Practical Estimator for Self-Tuning Automotive Cruise Control

This study deals with a practical method for the estimation of vehicle model parameters in the application of a self-tuning regulator to an automotive cruise control. The characteristics of any vehicle show a great variation in terms of the vehicle itself, its weight, engine, transmission and so on, as well as external disturbances caused by road incline, weather and the like. For optimal control of this control object with such continually changing characteristics, we used an adaptive control in the form of a self-tuning regulator. This self-tuning regulator consists of a dynamic characteristics estimator using a recursive least-squares nethod with vector-type forgetting factor, and a feedback gain calculator using a pole assignment. The problems involved in dynamic characteristics estimation are discussed and a procedure to solve them is proposed. Generally, in the estimation of dynamic characteristics, the static characteristic is a given, and the data from one operating point on the static characteristic are used for the estimation. However, in the case of the control object under study here, the operating point will change under external disturbances, and these changes cannot be measured. Therefore, since the estimation of the dynamic characteristics in this case does not fulfill the above conditions, the estimation procedure is extremely difficult. Thus, in the present study, the estimation of a vehicle model with a cruise control is shown as one approach to the estimation of the dynamic characteristics of the control object in which the operating point changes.