Research On Predictive Control Of Automobile Controllable Suspension Based On Road Identification

When cars are commonly used in people’s lives,the ride comfort and stability of vehicle have gradually become the focus of attention.The intensity of suspension vibration is directly related to the road surface conditions.Therefore,during the driving process of vehicle,online identification of the road surface level is performed.The identified road surface conditions are combined with the suspension control system to realize the function of automatically adjusting the parameters of the suspension system.In this paper,the road level information is identified based on the reverse analysis method of vehicle dynamics response.The explicit model predictive controller is designed to control the suspension,so that the vehicle can have good ride comfort and achieve good sustainable development.The main content is summarized as follows:First,the theoretical content of the random input road level identification method is introduced,including the specific process of road level identification.The methods of road level identification method are Hilbert-Huang transform and probabilistic neural network classification.The Hilbert-Huang transform is used to analysis the vibration signal.Theoretical research contents such as the basic principles of probabilistic neural network classification.Secondly,obtain simulation data by establishing 1/4 vehicle model,and the simulation data is analyzed by the Hilbert-Huang Transform.The corresponding feature parameters are extracted under different levels of road.The feature parameters are used as the input of the probabilistic neural network,and complete the design of random input road classifier.Then,the vehicle ride comfort experiment on the road is carried out.Acceleration sensors are installed on the bottom of vehicle body and bridge of vehicle to collect the vibration information.The experimental data is analyzed by Hilbert-Huang Transform.And the sensitive feature parameters are extracted and input into the trained probabilistic neural network classifier to obtain the level identification result of the experimental road.Finally,the explicit model predictive is based on the model predictive control theory,which replaces the offline calculation and online table look-up for rolling optimization.The explicit model predictive shortens the process of repeated optimization calculations.After completing the design of the explicit model predictive controller and the object model,and the control effect is verified by comparing with the passive suspension.The research results of the thesis show that the randomly road level is identified through the method described above,and the explicit model predictive used in vehicle suspension can effectively improve the driving performance of vehicle,and improved by about 30.45%.