Nonlinear identification of Hydraulic Actuators for Vehicle Power Transmission Systems Control

摘要

Nowadays, many of the passenger cars use automatic transmissions as power transmission control devices, and in accordance with this trend, research on the advanced power transmission control of automatic transmissions is extensively carried out. Current power transmission control algorithms have some limitations because they use only speed signals of the automatic transmissions for the feedback power transmission control. Since the ultimate performance of feedback power transmission control is determined by precise pressure control of hydraulic actuators, an enhanced power transmission control algorithm can be obtained if pressure signal can be measured. However, current automobiles have no pressure sensors mainly due to the cost problems. In this study, a method to access the pressure signals of the hydraulic actuators by recurrent neural-network-based identification of the hydraulic actuators is presented. Experimental analysis has been performed to grasp the system characteristics, and an identification procedure is carried out considering the experimental analysis results. Through the validation of the identified model of the actuator, it is shown that the estimated pressure signal by the method presented in this study can be practically used for high-quality feedback control of automotive power transmission systems.