Research On Hybrid Model Predictive Control Of Vehicle Height And Posture Adjustment System For Bus With ECAS

Electronically controlled air suspension(ECAS) can achieve active modulation of suspension stiffness, damping characteristics and vehicle height appropriate to improving ride comfort, handling stability and fuel economy of the vehicle during driving process, and it has resulted in a growing research interest in the vehicle engineering. As one of the main features of ECAS, vehicle height adjustment is achieved by conducting gas charging and discharging of the air springs, which provides important technical support for improving the vehicle overall performance. In addition, due to the uneven vehicle payload distribution and the difference between the system parameters of the front and rear air spring suspensions, undesirably asynchronous height adjustment at four corners of the vehicle may be emerged, which leads to unstable phenomenon of the vehicle posture. Therefore, an effective control strategy for vehicle posture during the height adjustment process is also very essential.High-speed on/off solenoid valves, which are the critical component for vehicle height and posture adjustment of ECAS, are used to adjust the input and output air mass flow rates of the air springs. However, discrete switchings between several distinct operation modes are formed due to the different onoff status combinations of the solenoid valves, while the evolutions of the system states are defined by typically continuous dynamic equations, thus the vehicle height and posture adjustment process of ECAS has obvious hybrid dynamical characteristics, and can then be regarded as a class of typical hybrid systems. This phenomenon inspires to consider presenting a novel approach in this dissertation to the control system design problem of vehicle height and posture adjustment system for bus with ECAS based on hybrid model predictive control theory. The proposed approach is assumed to provide new theoretical basis and technical support for improving the performance of ECAS.Firstly, the nonlinear mechanism models of the vehicle height and posture adjustment system for bus with ECAS were built in the dissertation. By analyzing and summarizing the vehicle height adjustment procedure of singlewheel ECAS, the mathematical expression between the major states of the airspring during the gas charging and discharging process was derivated based on thermodynamic theory for variable-mass gas charging/discharging system.Combined with the vehicle system dynamics, the nonlinear mechanism model of the vehicle height adjustment system for single-wheel ECAS was built. On this basis, to further achieve the vehicle posture control in the vehicle height adjustment process, the full-car nonlinear mechanism model for bus with ECAS was also established, which can reflect the real-time changes of the vehicle body roll and pitch angles.Secondly, the hybrid dynamic behaviors analysis of the vehicle height and posture adjustment process for bus with ECAS was completed. According to the actual operating characteristics of ECAS, the system working space was divided into several discrete operating modes. Combined with the switching conditions between each mode and the on-off statuses of the solenoid valves, a discrete event set, which reflects the system switching behaviors, was built. Thus, the evolution of the continuous dynamics process, the jump of the discrete events and the interaction relationship between the two, which lays an important foundation for establishing the system hybrid model, were disclosed for the vehicle height and posture control procedure for bus with ECAS.Thirdly, the hybrid dynamical modeling of the vehicle height and posture adjustment system for bus with ECAS was achieved. On the basis of some reasonable linear approximations for components nonlinearities in the system mechanism model, the establishments of the hybrid model for the vehicle height adjustment of single-wheel ECAS and the vehicle posture control of fullcar ECAS were conducted by using mixed logical dynamical modeling method. Based on propositional logic, the discrete events such as the on-off statuses of the solenoid valves and the boundary conditions of the piecewise linear approximations were described accurately. After that, the coupling behaviors between the continuous dynamics processes and the discrete events in the vehicle height and posture adjustment process of ECAS were compiled by HYSDEL language, and then the canonical form of the system mixed logical dynamical model was got finally.Fourthly, the hybrid model predictive controller of the vehicle height and posture adjustment system for bus with ECAS was designed. In order to achieve the accurate tracking of vehicle height and the effective control of vehicle posture as well as to prevent the frequent switching phenomenon of the on-off statuses of the solenoid valves appeared in the vehicle height and posture adjustment process, the design of the vehicle height controller for single-wheel ECAS and the vehicle height and posture associated controller were conducted based on hybrid model predictive control method. The control problem of vehicle height and posture adjustment for bus with ECAS was attributed to a class of constrained finite time optimal control problem, which can then be amounted to solving a mixed-integer quadratic programming problem. Finally,the effective hybrid model predictive control law of the vehicle height and posture adjustment system for bus with ECAS was synthesized.At last, the actual vehicle tests were conducted to demonstrate the potential advantages of the hybrid model predictive control law of the vehicle height and posture adjustment for bus with ECAS. To solve the system real-time control problem, the equivalent explicit expression of the system hybrid model predictive control law was completed based on multiparametric programming technology. On this basis, the control law was implemented in a rapid control prototype, which can then be connected to an actual vehicle with ECAS,thus the actual vehicle test platform of the vehicle height and posture adjustment of ECAS was constructed. Based on the platform, the vehicle height adjustment tests were conducted, and then the vehicle height tracking performance and the posture control effect were analyzed to verify the effectiveness and reliability of the system hybrid model predictive control law.The research results showed that the proposed control methodology can not only enable us to effectively adjust the vehicle height(the vehicle height adjustment accuration is 96%), but also obviously decrease the peak values of the vehicle body roll and pitch angles during the vehicle height adjustment process by 39.3% and 49.4% respectively, which improves the vehicle posture control performance. Moreover, it is noted that since the established systemhybrid model contains the logical variables which reflect the discrete on-off statuses of solenoid valves, the synthesized hybrid model predictive control law can also achieve the on-off statuses direct control of the solenoid valves,and on this basis, the frequent switching of the on-off statuses of the solenoid valves in short time can be prevented, which can’t be achieved by earlier presented control methods. The research work in this dissertation is of certain significance in improving the research level of the control of ECAS and making the research content of ECAS more abundant.