience.

[0020] Step four: based on the target gearbox control strategy parameters, control feedback evaluation is performed on the target gearbox, control feedback effect data is obtained, and the target gearbox control strategy parameters are optimized based on the control feedback effect data.

[0021] Specifically, based on the electronic control unit, control instructions are sent to the target gearbox according to the target gearbox control strategy parameters, including shift timing, shift speed, throttle response, etc.; at the same time, data collection is performed on the control results to obtain a control result data set, such as continuously monitoring the running state of the gearbox through sensors, including speed, torque, oil temperature, etc., and the sensors also monitor the running state of the vehicle, such as speed, acceleration, steering wheel angle, etc. Then, using a preset evaluation algorithm or model, the control results are evaluated based on the control result data set, where the evaluation indicators include the smoothness, response speed, fuel efficiency, etc. of the control effect, and the control feedback effect data is obtained. Then, according to the control feedback effect data, the target gearbox control strategy parameters are optimized, such as: when the control feedback effect of a certain indicator does not meet the expected requirements, parameter optimization analysis is performed for that indicator to obtain optimized control strategy parameters for adaptive control of the target gearbox; at the same time, during the parameter optimization analysis process, all data related to the stability and reliability of the gearbox control should be recorded to provide important basis for subsequent fault diagnosis, performance optimization and maintenance plan development, and these data records are crucial for ensuring long-term stable operation of the gearbox and improving overall vehicle performance. Through real-time monitoring and intelligent decision-making, the accuracy and timeliness of gearbox control parameter adjustment can be improved, making the gearbox shift more smooth and stable, thereby providing a more comfortable and safe driving experience for the driver, and optimizing vehicle driving performance.

[0022] The above method can solve the existing gearbox control method. Due to the low accuracy and response timeliness of the gearbox control parameter adjustment, there are often shift delays and inaccurate shifts, which causes poor vehicle driving experience and limited vehicle performance. First, a monitoring sensing module is installed on the target gearbox, which includes a speed sensor, a torque sensor and an oil temperature sensor. The gearbox operation data stream is collected and transmitted to the electronic control unit through the monitoring sensing module. Then, the electronic control unit extracts the running characteristics of the gearbox operation data stream to obtain the vehicle running state feature set, and obtains the steering wheel angle and brake pedal position through the vehicle control system. Based on the vehicle running state feature set and the steering wheel angle and brake pedal position, the driving intention prediction information is obtained. Then, the driving condition information is identified and obtained, and the gearbox control strategy space is constructed. The driving intention prediction information and the driving condition information are matched and analyzed as constraint parameters with the gearbox control strategy space to obtain the target gearbox control strategy parameter. Finally, the target gearbox is controlled and feedback evaluated based on the target gearbox control strategy parameter to obtain the control feedback effect data, and the target gearbox control strategy parameter is optimized and recorded based on the control feedback effect data. By installing a monitoring sensing module on the target gearbox to collect the speed, torque and oil temperature data of the gearbox, the gearbox operation data stream is obtained. Then, the running characteristics of the gearbox operation data stream are extracted to obtain the vehicle running state feature set. Further, the driving intention prediction information is obtained by driving prediction based on the vehicle steering wheel angle and brake pedal position and the vehicle running state feature set. On the other hand, the driving condition information is obtained, and the driving intention prediction information and the driving condition information are input into the gearbox control strategy space for matching analysis to obtain the gearbox control strategy parameter. Finally, the target gearbox is controlled according to the gearbox control strategy parameter, and the control effect is feedback evaluated, and the control feedback effect data is recorded to optimize the subsequent gearbox control strategy parameter. Through real-time monitoring and intelligent decision-making, the accuracy and timeliness of the gearbox control parameter adjustment can be improved, the gearbox shift is more smooth and stable, thereby providing a more comfortable and safe driving experience for the driver, and the vehicle driving performance can be optimized.

[0023] Further, as shown in the accompanying drawings, Figure 2 The second step of the present application includes: According to the vehicle driving application scenario, a driving intention label library is constructed; a vehicle driving database is collected and acquired, the vehicle driving database is labeled according to the driving intention label library, and a vehicle driving intention label database is obtained; a support vector machine is used to train the vehicle driving intention label database, a driving intention classification predictor is obtained; the driving intention classification predictor is multi-dimensionally evaluated and optimized, a driving intention classification optimized predictor is generated, and the vehicle running state feature set and the steering wheel angle and brake pedal position are predicted based on the driving intention classification optimized predictor, and the driving intention prediction information is obtained.

