Adaptive Particle Swarm Optimization based Self-Tuning Control for Combustion Engines

This paper presents a self-tuning adaptive particle swarm optimization (APSO) proportional integral derivative (PID) controller for the speed control of gasoline engine. The parameters exhibit strong uncertainties in combustion engine speed control; in particular, mass equivalent coefficient ηf and efficiency cf. Additionally, heat release Q from a unit air mass of gas is greatly influenced by these external conditions even if the air-fuel ratio is controlled to be constant and the ignition time is also well regulated. Strong uncertainty of parameters is the motivation of this research to develop an adaptive-based self-tuning control design scheme. In contrast to the model's structure, the considerable variability in parameters serves as the driving force behind this research endeavor, leading to the development of a control design scheme based on adaptive optimization of self-tuning controller gains. Based on feedback from the combustion engine, an optimal solution can be attained through the optimization mechanism. To enhance the efficiency of obtaining superior optimization solutions, we introduce the aggregation degree and evolution speed into APSO. These elements dynamically modify the inertia weight during the practical optimization process. The APSO system adapts PID gains to achieve smooth control of both speed and pressure with minimum cost of 1950 as compared to PSO (3.05 × 106) and ACO (1.2 × 107).