Intelligent Electromechanical CVT Ratio Controller: Adaptive Artificial Neural Network Controller for Electromechanical Continuously variable Transmission

Good ride performance is one of the most important key attribute of a passenger vehicle. One of the methods to achieve this is by using continuously variable transmission (CVT). This is because a CVT has the capability of providing an almost infinite ratio within its limits smoothly and continuously. In this thesis, the work focuses on developing an intelligent control system that will provide the appropriate CVT ratio based on a vehicle maximum power driving strategy. This will provide maximum acceleration performance at all time during driving. On-line values of the engine speed and throttle opening are used as inputs to the developed intelligent controller, which then will provide the appropriate required CVT ratio for the vehicle to perform accordingly. An electromechanical dual acting pulley CVT designed (EMDAP-CVT)and developed by the Drivetrain Research Group (DRG) UTM, is used in this work. EMDAP-CVT with zero belt misalignment give a great performance and longer belt life. Adaptive artificial neural network (AANN) is then applied as intelligent controller and the results show that the intelligent controller using AANN performs satisfactorily. Read more