Research About Active Suspension Based On The Distributed Control Theory Of The Large System

The automobile suspension system is the generic terms of the force actuating devices, which connect the wheels with the car body. And the characteristic of suspension system directly affect the car’s Riding Comfort and vehicle handling stability, so it always is one of the hotspots in automobile engineering technology research to improve the performance of the suspension system. At present, the research in active suspension control strategy, vast majority is based on a quarter or1/2suspension model, which can only improve the car’s part performance, and a few is based on the carload suspension research. Because that the vehicle suspension system is a multiple input multiple output (MIMO) nonlinear and uncertainty of complex system, adopting centralized control is hard to reach to a good control effect. By combining the great system control theory and intelligent control technologies, this paper will carry on research about active suspension control system based on the carload suspension model.The vehicle active suspension system belongs to the category of the big system, and distributed control strategy is the priority selection control method in large complicated system. This paper use the distributed control theory, to translate the complex multiple input multiple output control system into a few single input single output subsystems. Due to using the conventional control strategy, it is hard to obtain satisfactory results, here use the human reasoning and intelligent technology as its control strategy combining, and discusses a study from this control strategy.In this paper, using the vertical acceleration, pitching Angle acceleration, listing Angle acceleration, as the evaluation indexes, suspension dynamic deflection as, to design a adaptive fuzzy PID control strategy, and do online fixed the PID three parameters by using the fuzzy control rules. Also this paper puts forward a adaptive fuzzy neural network control strategy, which uses the fuzzy control rules to regulate the control parameters, in order to achieve the best performance under the condition of suspension, and uses the neural network identification ability, to cope with uncertain input data and design control network layer for improving suspension performance. In the two of conditions of random road input and heave pavement input, the simulation results of simulation on passive suspension, fuzzy PID control active suspension, fuzzy neural network active suspension experimental by using MATLAB, shows that if the performance of the suspension system is remarkably improved, fuzzy neural network suspension fuzzy PID control have better cushion capacity and better robustness than suspension.This study, not only has the certain reference value for active suspension system, and provides a good basis to domestic car’s applying active suspension system, but also lays a good foundation on making future study on vehicle dynamic performance and computer control theory.