Real-time Engine Control Parameters Optimization Method for Small Diesel Engine by Multi Objective Genetic Algorithm

Diesel fuel injection control systems, which are one of the expected solutions for meeting emission regulations, and achieving economical fuel consumption, consist of many parameters. To optimize this system, we have developed an adaptive multi-objective genetic algorithm (MOGA) method for real-time optimization. This method provides Pareto approximation set for the required objectives. Thus engine specifications can be precisely measured and minimize evaluation time. Pre-selection, which is an acceleration method for MOGA, is employed. By using this method we achieved the faster search of three-stage injection sets and PID controller gain.