Optimal Idle Speed Control of a Natural Aspirated Gasoline Engine Using Bio-inspired Meta- heuristic Algorithms

In order to lowering level of emissions of internal combustion engines (ICEs), they should be optimally controlled. However, ICEs operate under numerous operating conditions, which in turn makes it difficult to design controller for such nonlinear systems. In this article, a generalized unique controller for idle speed control under whole loading conditions is designed. In the current study, instead of tedious time-consuming trial-and-error based methods, soft computing techniques are employed to tune a proportional-integral-derivative (PID) controller which controls idle speed of engine. Since model based design technique is employed, a mean value model (MVM) is taken advantage due to its evidenced merits. Moreover, a brief introduction to the selected meta-heuristics is given followed by a flowchart to show how the engine model is linked to the optimization algorithms. A set point of 750 rpm is fed to the system, and the weighted sum of the three characteristics of mean squared error, control energy, and percent overshoot of the control system is set to the problem objective function to be minimized. It is evidenced that of all the examined meta-heuristics, Bees Algorithm (BA) converges to a better solution. Finally, to consider the effectiveness of the developed optimal controllers in disturbance rejection, they are implemented to the engine MVM model. The results of the research indicate, all the four optimally designed control systems, albeit the intermediate superiority, are of conspicuous success in compensating for the input disturbances of the load torque.