Real time path planning based on the potential method for an autonomous mobile overhead traveling crane
Takeshi Akamatsu, Akihiro KANESHIGE, Kensei Terashima · 2004
This paper presents a method to establish real time path planning for three-dimensional (3-D) autonomous mobile overhead crane system for prospective factories. Proposed crane system includes a visual system and a planning system. In the visual system, CCD image sensors, laser line markers and ultra sonic sensor (USS) are used to scan the environment of transfer space. Environmental map that describes the original transfer environment is built by CCD image sensors and laser line markers by oil-line while USS is used to renovate environmental map as it recognizes partial changes of transfer environment in terms of the quick scanning by on-line. The planning system that establishes the on-line path planning is proposed by expanding the previous research of the off-line path planning. By means of the potential method as its algorithm, updated potential field that describes the latest environment of transfer space is developed from the potential field developed in previous path planning and renovated environmental map in this system. The transfer path for obstacle avoidance is then constructed from the updated potential field and used as crane control references.