A vector polar histogram method based obstacle avoidance planning for AUV

In this paper we will present a vector polar histogram method for obstacle avoidance planning of autonomous underwater vehicle (Abbr. AUV), which is based on underwater sonar measurement. By applying this new method, we can solve the local minimum and frequent vibration problem in potential field method, and overcome the shortcoming in enhanced vector field method where different effect of obstacles in various directions on AUV motion is not considered and AUV real-time speed is neglected in construction of threshold function. This new method is much more suitable for calculation of optimal moving direction when AUV makes online planning of obstacle avoidance. In the beginning of design, distance between AUV and each obstacle is measured by underwater sonar from different directions and information relating to distance distribution is transformed and expressed as the histogram in polar coordinates. Then threshold function is constructed and used for transformation from vector polar histogram to dualistic histogram. Finally the optimal moving direction is adopted through calculation of cost function in dualistic histogram. Simulation results show that the method presented in this paper is more efficient for calculation of optimal AUV moving direction and much more suitable for online planning of obstacle avoidance.