Research On ARC Based Vehicle Active Suspension With Road Profile Adaptiveness

The performance of traditional suspension is unable to be improved for the fixed structure and the balance between ride comfort and steering stability.In response to this disadvantage,by connecting the vehicle body and axle using an active actuator,the active suspension system is capable of making a large improve on both ride comfort and steering stability which makes it be a focus research directions.However,most of the actuators used in active suspension research are double-rod actuators with same area on both piston sides.In contrast,single-rod hydraulic actuators are seldom found in this research field.Moreover,under the condition of uncertain parameters appearance,the controller design has no relation with the road profile which makes the control object unchangeable.Correspponding to these problems,based on the electro-hydraulic system,an elaborate model was built and identified using the 1/4 vehicle active suspension test bed.Moreover,the terrain adaptive active suspension was structured using filter technique and Takagi-Sugeno Fuzzy modelling method.Here,the Adaptive Robust Control(ARC)method was employed to deal with the uncertain nonlinearities and modify the force tracking performance.In the end,the advantages of active suspension are testified through the test bed experiment of 1/4 vehicle active suspension.The main idea of this thesis can be outlined as following:(1)The elaborate modelling of servo-valve and single-rod actuator of active suspension.As to the more general single-rod actuator,three types of its realization forms,that is actuator with fixed cylinder barrel,the full active suspension which only constructed by a actuator and ordinary active suspension supported by parallel-linked spring,damper and actuator,are made to build the transfer functios.Then,an 8 orders system with fixed cylinder barrel is obatained.Corresponding to the other two 10 orders system,a full active suspension and an active suspension including a spring and damper,two different parts are derived.The first part is the transfer function from current to sprung mass displacement with zero disturbance,the second part shows the transfer function from road disturbance to sprung mass displacement with zero current input.This kind of model supply the prior knowledge for system identification.(2)The parameters of active suspension system are estimated and valided.Using 20groups step digital current signal separatly,the transfer function paramters of two different realizations,fixed cylinder barrel type and hydraulic part of full active suspension,are identified.The parameters in disturbance part of full active suspension are identified with sinusoidal signal.The simulation results using the identified parameters are compared with the experiment results to valid the model precision.As far as leakage error is considered,as to the disturbance part of full active suspension,random signal is selected to proceed the FRF frequency domain identification with Hanning window and Diff window.Then,the time domain ARX model,Output-Error model and Box-Jenkins model are identified to obtain the precise system model.(3)Research and build the terrain adaptive active suspension out loop control method which is capable of adjusting according to the road condition.The road disturbance observer are proposed using the pole placement method to deal with the issue that absence of the measurement.To solve the problem that the controller can't adapt to different road input,the linear filter method integrated to Backstepping method is proposed.In response to the constant low-pass bandwidth,another adaptive nonlinear filter is proposed.In order to improve the robustness,based on the two degree and 7 degree model,the Takagi-Sugeno modelling technique is combined with the H_?controller.The controller modification problem is transferred to a Linear Matrix Inequalities resolution issue which including controller stability analysis,the suspension rattle space and the actuator fore limit.Combine with the Parallel Distributed Compensation method,the gain of the controller which can adapt to the different road input is able to be obtained.In this way,the active force is reduced.(4)For the purpose of tracking performance promotion of single rod actuator,the Adaptive Robust Control(ARC)method is employed and researched.With the combination of adaptive control and robust control,the adaptive term with regard to the control speed and robust term with regard to the control accuracy are used to modify the uncertain element in active suspension system.To reduce the sensitiveness to the unmodeled part and uncertain signals,the online adaptive robust control is proposed.In order to take the adaptive parameter convergence issue,the identification module is separated from the control module,through adding a parameter identifier,to form a modular ARC controller.In the end,the tracking property of actual force and desired force is confirmed.(5)Based on 1/4 vehicle active suspension test bed and RCP real-time platform,the active suspension control effect and road profile adaptiveness property are testified.The data collection system is defined and the key parameters are selected.Aimed at active suspension1/4 test bed,the PID controller is designed and the displacement control is realized.The experiment results are compared to the passive suspension results and simulation results to verify the effectiveness of active suspension and accuracy of its model.Based on the dSPACE MicroAutoBox,the RCP real-time simulation system platform is built to verify the control objective shift function of the road profile adaptive controller.