Control Design Of Magnetorheological Semi-active Suspension Based On Explicit Model Predictive Control

With the development of the times, people pay more and more attentions to performance of the car. As a critical part of modern vehicles, suspension system plays a very important role to improve vehicle’s safety and comfort performance. Comparing to other types of suspension systems, Semi-active suspension system which is based on magnetorheological (MR) dampers has a great value of research, and will have a wide range of applications in the future. Research on this kind of suspensions usually involves design of the suspension system and damper, system modeling, design of control strategy, system simulation, hardware-in-the-loop test and real car test, etc. Main purpose of this article is to design a control strategy for MR-damper-based semi-active suspension system to improve its performance. Main researches are including following aspects:1). Builded a hardware and software platform based on ControlBase_VT controller, and according to characteristics of this platform’s output, designed a current drive circuit for MR damper. As a rapid prototyping development platform for designing a controller, it provides a software and hardware infrastructure for downloading and testing of the control algorithm in a bench test or in a real car.2). According to the structure and working mechanism of a semi-active suspension system, this paper established a mathematical model for semi-active suspension system. And as an essential part and the actuator of a semi-active suspension system, a precise model of the MR damper is necessary. So, by analyzing characteristics of MR damper, collecting real characteristic data of MR damper from bench test, A hysteresis polynomial model has been established and tested in this paper.3). According to suspension’s performance index and model of semi-active suspension system, this paper set up an optimization problem with constraints of actual suspension system. After this optimization problem processed to a mixed logical optimization problem by MLD(Mixed Logical Dynamic) theory, based on explicit model predictive control(EMPC) theory, this article designed an EMPC controller. Then, by comparing simulation results of EMPC with results of skyhook control in Matlab, it is easy to verify that EMPC algorithm is more effective than skyhook control strategy to improve suspension’s performance.4). Using CAN bus communication and management software CANape to download EMPC strategy into ControlBase_VT, and then using CANape to combine ControlBase_VT hardware with semi-active suspension mathematical model which is modelled in Simulink to do hardware-in-the-loop test, results further proof that EMPC algorithm is very effective.In summary, this paper established corresponding mathematical models of semi-active suspension system and MR damper by analyzing their characteristics, set up an optimization object with actual constraints, designed an EMPC controller and a skyhook controller, finally, proved that EMPC algorithm is more effective than skyhook to improve safety and comfort performance of semi-active suspension by doing Matlab simulation and hardware-in-the-loop testing.