Research On Gain Scheduling Robust Control For Gasoline Engine Speed Regulation

With modern vehicles performance improved continuously, a higher demand is proposed to address engine speed control for the full operating envelope, such as the speed control robust stability and adaptability, the adjustment process in the rapid and smooth, so that engine speed control system is more challenging and needs more advanced design theory and methods.The existing robust engine speed regulation algorithms are very conservative. It is not guaranteed that the engine has good control quality for the full operating envelope. The gain scheduling idea is introduced in this dissertation. Gain scheduling robust control is studied for gasoline engine speed regulation. The main research contents and contributes are presented as follows.Base on the modeling idea from the Mean Value Engine Model (MVEM), the two-stroke gasoline engine plant model is set up according to the engine structure parameters and test data, in which torque load and throttle actuator characteristics are considered. The proposed engine model is validated to be able to reflect the dynamic characteristics of engine by Matlab\Simulink simulation analysis, which lay a foundation for gasoline engine speed control design analysis and simulation.A new H_∞fuzzy gain scheduled approach with pole constraints is proposed based on Takagi-Sugeno (T-S) theory and regional pole placement theory. The closed-loop global stability and performance is proved. The gasoline engine speed controller is designed which shows a continuous gain scheduling and meets the H_∞disturbance rejection and pole constraints for the full operating envelope. The gains switch deficiency of the conventional gain scheduled controller is overcome. And the TS model with uncertainties and disturbances is considered, in which the modeling errors are absorbed in the uncertainty part. A new robust H_∞fuzzy gain scheduled approach is proposed. The closed-loop global stability and performance is proved. Based on the proposed method, engine speed controller is designed. The performances of the designed robust speed controller are validated via simulation. The fuzzy gain-scheduling approach discussed above is linked by a divide and conquer type of design procedure whereby the nonlinear control design task is decomposed into a number of linear sub-problems. In addition with no requirement for precise mathematical model, the fuzzy gain-scheduling control provides an effective fuzzy intelligent control approach for complex nonlinear systems.Basd on the linear parameter varying (LPV) and polytope concept, a polytopic gain scheduled H_∞controller is designed with pole constraints. The global performance is proved. Based on the quadratic D-stability and quadratic H_∞performance concept, the controller design is simplified to the vertex controller design of the plant polytope. As a result, the calculating load is decreased greatly. Through the introduction of throttle and speed real-time measurement, the controller can be automatically scheduled as gasoline engine dynamic characteristics change, so that good stability and dynamic performance can be guaranteed along parameter varying trajectories.The advantage and validity of this controller is demonstrated via simulation. The proposed polytopic gain scheduled H_∞approach is applicable to nonlinear system, which has precision LPV model and has a certain degree of conservatism.To reduce the conservatism of the above polytopic gain scheduled H_∞controller design, a new approach is proposed and proved. In addition to the varying-parameter rate, the parameter-dependent Lyapunov function approach is considered in the new controller design. A slack variable is introduced to avoid the coupling between the Lyapunov function and system matrices. The multiconvexity concept is applied to formulate the controller synthesis as a more flexible finite-set LMI feasibility conditions at vertices, so that the controller design is cast into a convex optimization problem, and a more easily tractable condition is obtained. Based on the proposed method, engine speed gain-scheduled H_∞controller is designed. The method effectiveness is validated by simulation.To validate the proposed engine speed regulation approaches, a controller Hardware-in-the-Loop (HIL) simulation platform is set up. The simulations are implemented for the above three control approaches. It is showed that the three proposed methods are better than traditional gain scheduling PID control, and the engine speed control quality and precision are meet in the presence of variable speed set points, and the speed overshoot is limited in the specified control target under torque load disturbances. The controller HIL simulation is confirmed to reduce the experimental cost, and shorten the development cycle.Via deeper research on gain scheduling robust control for gasoline engine speed, the proposed approaches provide not only new solutions for modern gasoline engine speed regulation, and also important academic consult value for other power plant speed regulation.