A New Path Planning Algorithm for Marine Engine Room Inspection Based on RA*-ADWA*

Path planning plays a key role in robotic inspection, especially on ships. Previous research does not consider the unique terrain condition of the engine room, such as ladder steps and narrow areas, which greatly influence the walking efficiency of robots. In addition, how to realize path correction when the direction is deviated due to bumps caused by waves remains to be explored. Therefore, the RA*-ADWA algorithm combined by global and local planning is proposed in this paper, which provides a new idea for inspection in multi-terrain environment of the marine engine room. The globally optimal path(GOP) is obtained by adding terrain factors and using the rectangular safety frame filtering (RSF) method proposed in this paper. And local path planning realizes adaptive path correction by calculating the data from environmental information and introducing the distance loss function on the basis of the GOP. The method demonstrates superior performance compared to other traditional methods in terms of path length, path angle, efficiency and stability.