Motion planning for an under-actuated autonomous underwater vehicle based on fast marching nonlinear model-predictive quantum particle swarm optimization

Abstract

The motion planner determines the autonomy of the autonomous underwater vehicle. However, many studies often ignore the characteristics of autonomous underwater vehicles and the underwater environment, resulting in unreasonable trajectories. To solve this problem, this paper plans the motion force through prediction and optimization to integrate the characteristics of the autonomous underwater vehicle. The fast travel algorithm is modified to take advantage of favorable currents and output arrival times, taking into account the peculiarities of the underwater environment. A novel objective function is proposed, which incorporates the arrival times to imply the requirements of obstacle avoidance and distance optimization. A penalty item is included in the objective function to avoid strong side flow. At the same time, the force and its changing rate limited in appropriate ranges are also included to balance energy consumption. In order to meet the high running time requirements in practical applications, the warm-start improved quantum particle swarm optimization is introduced into force optimization. Simulation studies are performed on a self-developed autonomous underwater vehicle. The method has good application effects in multi-island areas, ocean currents, and dynamic three-dimensional environments. Comparative simulations show that the planner plans shorter paths with good robustness and adaptability.