| As one of the most important parts of a vehicle, suspension plays an important role on attenuating the excitation caused by road unevenness, and holding adhesion tire and road. Spring stiffness and damping coefficient of passive suspension are chosen according to experience or a specific optimization method. Once chosen, they can't be changed so that passive suspension can't adapt the complex levity of vehicle parameters and working condition. Profound researches on active suspension system and its control technology have been carried out to overcome the limitation of passive suspension. With contrast to passive suspension, active suspension can provide timely active forces according to vehicle movement state and road profile to make suspension working with its optimal state, so active suspension plays an important role on improving vehicle performance. As an active research domain, research on active suspensions has become an advanced subject in the field of vehicle dynamics and control.This dissertation combined the Jiin University 985 automobile innovation project and new century excellent talents in ministry of education-funded project "inertial control active/semi-active suspension technology research", which mainly study the output feedback robust control of active suspension to solve the complexity and bad practicality of robust controller. This dissertation chooses different optimization performances to design several kinds of active suspension according to different feedback signals. After comparing and analyzing the designed active suspensions, this dissertation concludes the best control strategy. The main researches are stated as follows:Chapter 1 Exordium. Introduce the function of active suspension briefly; summarize its status, trends and performance requirements, concludes the existing problems on robust control of active suspension home and abroad. On this basis, the main study content is presented.Chapter 2 Model of active suspension and its performance evaluation. First, different road profiles and their math models are introduced. Second, establish the half-car active suspension model based on suspension system dynamics equation, and clearly illustrates the performance requirements of active suspension. Finally, on the basis of establishing the uncertain half-car suspension model to analyze robustness, method is stated for evaluating active suspension system, which lays a theoretical basis for suspension design and performance appraisal.Chapter 3 Noval optimization algorithm of active suspension. First, the basic knowledge of linear matrix inequality (LMI) is introduced. Second, introduce a new evolution technology—differential evolution by illustrating its characteristic, realization process and application. Finally, a hybrid algorithm by mixing DE and LMI as well as its steps and flow chart are presented in order to solve BMI optimization problem.Chapter 4 Single-object robust output feedback control of active suspension. Taking H∞and H2 performance of active suspension as optimization targets, we study the optimal H∞and H2 active suspension control. According to different feedback signals, we design several kinds of active suspensions by DELMI, and then compare them from the aspects of frequency domain, time domain and robustness. Simulation results show that active suspension designed by suspension deflection owns the best performance.Chapter 5 Multi-object robust output feedback control of active suspension. In order to meet the demands of various performances, we choose H∞and H2 performance of optimization outputs as optimized target, and use generalized H2 (GH2) performance to make hard constraints changing in their allowable ranges. We design H∞/GH2 and H2/GH2 active suspension using DELMI, and then compare them with the reported active suspensions designed by state and output feedback control.Chapter 6 Non-fragile H2/GH2 robust output feedback control of active suspension. Traditional robust controller is usually fragile. In order to overcome the shortcoming, we study the H2/GH2 non-fragile controller design based on output feedback using DELMI algorithm. Then, we design non-fragile active suspension (NF active suspension) using suspension deflection. Simulation results confirm that NF active suspension is non-fragile and robust within the framework of the perturbation.Chapter 7 Experiment section. First, we test the performance of passive suspension, illuminating the necessity of robust control. Second, disscuss the importance of real-time simulation in the early development of active suspension, and make dSPACE-based real-time simulation experiments to validate the correctness of the proposed control strategy.Chapter 8 Conclusion and expectation. The main study achievements, creation parts in study are advanced, and the expectation for the future work is also put forward.The innovative research works in this dissertation is as follows:1. Present the existence condition of active suspension robust output feedback controller using BMI, and propose a noval optimization algorithm based on DE and LMI. The algorithm is simple and gives an important innovation in theory. DELMI algorithm is applied to the robust active suspension controller design, which demonstrates the practicability and effectiveness.2. In order to design a simple robust active suspension controller with high performance, a single target and multi-objective robust output feedback control is studied. Simulation results show that H2MC2 suspension has the best performance which can achieve even surpass the reported suspensions.3. From the point of practicality, non-fragile H2/GH2 output feedback control is studied considering the influence caused by controller perturbation. Based on DELMI algorithm, we design NF active suspension according to suspension deflection. Analysis results validate that it can restrain the bad influence caused by controller perturbation but also have strong robustness.The research results in the dissertation give reference for the practicality of the vehicle active suspension. |
PDF Full Text Request