Enhanced Model-Free Control For Vehicle Active Suspension Systems With Uncertainties

Vibrations have a critical influence on objects which are transported in road vehicles;among these,human bodies of the driver and passengers need to be especially concerned.Vehicles active suspension system,in comparison with its counterparts,plays a significant role in adequately guarantee the stability of the vehicle and improve the suspension performances.The most valuable advantage of active suspension systems is the flexibility afforded by the actuator parts.If this flexibility is not used,the tuning of the active suspension cannot achieve optimal results for every driving state.Thus,the full potential of system performance cannot be exploited,since the requirement to maintain the safety limits of different types of road profiles imposes a reservation on the design of the controller.From this observation,the idea of developing the model-free control concept has evolved.To overcome the drawbacks of active suspension systems,i.e.,primarily their high power demand and achieving all active suspension system requirements.With a full understanding of state of the art in vehicle control systems,this thesis identifies key issues in enhanced model-free control based intelligent proportional integral derivative(i PID)strategy for the nonlinear active suspension systems with uncertainties and outside disturbances,contributes to improve the suspension performances via handling conflict between ride comfort,road holding and suspension deflection.The priority of this thesis is to highlight the contributions in performing the robustness with simplicity of control method challenges and suspension model parameter uncertainty.This work provides firstly,a new model-free fractional-order sliding mode control(MFFOSMC)based on an extended state observer(ESO)for the two degrees of freedom(2DOF)quarter car active suspension systems.The main goal is to increase the ride comfort while the dynamic wheel load and the suspension deflection remain within safety critical bounds.The model with nonlinearity,parameter variation and/or external disturbance which includes the friction force effect are simultaneously considered to provide a realistic framework.Moreover,modeling was performed using the software LMS AMESim,while the control part was configured on Matlab/Simulink.Lyapunov stability theory is used to analyze the closed-loop system and finite-time convergence stability.Secondly,it provides a new model-free fuzzy logic controller based on particle swarm optimization(PSO-MFFLC)for the 2 DOF nonlinear quarter-car active suspension systems.The proposed method comprises the model-free based intelligent PID controller,the fuzzy logic controller,the time-delay estimation,and the particle swarm optimization.The control objective is to deal with the classical conflict between minimizing vertical chassis acceleration to increase the ride comfort and keeping the dynamic wheel load small in order to ensure the ride safety.The advantage of PSO-MFFLC is quite a simple structure and easy to be regulated.Besides,it is proposed to achieve a good tracking performance with excellent robustness against uncertainties and external disturbances,which satisfies all evaluations of suspension performance simultaneously,for different road disturbances.Moreover,unlike the MFFOSMC method,all parameters are selected by the particle swarm optimization technique in the proposed PSO-MFFLC.To validate the proposed controller,the nonlinear active suspension systems with uncertainties such as the parameters variation,external disturbance,and friction force effect are simultaneously taking into account to provide a realistic framework.Finally,it presents a novel model-free adaptive fuzzy logic controller(MFAFLC)for vibration control of the 4 DOF half-car active suspension system.In order to test the model-free control strategy for a more complicated model,which adds angular dynamic degrees of freedom into account.The physical constraint of the half-car model is considered so that the parameters variation and external disturbances which are simultaneously taken into account may provide a realistic framework.The control objective is to deal with the classical conflict between minimizing both the vertical and angular chassis accelerations,which increase the ride comfort.On the other side,the dynamic wheel and suspension deflection remain within safety limits to ensure ride safety.The proposed scheme contains three parts: the first one is the model-free based intelligent PI controller,which overcomes the existing controller's complexity,introduces requisite performances,and reduces the output high order derivative.The second one is the extended state observer(ESO),which is utilized to estimate the model unknown uncertain dynamics.In the third one,the adaptive fuzzy logic based fractional order control is used to compensate for the ESO estimation error online.The feature of MFAFLC is quite a simple structure and easy to be implemented.The entire closed-loop system stability is proved based on the Lyapunov theory and the Barbalat Lemma.