Research on Online Calibration System of Control Parameters Based on MADS-BO

In order to shorten the calibration cycle and reduce development costs, and solve the problems of long calibration time and getting stuck in local optima in current engine control system's calibration methods, this paper designs an online calibration system based on Mesh Adaptive Direct Search Bayesian optimization algorithm. An experimental platform is designed and built specifically for the calibration of parameters of electronic throttle control system of the engine. Experimental results demonstrate the real-time performance and accuracy of the system during the online calibration process of control parameters. The research findings provide certain guidance for the online calibration of control parameters in engine systems.