Path planning algorithms for mobile robots

Suboptimal algorithms for path planning in real time are described. The proposed system uses inexpensive general-purpose microprocessors for image processing and for guidance and control to determine a suboptimal path between two nodes and to recalculate the path in the event of path blockage by an obstacle. In addition, it has obstacle avoidance capabilities if there is enough clearance to allow vehicle passage in the presence of an obstacle. The algorithms are for global path planning along predefined world pathways and for lateral path planning within a pathway, for obstacle avoidance. The world is based on a factory floor layout model. The presence of potential obstacles precludes storage of the world map in memory. Information on pathway layout and obstacle presence is acquired by means of a pair of linear image arrays (LIAs). Two LIAs are needed for stereo vision, used in obstacle detection.<
>