1994 Volume 56 Issue 2 Pages 83-92
In order to utilize the agricultural vehicle as an autonomous one, it is necessary to schedule the working route in advance. The vehicle simulator was constructed for both forward and backward movements by neural networks. And genetic algorithms were adopted as an optimization method of a vehicle kinematic. By use of these simulators and genetic algorithms, the time series of the steering angles, that are the control input, were optimized and consequently the working routes of the vehicle were scheduled optimally. As this method divides the route between the initial and terminal conditions into some segments under some constraints and finds the optimal route for each segment, it is possible to create the optimal working route for the complicated conditions including both forward and backward movements.