1995 Volume 57 Issue 6 Pages 61-67
This study intends to make an appropriate control system for a nonlinear vehicle kinematics in view of developing autonomous agricultural vehicles. In order to evaluate the results of computer simulation reported in Part 1 and to develop an autonomous vehicle, a model vehicle was designed and fabricated. To make neuro-controller effective, a learning procedure by using vehicle simulator was contrived. As the vehicle velocity was necessary for following control using the neuro-controller, the velocity was estimated using data from a geomagnetic sensor and rotary encoders. Finally, following control examinations were carried out, and it was clarified that neuro-controller was effective for vehicle control, which corroborated the results of Part 1.
