A fully modular online controller for robot navigation in static and dynamic environments

摘要

In this paper, a new approach for behavior based architecture in mobile robot navigation in unknown environments is presented. In this approach, the general actions of a mobile robot are classified into three thoroughly separated tasks. Then, an individual fuzzy controller is designed and implemented for each task (behavior). Finally, a behavior selector is designed to switch between the robot tasks (behavior patterns) and manage all separated fuzzy controllers to navigate a mobile robot in a fully unknown environment. Another case considered in this paper is to test and examine the proposed algorithm in an environment including the static and dynamic (moving) obstacles. The most difficult case of robot navigation examined in this paper is when the obstacles in the robot's environment are assumed to be both static and dynamic, while the target of the robot has a dynamics position (like a moving carriage) too.