[0024] Specifically, first, the vehicle driving application scenarios are analyzed, such as analyzing different driving scenarios and purposes, such as city driving, highway driving, mountain road driving, emergency avoidance, etc., and then different driving intentions that the driver may take in these scenarios are identified, and the driving intentions are divided into several main categories, such as acceleration, deceleration, turning, and keeping straight, etc.; then for each category, the intentions are further subdivided, such as slow acceleration, smooth acceleration, fast acceleration, etc., different levels are defined for each intention category to reflect the strength and urgency of the driver's intention, for example, deceleration can be divided into slight deceleration, medium deceleration, and emergency braking, etc.; further, a driving intention label library is constructed according to multiple driving intentions and corresponding multiple levels, the driving intention label library contains all defined driving intention labels, and each label has its corresponding category and level, and the driving scene and the driver's operation can be mapped to the corresponding label, for example, on the highway, the driver may perform smooth acceleration, which can be mapped to the smooth acceleration category in the label library.

[0025] The vehicle driving database is collected and acquired, such as retrieving and acquiring historical vehicle driving data to construct a vehicle driving database under the constraint condition of a target vehicle, wherein the vehicle driving data includes vehicle speed, engine speed, steering wheel angle, brake pedal position, throttle pedal position, etc. Then the vehicle driving database is labeled according to the driving intention label library, i.e., the vehicle driving data is labeled and identified, the mapping relationship between the vehicle driving data and the vehicle driving intention label is established, and the vehicle driving intention label database is constructed.

[0026] The support vector machine is a widely used supervised learning algorithm for classification and regression analysis, which is very effective in solving linearly separable and non-linearly separable problems, and its core idea is to find a best decision boundary to make the boundary separate different categories of data as much as possible, while keeping the distance from the data points on both sides of the boundary to the boundary as large as possible. The vehicle driving intention label database is used as training data for supervised training of the support vector machine, wherein the input data of the support vector machine is vehicle driving data, including vehicle operating state features, steering wheel angle and brake pedal position, and the output data is driving intention, to obtain a driving intention classification predictor that meets the expected convergence constraint, wherein the expected convergence constraint is an expected output accuracy indicator, which can be set according to actual needs.

[0027] Then the driving intention classification predictor is optimized by multi-dimensional evaluation for improving the performance and accuracy of the model, wherein the multi-dimensional evaluation includes accuracy, precision, recall, generalization ability, etc., and then the driving intention classification predictor is optimized according to the evaluation results to obtain a driving intention classification optimized predictor. Further, the driving intention classification optimized predictor is used to predict the driving intention of the vehicle operating state feature set and the steering wheel angle and brake pedal position, and output the driving intention prediction information.

[0028] By building a driving intention classification predictor based on a support vector machine and optimizing the driving intention classification predictor according to the multi-dimensional evaluation results, the performance and accuracy of the predictor can be improved, thereby improving the accuracy, reliability and efficiency of the driving intention prediction information.

[0029] Further, the present application further comprises the following steps: The driving intention classification predictor is evaluated by multi-dimensional performance evaluation through a data validation set to determine the multi-dimensional performance evaluation parameters of the predictor; a predictor optimization parameter space is constructed, which includes predictor performance evaluation parameters, model optimization parameters and corresponding model optimization effect data; parameter matching and screening are performed based on the predictor multi-dimensional performance evaluation parameters and the predictor optimization parameter space to obtain a predictor optimization parameter set; the model optimization parameters are determined by comparing and optimizing the model optimization effect data in the predictor optimization parameter set; the driving intention classification predictor is iteratively optimized based on the model optimization parameters to generate the driving intention classification optimized predictor.

