Research On Online Cognitive Technology Of Vehicle Types Based On Vehicle Motion Characteristics

The future of smart vehicles can be accepted for the user,whether it's driving style is similar to the driver's and whether the driver can build trust in the intelligent driving system are an important determinant.The intelligent drive controller and car carrier decoupling and emphasizing personal style of controller are the starting point of this paper.Therefore,this paper intends to explore an adaptive technology that enables the intelligent driving system to identify the type of the vehicle online,so that the dedicated controller can identify the universal vehicle type online and adjust the vehicle dynamics characteristics in the control strategy,so as to develop the vehicle dynamics characteristics simultaneously.At the same time,technical support is provided for the driver's personal style intelligent driving control system.In addition,the current smart car controller needs to calibrate the dynamic characteristics of the controlled vehicle through a large number of field tests to achieve precise control of the vehicle.The technology can also solve the above-mentioned process of human and material cost of huge demand problem to some extent.At present,there is still less research in related fields.In view of the above engineering requirements and research status,this paper attempts to study an adaptive technology that can recognize the dynamic type of the vehicle online,and construct a vehicle type online cognitive system.The system can use the motion state information of the vehicle to output the type of vehicle to which it belongs,and provide a basis for the adjustment of the control strategy.The specific research content of this paper includes the following aspects:Firstly,a vehicle classification method based on dynamic characteristics is selected in comparison.Firstly,the main factors affecting the lateral motion characteristics of the vehicle are analyzed from the perspective of dynamics,and the structural parameters that can distinguish different vehicle dynamics types are determined.Based on this,after investigating the existing vehicle classification criteria,the classification criteria of Volkswagen AG of Germany were chosen to define the vehicle dynamics type in this paper.Secondly,screening the characteristic physical quantities that are oriented to different types of dynamics.In order to ensure the effective recognition of vehicle types,this paper analyzes and determines the characteristic physical quantities that demonstrate the lateral motion characteristics of the vehicle,and establishes the evaluation method of similarity and redundancy.According to the calculation results,the characteristic physical quantities are furtherly screened and finally determined.Some of which are the longitudinal speed of the vehicle,the yaw rate and the steering wheel angle.Thirdly,establish a cognitive technology of vehicle types which is based on the hidden Markov model.Fully research the scientific methods in theory of pattern recognition,in view of statistical characteristics and continuity in time serieso of this paper,creatively apply the hidden Markov model which is widely used in the field of speech recognition to the online cognitive process of vehicle types in this paper,and identify specific methods.Finally,this paper verifies the cognition method of vehicle types from the offline and online layers.Using the general commercial vehicle dynamics software for off-line simulation,the feasibility of the method of offline training to obtain the template library is proved.The driving simulator is used to carry out the driver-in-loop test of online simulation,which proves the effectiveness of online cognition.Improve and conduct real-time online cognitive testing.The test results show that the online cognitive method of vehicle types can effectively identify the dynamic type of the vehicle and meet the requirements of actual engineering.