A Novel Path Planning for AUV Based on Objects’ Motion Parameters Predication

The path planning for autonomous underwater vehicle in unknown complicated oceanic environment is still a task with challenges up to now. A novel path planning algorithm is proposed to deal with the above problem based on moving objects' motion parameters predication (OMPP). Rapid random search tree algorithm is presented to divide sonar data into segments to obtain separate objects, and virtual centroid is introduced as the replacement of the object in motion parameters' predication. To predict the motion parameters of moving object including velocity and heading angles, an adaptive neuro-fuzzy inference system is adopted. Then, the OMPP algorithm is executed to plan an optimal path, where particle swarm optimization is used to produce temporary waypoints for obstacle avoidance, and a smooth path to destination is automatically produced under the guiding of temporary waypoints and destination. Finally, simulations are conducted in MATLAB soft environment, the results show that the OMPP algorithm is feasible, and the path planned by this algorithm is optimal comparing with other two algorithms.