1A2-C17 車両の搬送を目的とした複数ロボット協調システムiCART : 第3報:安定な車両搬送を実現する分散協調アルゴリズム

Abstract

This study proposed the car trasnportation system termed as iCART (intelligent Cooperative Autonomous Robot Transporters). This system composes of two omni-directional robots, which have end-effecter for supporting the car, so that the system transports and handles cars at the narrow space such as parking into the mechanical parking system. This system also realizes various functions such as parking, valet parking, twoing, and the transportation of cars to a factory, ferry, parking areas, etc. The motion control algorithm, which controls two robots in coordination, was proposed. This algorithm is the leader-follower type distributed motion control algorithm, and realized the coordinated transportation of the car. However, the system occasionally undergoes oscillations; as a result, the trajectory estimation by the follower is affected, and coordinated transportation becomes difficult because of the noise and vibration signals recieved from the force sensor. Normally, these signals are very small and they can be neglected. However, the sensor noise is amplified in this algorithm because the distance between the control center and the sensor is large; the control center is defined as the representative point of the car and the force sensors are attached to the robots. It is necessary to transform the force information from the position of the force sensor to the control center. This transformation amplifies the noise recieved form the force sensor. In this paper, a leader-follower-type motion control algorithm that reduces the amplification of the noise and vibrating singnals is described. In particular, the control center of a robot is defined as its representative point. The distance between the control center and the foce sensor is small in this algorithm. Therefore, the amplification of the noise received from the force sensor is reduced.