Engine Speed System Identification And The Research Of Model-based Speed Tracking Control Strategy

Engine modeling plays a very important role in the design and evaluation of controlsystem, and provides a strong theoretical basis for the study and validation of controlstrategy. The engine speed is an important performance index of the engine control system,and relates to the comfortability, drivability, emission and fuel economy of a car. In thispaper, control-oriented gasoline engine modeling and model-based engine speed trackingcontrol strategy are studied.First, focused on the modeling problem, this paper demonstrates a case study for theidentification of gasoline engines speed control system model from the perspective ofengine speed control. Considering the nonlinearity and time-varying of the engine model,a statistical two-dimension nonlinear mean-value model is introduced, which isconstructed by curve fitting technique and dynamic identification algorithm. The modelingapproach is based on a large number of experimental data which was conducted on asix-cylinder gasoline engine control test bench. Experimental validations are alsoconducted to validate the accuracy of identified model by giving several cases of modeloutput and the experimental output comparative results with different operatingcoonditions.Then, focused on the spark ignition engine speed tracking problem, a model-basedreceding horizon control scheme is proposed. In order to reduce the tracking error andimprove the transient tracking performance of the system, the error system is derived onthe basis of the identified model, and based on the error model a receding horizon optimalcontroller is designed. A fast continuation/generalized minimum residual method withrestarts(C/GMRES) is adopted to solve the solution of nonlinear receding horizon problemto obtain the online real-time solution of the system.Finally, in order to validate the controller performance and the efficiency of controlalgorithm, the simulation validating the engine speed constant value regulating and speedtracking performance is carried out in the condition of MATLAB/Simulink. At the sametime, the experimental validation is conducted on the engine test bench, and the results are also analyzed. The simulation and experimental results demonstrate that the designedoptimal controller can achieve a good control performance.