Dynamic path replanning algorithm for data gathering AUV

Underwater dynamics can include static and dynamic obstacles and drift. Following the estimated shortest path for Autonomous underwater vehicle(AUVs) to reach the target location is impractical due to these underwater dynamics. Thus an efficient path replanning mechanism for AUVs that attempt to re-align itself with the estimated shortest path is required. This paper presents an efficient algorithm which estimates path dynamically towards the target. It uses a local search approach using image acquisition and segmentation that is simple to implement. A delay of 2 milliseconds is considered between capturing of two successive images on every local search to check the direction of movement of obstacle. The proposed algorithm re-plan its path from source to target faster and near to shortest distance in dynamic underwater scenario. Experiments were conducted to study the percentage deviation in the re-planned path distance to the target location indicated an average of 6.5 percent and 50 percent deviation for the cases of no drift and with drift respectively from the original shortest path distance.