A backpropagation neural network is proposed as a controller for an automated guided vehicle (AGV) system. At the present stage of development, the input layer consists of two neurons and receives the state signals of the tracking errors from the camera image processor, and the sole neuron in the output layer provides the command signal of a reference yaw rate signal for the vehicle. Simulations and preliminary experimentation on a prototype vehicle showed that one hidden layer was adequate to provide good driving for such a time-varying nonlinear dynamic system. A comparison with a previous proportional controller is included.>

The authors present a guidance scheme that provides accurate tracking and faster minimization of tracking errors of the front and rear ends of automatic guided vehicles (AGVs). Sensors provided at the front and rear provide the position and orientation of the front and rear ends of the vehicle relative to the track. Control laws that make use of this information have been devised to speedily achieve accurate tracking of the front and rear ends of the vehicle with minimum overshoots. The control laws are chosen based on the presence or absence of curvature and also based on the relative location of the longitudinal axis of the vehicle relative to the track. The gains are modified online to achieve proper tracking. Simulation results are provided for illustration.>