Research On The Analysis, Design And Control Of Automobile Suspension System Based On Multibody Theory

Suspension system is a very important component of an automobile. It not only takes the function of transferring the forces(torque),but influences the ride comfort and handling stability also. Weather the design of suspension success or not is a key factor that decides the design level of the whole vehicle. The theory of multi-body dynamics and intelligent control that developed with mathematics, mechanics and computer science is finding its wider and wider application in the engineering.In this dissertation, summarizing the development, the present state and the modeling feather of multi-body dynamics lead to the suggestion of the application of multi-body dynamics to the analysis, design and control of suspension system. After introduction of the different control strategies applied to the control of semi-active suspension, the using of genetic algorithm based fuzz-neural networks control strategy is pointed out.The multi-body model of double wish-bone independent suspension is established. Using maximal coordinate methods, the kinemics of this suspension is analyzed. To design the arrangement of this kind of suspension, the design method based on the kinemics simulation is put forward. For the first time the multi-body kinemics and genetic algorithm are used to solve such a multi-objective optimization problem. The code is also used to analysis and to design a specific suspension. The experiment results prove that the design method put forward is correct.A new method to deal with the constraint violation is introduced. The simplified method of leaf-spring is studied .The dynamics character simulation of leaf-spring based on multi-body model is carried out. The results fit with the experiment results quite well.The sensitivity analysis of multi-body system described by different types of governing equations is performed and the formulas are deduced. The multi-body dynamics models of main components are established and then integrated to form the model of whole vehicle. After the road model is established, the dynamics simulation of the automobile is carried out. The thought of automobile suspension CAD based on dynamics simulation is raised. To a specific automobile, based on the optimization of character parameters, the improvement design of theleaf-spring structure is performed. The newly designed leaf-springs are mounted on the automobile to be tested. The results show that design thought is correct.The non-linear formulas of the common force elements in automobile are established. The common state space equations of automobile are deduced. Appling genetic algorithm based fuzz-neural networks control strategy to the control of semi-active suspension system. The weights of neural networks are optimized. Genetic algorithm is used to optimize the parameters and structure of fuzz-neural networks system. The control simulation of semi-active suspension system based on the multi-body is carried out. The table test of semi-active suspension is performed. The results reflect the correctness of the model and control strategy studied in the thesis.