[0030] Specifically, the method of generating the driving intention classification optimization predictor is as follows: first, the driving intention classification predictor is subjected to multi-dimensional performance evaluation through a data validation set, to obtain predictor multi-dimensional performance evaluation parameters, wherein the predictor multi-dimensional performance evaluation parameters include accuracy, precision, recall, generalization ability, etc. Then, a predictor optimization parameter space is constructed, wherein the predictor optimization parameter space includes predictor performance evaluation parameters, model optimization parameters, and corresponding model optimization effect data. The predictor performance evaluation parameters are standard performance evaluation parameters, such as accuracy indicators, precision indicators, recall indicators, generalization ability indicators, etc. The model optimization parameters include learning rate update step, iteration number, regularization parameter, etc. The model optimization effect data is used to evaluate the optimization effect of different parameter combinations.

[0031] Then, the predictor multi-dimensional performance evaluation parameters are compared according to the predictor performance evaluation parameters in the predictor optimization parameter space, and the parameters that do not meet the predictor performance evaluation parameters are set as predictor optimization parameters, to obtain a predictor optimization parameter set. For example, if the accuracy is less than the accuracy indicator, then the accuracy is the predictor optimization parameter. Further, the model optimization effect data is used to compare and optimize in the predictor optimization parameter set, and the model optimization parameter with the best optimization effect is selected. Then, the driving intention classification predictor is subjected to iterative optimization according to the model optimization parameter, such as increasing the iteration training number, adjusting the learning rate update step, etc., until the driving intention classification optimization predictor that meets the expected indicators is obtained.

[0032] Further, step three of the present application includes: An attribute classifier is constructed to divide and cluster the gearbox control strategy space, to obtain a gearbox attribute feature strategy space. The driving intention prediction information and the driving condition information are used as constraint parameters, and the constraint parameters are labeled based on the attribute classifier, to obtain driving calibration feature parameters. The driving calibration feature parameters and the gearbox attribute feature strategy space are mapped and matched, to obtain a gearbox calibration strategy library. A gearbox control cost function is constructed, and the gearbox calibration strategy library is evaluated and optimized based on the gearbox control cost function, to obtain the target gearbox control strategy parameters.

[0033] Specifically, an attribute classifier is constructed based on the principle of decision tree to better understand the diversity of gearbox control strategies and divide them into strategy groups with similar attribute characteristics; then the gearbox control strategy space is divided and clustered by the attribute classifier to further refine and optimize the strategy space to make it more consistent with actual driving needs and obtain the gearbox attribute characteristic strategy space. The driving intention prediction information and the driving condition information are further used as constraint parameters, and the attribute classifier is used to mark the driving intention prediction information and the driving condition information, where marking refers to identifying the feature parameters of the constraint parameters to obtain driving calibration feature parameters, where the driving calibration feature parameters are the feature parameters after classification by the attribute classifier.

[0034] Then the driving calibration feature parameters and the gearbox attribute characteristic strategy space are mapped and matched, i.e., the driving calibration feature parameters identify the gearbox attribute characteristic strategies in the gearbox attribute characteristic strategy space to construct a gearbox calibration strategy library; on the other hand, a gearbox control cost function is constructed, which is used to evaluate the control effect of the gearbox, where the better the control effect, the smaller the cost function evaluation result, then the gearbox calibration strategy library is evaluated according to the gearbox control cost function to determine multiple cost function evaluation results, and the gearbox calibration strategy corresponding to the minimum cost function evaluation result in the multiple cost function evaluation results is further selected as the target gearbox control strategy parameter to obtain the target gearbox control strategy parameter.

[0035] Further, the present application further includes the following steps: Obtaining gearbox attribute factor information, the gearbox attribute factor information including driving intention information and driving condition information; sequentially performing attribute content node extraction on the gearbox attribute factor information to obtain a gearbox attribute content node set; performing feature parameter assignment on each content node in the gearbox attribute content node set to obtain an attribute content node feature parameter set; and constructing the attribute classifier based on the gearbox attribute content node set and the attribute content node feature parameter set.

[0036] Specifically, the method for constructing the attribute classifier is as follows. First, gearbox attribute factor information is obtained, the gearbox attribute factor information including driving intention information and driving condition information, wherein the driving intention information includes acceleration, deceleration, turning, keeping straight, etc., and the driving condition information includes road type, traffic condition, weather condition, etc. Then, attribute content node extraction is performed on the gearbox attribute factor information in sequence to obtain a gearbox attribute content node set, wherein the gearbox attribute content node refers to specific content information of each attribute factor, for example, the driving condition includes highway, urban road, mountain road, etc. road type, and congestion, smooth, etc. traffic condition, and each attribute content is taken as a node.

