A novel particle swarm and genetic algorithm hybrid method for improved heuristic optimization of diesel engine performance

This study explores a novel application of the Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) heuristic methods in a hybrid construction on a 4 cylinder medium-duty diesel engine at part-load conditions. The application of the hybrid PSO-GA approach is compared with a basic PSO in the optimization of the control parameters of a diesel engine utilizing high EGR capability, modestly high fuel pressure capability, and a two-injection fuel strategy. The results indicate that the application of the GA to the basic PSO method improved the search breadth and convergence rate when compared to the basic PSO method alone. The novel approach of applying the GA to the fuel schedule is found to be worthy of further investigation. Applying the GA to specific parameters as way to improve optimizations on was effective in reducing the iterations and time taken to achieve satisfactory objective values. The hybrid method showed up to a 49% improvement in objective value over the basic PSO with less operational time in testing.