Weight optimal Proportional-Integral-Plus control of a gasoline engine model

Herein, addressing the optimal control problem of a simulated gasoline engine is aimed. The challenge demands tracking a desired torque profile with the minimum possible fuel consumption while promising no knock and misfire in all cylinders. The present study suggests a solution by means of Proportional-Integral-Plus (PIP) control algorithm. In this way, a MIMO model is extracted for the system using refined instrumental variable, which is known as a linear identification method. An LQ optimal design is used to obtain control gain matrix for PIP control formulation. Controller hyper parameters are then optimized through a constrained nonlinear evolutionary search. The constraint function from the plant ensures avoidance of the two mentioned threatening situations for the engine. Final results show that, in comparison to the controller tuned manually, the optimized system consumes lower amount of fuel beside a better tracking performance of the desired input and no reported constraint violation all during the simulation course.