[0037] Then, feature parameter assignment is performed on each content node in the gearbox attribute content node set, wherein the feature parameter assignment refers to different feature parameter coding for each content node, wherein the parameter coding form can be set by oneself, for example, letter plus number coding is adopted to obtain an attribute content node feature parameter set; then, based on the decision tree principle, the gearbox attribute content node is taken as a child node, and the corresponding attribute content node feature parameter is taken as a leaf node of the child node, and the gearbox attribute content node set and the attribute content node feature parameter set are taken as construction data to generate the attribute classifier.

[0038] Further, the application further includes the following steps: The gearbox control target is defined, effect evaluation index extraction is performed on the gearbox control target to obtain a gearbox control effect evaluation index set, weight distribution is performed on each evaluation index in the gearbox control effect evaluation index set to determine an effect evaluation index weight factor set, and data fitting is performed on the gearbox control strategy space based on the gearbox control effect evaluation index set and the effect evaluation index weight factor set to construct the gearbox control cost function.

[0039] Specifically, first, the gearbox control target is defined, i.e. the performance and functional targets that the gearbox control system needs to achieve, such as shift smoothness, fuel efficiency, driving experience, etc.; then the effect evaluation index extraction of the gearbox control target is performed, i.e. according to the defined gearbox control target, the key indicators reflecting the control effect are extracted, including shift delay, shift impact, fuel consumption rate, driving comfort, etc., to obtain a gearbox control effect evaluation index set. Further based on the coefficient of variation method, the weight distribution of each evaluation index in the gearbox control effect evaluation index set is performed, wherein the coefficient of variation method is a commonly used weighting method, which can determine the importance of each index data in the overall evaluation according to the variation degree of the data. First, for each evaluation index in the gearbox control effect evaluation index set, the mean and standard deviation of all samples are calculated; then the coefficient of variation is calculated, wherein the coefficient of variation is the ratio of the standard deviation to the mean, which is used to measure the variation degree of each evaluation index; finally, according to the coefficient of variation of each evaluation index, the weight of the overall evaluation is calculated, wherein the larger the coefficient of variation, the greater the weight of the evaluation index, and a set of effect evaluation index weight factors is determined. By using the coefficient of variation method for weight distribution, the gearbox control effect evaluation can be more reasonable and objective. This method can automatically adjust the weight of each evaluation index according to the characteristics of the data, thereby improving the accuracy and reliability of the evaluation results.

[0040] Finally, according to the gearbox control effect evaluation index set and the effect evaluation index weight factor set, the gearbox control strategy space is data-fitted to generate a gearbox control cost function, wherein the evaluation result of the gearbox control cost function is the weighted calculation result of the control effect evaluation index, and the cost function evaluation result is inversely proportional to the control effect, i.e. the better the control effect, the smaller the corresponding cost function evaluation result.

[0041] Further, step four of the present application includes: Identifying the effect to be optimized in the control feedback effect data, performing parameter optimization analysis on the effect to be optimized, obtaining a control strategy parameter optimization direction; based on the control strategy parameter optimization direction, dynamically optimizing the target gearbox control strategy parameter, determining the gearbox optimization control strategy parameter, and based on the gearbox optimization control strategy parameter, adaptively controlling the target gearbox.

[0042] Specifically, the effect to be optimized in the control feedback effect data is identified, where the effect to be optimized refers to a control index that does not meet the expected control requirement, such as control stability, control reliability, shift smoothness, etc.; then parameter optimization analysis is performed based on the effect to be optimized, that is, control strategy parameters that may affect the control effect are analyzed, such as shift timing, throttle response, clutch control, etc., to determine the optimization direction of the control strategy parameters, so as to improve the effect to be optimized, and the optimization direction includes adjusting parameter values, changing parameter ranges, optimizing parameter combinations, etc. Then the target gearbox control strategy parameters are dynamically optimized based on the control strategy parameter optimization direction, to obtain gearbox optimized control strategy parameters, and finally the target gearbox is adaptively controlled according to the gearbox optimized control strategy parameters, that is, the control strategy automatically adjusts the control parameters of the gearbox, such as shift timing and shift speed, according to the current driving condition. This method can improve the control performance of the gearbox, enhance the driving experience, and ensure the stability and reliability of the vehicle under various driving conditions.

