Engine Calibration Using Global Optimization Methods with Customization

The automotive industry is subject to stringent regulations in emissions and growing customer demands for better fuel consumption and vehicle performance. Engine calibration, a process that optimizes engine performance by tuning engine controls (actuators), becomes challenging nowadays due to significant increase of complexity of modern engines. The traditional sweep-based engine calibration method is no longer sustainable. To tackle the challenge, this work considers two powerful global optimization methods: genetic algorithm (GA) and Bayesian optimization for steady-state engine calibration for single speed-load point. GA is a branch of meta-heuristic methods that has shown a great potential on solving difficult problems in automotive engineering. Bayesian optimization is an efficient global optimization method that solves problems with computationally expensive testing such as hyperparameter tuning in deep neural network (DNN), engine testing, etc. In real engine testing platform, only the limited number of testing (function evaluations) is available. We customized GA to cope with limited resource by tuning population size and designing self-adaptive mutation operators. Another challenge of engine calibration is that, in real engine testing platform, some solutions cannot even run completely due to the engine hardware limitations. These solutions, called non-operational solutions, are part of infeasible solutions and do not have any information about either objectives or constraints. A constraint repair algorithm is applied to handle non-operational solutions. The experimental study on high-fidelity engine model demonstrated that both customized GA and Bayesian optimization efficiently find solutions very close to global optimum both using less than 400 function evaluations. Bayesian optimization shows more stable performance and has better results even in the worst-case.