NEURO FUZZY TRANSMISSION CONTROL FOR AUTOMOBILE WITH VARIABLE LOADS

A transmission controller for an automobile with variable loads is des igned and tested by using a transmission system (Isuzu NAVi-5) consist ing of actuators and sensors besides gears, a throttle, and a clutch. This system engages and disengages the clutch and gears up and down au tomatically, In other words, a manual gear-shift system is automated. As neuro and fuzzy approaches are promising methods to interface betwe en a vehicle operator and an automobile, they are used to design a tra nsmission controller. Automobile loads are estimated from the signals of status sensors by fuzzy logic, Vehicle driver's intention is inferr ed by fuzzy logic using the information from automobile status sensors on which driver's intention is reflected, These are then fed into a n eural network, An experienced driver (much data) teachers the neural n etwork an optimal gear-shift scheduling such that a vehicle operator f eels comfortable even when automobile loads change (for example when a n automobile is traveling up and down hill).