[0043] Further, the application further includes the following steps: Based on the control strategy parameter optimization direction, the control feedback effect data is analyzed for regulation and correlation, to obtain a control strategy parameter optimization pace; the target gearbox control strategy parameters are optimized and updated according to the control strategy parameter optimization pace, to obtain a control variation update strategy parameter set; and the control variation update strategy parameter set is evaluated and optimized based on the transmission control cost function, to determine the gearbox optimized control strategy parameters.

[0044] Specifically, the method for determining the gearbox optimized control strategy parameters is as follows. First, the control feedback effect data is analyzed for regulation and correlation based on the control strategy parameter optimization direction, that is, the effect of changes in the control strategy parameters on the control effect is analyzed, as well as the effect of fluctuations in the control effect on the parameters. Then, the optimization pace of the control strategy parameters is determined according to the regulation and correlation analysis results, where the optimization pace includes the frequency and amplitude of parameter updates. Next, the target gearbox control strategy parameters are optimized and updated according to the control strategy parameter optimization pace, to obtain an updated control variation update strategy parameter set. Finally, the control variation update strategy parameter set is evaluated and optimized based on the transmission control cost function, and the control variation update strategy parameter with the smallest cost function evaluation result is selected as the gearbox optimized control strategy parameter, to obtain the gearbox optimized control strategy parameters.

[0045] In summary, the adaptive optimization control method for the gearbox provided by the application has the following technical effects: 1. Through the installation of a monitoring induction module on the target gearbox, the gearbox speed, torque and oil temperature data are collected to obtain the gearbox operation data stream; then based on the gearbox operation data stream, the running feature extraction is carried out to obtain the vehicle running state feature set, and further according to the vehicle steering wheel angle and brake pedal position, the driving prediction is carried out combined with the vehicle running state feature set to obtain the driving intention prediction information; on the other hand, the driving condition information is obtained, and the driving intention prediction information and the driving condition information are input into the gearbox control strategy space for matching analysis to obtain the gearbox control strategy parameters; finally, the target gearbox is controlled according to the gearbox control strategy parameters, and the control effect is feedback evaluated, and the control feedback effect data is recorded to optimize the subsequent gearbox control strategy parameters; through real-time monitoring and intelligent decision-making, the accuracy and timeliness of the gearbox control parameter adjustment can be improved, so that the gearbox shifting is more smooth and stable, thereby providing the driver with a more comfortable and safe driving experience, and the vehicle driving performance can be optimized.

[0046] 2. By building a driving intention classification predictor based on support vector machine, and optimizing the driving intention classification predictor according to the multi-dimensional evaluation results, the performance and accuracy of the predictor can be improved, thereby improving the accuracy, reliability and efficiency of the driving intention prediction information.

[0047] The above description of disclosed embodiments enables one of ordinary skill in the art to make or use the application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0048] Obviously, various modifications and changes can be made to the present application by those skilled in the art without departing from the spirit and scope of the present application. Thus, it is intended that the present application cover the modifications and changes as falling within the scope of the application and its equivalent technologies.

Claims

1. An adaptive optimization control method for a transmission, characterized in that, The method includes: A monitoring and sensing module is installed on the target transmission. The monitoring and sensing module includes a speed sensor, a torque sensor, and an oil temperature sensor. The transmission operation data stream is collected through the monitoring and sensing module and transmitted to the electronic control unit. The electronic control unit extracts the operating features of the transmission operating data stream to obtain a set of vehicle operating status features. At the same time, the vehicle control system obtains the steering wheel angle and brake pedal position. Based on the set of vehicle operating status features and the steering wheel angle and brake pedal position, driving prediction is performed to obtain driving intention prediction information. The driving condition information is identified and acquired, and a transmission control strategy space is constructed. The driving intention prediction information and the driving condition information are used as constraint parameters to match and analyze with the transmission control strategy space to obtain the target transmission control strategy parameters. Based on the target transmission control strategy parameters, the target transmission is evaluated for control feedback to obtain control feedback effect data, and the control strategy parameters of the target transmission are optimized and recorded using the control feedback effect data.

