An approach to control design for cooperative multiple mobile robots

Abstract

This paper presents a high-level design approach to cooperative control of multiple mobile robots with local sensors. The method consists of two stages. In the first stage, we design global control plans for multiple robots, which include task allocation and task change plans to complete the goal. In the second stage, we design control logic and error recovery methods for an individual robot. Sensor information is integrated by merging similar sensor states together using AND or OR logic. Unique control commands are generated by eliminating logical conflicts using transition firing sequences in the Petri net models. Errors are recovered by changing control commands dynamically using reference and control models. To demonstrate the effectiveness of the proposed method, an object-searching task is evaluated. This task is to search an object such as a box using two robots, and consists of two sub-tasks, i.e., a wall-tracking task and a robot-tracking task. Simulation results of the object-searching task, the wall-tracking task, and the error recovery are presented.