Fuzzy model predictive control of normalized air-to-fuel ratio in internal combustion engines

In this paper, a fuzzy model predictive controller is developed to reduce the emission pollutants in spark ignition internal combustion engines. The path to this control goal is regulating the amount of normalized air-to-fuel ratio in the engine. In order to generate the simulation data, mean value engine model is simulated. To approximate the nonlinear and fast time-varying dynamics of the engine, a modified fuzzy relational model is trained offline in batch mode. For training, gradient descent back propagation algorithm along with evolutionary asexual reproduction optimization algorithm is used. Nonlinear structure of the fuzzy model of the engine imposes nonlinear optimization to produce control signals. Hence, gradient descent algorithm is used to generate online control signals. The effectiveness and robustness of the controller are evaluated through simulations.