Fuzzy control strategy design for an autopilot on automobile chassis dynamometer test stands

[[abstract]]An optimal control strategy designed for an autopilot system to aid driving pattern simulation on automobile chassis dynamometer test stands is described. The control strategy is based on modern fuzzy theory to simulate driver behavior under dynamic road-driving conditions. Using a phase plane analysis approach, the performance characteristics of a conventional nonfuzzy proportional-integral (PI) control, a linear fuzzy control, and a nonlinear fuzzy control are compared. A devised optimal model, based on the combination of both linear and nonlinear fuzzy theories, is introduced. The validation of the effect of each model on the dynamic vehicle performance is also carried out using computer simulation. An IM 240 driving pattern is chosen as an example driving condition. The results show that the new fuzzy control strategy is able to achieve the target of fast response and high accuracy during the driving test. Hence it is adequate to be implemented into the autopilot system to replace an experienced driver in the vehicle performance laboratory.[[fileno]]2020208010013[[department]]動機