The USV Path Planning Based on the Combinatorial Algorithm

The robot path planning is the important element of .the robot navigation. There were several algorithms for obstacle planning, but the most efficient obstacle avoidance algorithm has not been found. So we should continue to research the problem. Artificial Potential Field (APF) and Ant Colony Optimization (ACO) are often used in local and global path planning. However there are some inherent problems, for examples, the problem of GNRON in APF, slow convergence and prematurity in ACO. In order to solve this issue, the global path planning should be compensated by the local path planning. In this paper, we combine the ACO algorithm with Artificial Potential Field (APF)algorithm in goal path planning, and then the modified ACO algorithm can drive the USV to the target. Otherwise, when the environment is changeable, the algorithm switches to the Angle Potential Field method and the robot can escape from the dilemma smoothly. Then the initial global planning algorithm continues to drive the robot to the target. Finally, the simulation results demonstrate that the modified method is with high quality in optimal path planning. Path planning is the key technology of the USV navigation, and the algorithm of path planning influence its rate and quality. Path planning is divided into global path planning and local path planning. Global path planning means that the robot searches the optimal path in the environment model which is established by the related data of obstacles in the real environment(Yang,2012.,Gade,2013).Otherwise local path planning means that the robot utilizes its sensors to explore the unknown environment and change the current path according to the feedback data (Steven, 2015).However, how to combine global path planning with local path planning is a key issue. In the complex environment, map-building of global path planning algorithm can't satisfy the requirement because of changeable environment, and it is necessary to utilize the local path planning algorithm. So searching for a combinatorial algorithm is an important research target. This paper studies the issue of combinatorial path planning algorithm. Firstly, ACO algorithm is used to drive the robot in global path planning (Chen, 2013., Chen, 2015). In view of prematurity, this paper adopts chooses modified ACO algorithm instead of the convention alone (Khosla, 2015., Hameed, 2014).Based on the above methods, most researchers always neglect the unexpected objects after map-building, and the robot maybe trapped into local minimum. Because the traditional APF is not suitable for fast and efficient path planning in complex environment, this paper applies Angle Potential Field into the local path planning. The key points of the paper are as follows: (1) In global path planning, this paper proposed a modified ACO algorithm aiming at solving the prematurity of ACO; (2) if the robot may collide with the unexpected obstacles, the local path planning algorithm drives the robot avoid the obstacles, then the robot continues to move to the target using the method of the global path planning algorithm; (3) the combinatorial algorithm in this paper is the combination of the above two algorithms, and different algorithm is applied into path planning according to different environment; (4) a range of simulations are done to verify the viability of the combinatorial algorithm.