Study On Mathematical Model And Adaptive Neuro-fuzzy Control Of Semi-active Suspension

Ride comfort is the basic evaluation criterion for a vehicle. As a compromise between passive and active suspension, the semi-active suspension is economical and has brief configuration as the former, and has excellent performance in damping vibration as the latter, so research on its control has become an advanced subject in the field of theory and engineering of automobile.After establishing the mathematical modal of semi-active suspension, constructing an evaluation function of ride comfort, designing adaptive neuro-fuzzy control system and carrying out simulation, we finally improve the ride comfort characteristics. The main content of this paper is as follows:1. Establish the simulation models of road input and semi-active suspension. First, analyze the statistics of road random vibration signal, then establish its simulation model; Second, establish the mathematical models of suspensions with different degree-of-freedom, and compare their dynamic equations, then deduce the equations with which the requirements of simplification could be met and finally prove their correctness by simulation; Third, establish a quarter vehicle modal of semi-active suspension according to the simplification equations.2. Construct evaluation function of ride comfort. First, analyze the relationship of the main three performance indexes of ride comfort; Second, briefly introduce the traditional evaluation function which uses the acceleration as its only index; Third, construct a new evaluation function which integrate the main three indexes.3. Design adaptive neuro-fuzzy control system. First, create a computer program for generating training data of neural network without practical manipulation experience of suspension; Second, design control systems according to different evaluation functions, then validate their efficiency by simulations; Third, compare the results of the former two simulations, and summarize their advantages and disadvantages.In conclusion, as an intelligent control method, adaptive neuro-fuzzy control is applied in semi-active suspension system and conduces to effectiveness of improving the performance of ride comfort. The control system which is proposed in this paper has no need of practical manipulation experience of suspension, adaptive to different evaluation functions and efficient for vibration damping, so it can provide a convenient and effective way for the research on suspension's control.