2. The adaptive optimization control method for a transmission as described in claim 1, characterized in that, The acquisition of driving intention prediction information includes: Build a driving intent tag library based on vehicle driving application scenarios; Collect and acquire a vehicle driving database, and tag the vehicle driving database according to the driving intention tag library to obtain a vehicle driving intention tag database; A driving intent classification predictor is obtained by training the vehicle driving intent label database using a support vector machine for intent classification. The driving intention classification predictor is evaluated and optimized in multiple dimensions to generate a driving intention classification optimized predictor. Based on the driving intention classification optimized predictor, driving prediction is performed on the vehicle operating state feature set, the steering wheel angle, and the brake pedal position to obtain the driving intention prediction information.

3. The adaptive optimization control method for a transmission as described in claim 2, characterized in that, The generated driving intention classification optimization predictor includes: The driving intention classification predictor is evaluated in multiple dimensions using a data validation set to determine the multidimensional performance evaluation parameters of the predictor. Construct a predictor optimization parameter space, which includes predictor performance evaluation parameters, model optimization parameters, and corresponding model optimization effect data; Based on the multidimensional performance evaluation parameters of the predictor and the optimization parameter space of the predictor, parameter matching and filtering are performed to obtain the set of optimization parameters for the predictor. The model optimization effect data is compared and optimized within the predictor optimization parameter set to determine the model optimization parameters. Based on the model optimization parameters, the driving intention classification predictor is iteratively optimized to generate the driving intention classification optimized predictor.

4. The adaptive optimization control method for a transmission as described in claim 1, characterized in that, The obtained target transmission control strategy parameters include: Construct an attribute classifier, and use the attribute classifier to divide and cluster the gearbox control strategy space to obtain the gearbox attribute feature strategy space; The driving intention prediction information and the driving condition information are used as constraint parameters. The constraint parameters are labeled based on the attribute classifier to obtain driving calibration feature parameters. Based on the driving calibration feature parameters and the transmission attribute feature strategy space, a mapping and matching is performed to obtain the transmission calibration strategy library; A transmission control cost function is constructed, and the transmission calibration strategy library is evaluated and optimized based on the transmission control cost function to obtain the target transmission control strategy parameters.

5. The adaptive optimization control method for a transmission as described in claim 4, characterized in that, The construction of the attribute classifier includes: Obtain transmission attribute factor information, which includes driving intention information and driving condition information; The attribute content nodes of the gearbox attribute factor information are extracted sequentially to obtain a set of gearbox attribute content nodes; Assign feature parameters to each content node in the gearbox attribute content node set to obtain the attribute content node feature parameter set; The attribute classifier is constructed based on the set of attribute content nodes of the gearbox and the set of feature parameters of the attribute content nodes.

6. The adaptive optimization control method for a transmission as described in claim 4, characterized in that, The construction of the transmission control cost function includes: Define the gearbox control objective, extract the effect evaluation index of the gearbox control objective, and obtain the gearbox control effect evaluation index set; Weights are assigned to each evaluation index in the gearbox control effect evaluation index set to determine the set of effect evaluation index weight factors. Based on the set of evaluation indicators for transmission control performance and the set of weight factors for the evaluation indicators, data fitting is performed on the transmission control strategy space to construct the transmission control cost function.

7. The adaptive optimization control method for a transmission as described in claim 4, characterized in that, The method includes: Identify the effects to be optimized in the control feedback effect data, perform parameter optimization analysis on the effects to be optimized, and obtain the optimization direction of the control strategy parameters; The control strategy parameters of the target transmission are dynamically optimized based on the optimization direction of the control strategy parameters to determine the optimized control strategy parameters of the transmission, and the target transmission is adaptively controlled based on the optimized control strategy parameters of the transmission.

8. The adaptive optimization control method for a transmission as described in claim 7, characterized in that, The determination of the transmission optimization control strategy parameters includes: Based on the optimization direction of the control strategy parameters, the control feedback effect data is adjusted and correlated to obtain the control strategy parameter optimization pace. The control strategy parameters of the target gearbox are optimized and updated according to the optimization pace of the control strategy parameters to obtain a set of control mutation update strategy parameters. The control mutation update strategy parameter set is evaluated and optimized based on the transmission control cost function to determine the optimal control strategy parameters for the transmission.

Images