Research On Control Strategy Of Active Suspension Based On Co-simulations

Vehicle suspension system is an important part of the ride system, as has great impact on ride comfort and operation stability. Especially, based on certain control strategies, the active suspension system actively changes and generates desired control force and controls car body vibration, as has vehicle maintain well ride comfort and operation stability。And the superiority of control strategies has important affect on the function of active suspension, therefore, the research of active suspension has been paid increasingly attention. Based on co-simulation, this thesis makes analytical study on several control strategies.Time domain road roughness is structured by means of filtering white noise. Simplified suspension system mechanical model is structured with corresponding input and output variable amid the ADAMS, and vibration acceleration is chosen as control target using CONTROL plug-in unit educing subsystem, which can be used for Co-simulation.Based on PID domination principle and fuzzy control theory, traditional PID-controller is structured, as well as fuzzy controller is proceeding online parameter of auto tuning, also regarded as PID-controller. Fuzzy control unit of the controller is double input three output controller, whose input data is adjusting parameters of PID controller. Besides that, an compound controller is structured with the function of shifting to fuzzy control or PID control in accordance with error value.Neural network control theory is given brief introduction, and active suspension with the single neuron controller is structured by means of S-function. Furthermore, online identification neural network control system is structured by means of the online identification function of three-layer BP neural network, on the basis of it, weight of neural PID-controller is adjusted online.Controllers and export modules of suspension models mentioned in the above are respectively connected to the controller in order to proceed with Co-simulation. The output data derived from it, of vibration acceleration, suspension dynamic deflection and dynamic loads of the wheels is analyzed and compared in order get specialty, merit and weakness of different control strategies.