抄録
This paper describes the configuration of an intelligent vehicle, algorithms of the driving control and experimental results. The vehicle is completely autonomous and can be driven without a human driver. The intelligent vehicle consists of a vision system, a dead reckoning system, a driving controller and an automobile. The vision system consists of two TV cameras and a processing unit for video signals from the cameras. The cameras are mounted in the front part of the vehicle with a fixed interval along a vertical line. The system functions as real time obstacle detection by triangulation. It locates obstacles in front of the vehicle within the range of 5m to 20m with the visual angle of 40 degrees. The dead reckoning system measures the position and the direction of the vehicle on real time base from revolution counts of both of the rear wheels. Two algorithms are proposed as a driving control method. One algorithm employs a table-look-up method, in which the optimal control is indexed in the table of control by key words generated based on the data from the vision system. Another algorithm is the target point following method. This algorithm determines the steering control based on the positions and directions of the vehicle and the present target points. The vehicle is driven so that it may hit a sequence of target points one after another that represents a route of the vehicle. On the test track we confirmed experimentally autonomous driving of the intelligent vehicle at the speed up to 30km/h.
