An improved genetic algorithm for global path planning of underwater vehicle is presented. After initializing the feasible initial population; the fitness takes the path safety; smoothness and path length into account. The pareto principle is applied to carry out strict natural selection; and the individuals whose fitness below zero in the offspring is directly eliminated; making the size of offspring population variable. Through simulation in two-dimensional and three-dimensional environments; the results show that the optimized algorithm can improve the performance and efficiency of path planning; and is more suitable for the complex underwater environments.
The total area of the ocean accounts for about 71% of surface area in the earth. The changes in land resources and climate further promote human exploration of the ocean. As our eyes and feet in the ocean; the underwater robot is the research hotspot at present and in the future. Path planning is the key to realize the autonomous and intelligent operation of underwater vehicles which means to independently plan a collision-free optimal route to the target location in the known local or global map environment.
Genetic algorithm is a simulation of the evolution of genetic material based on "Survival of the fittest". In the path planning; each chromosome composed of genes is a feasible path. Through fitness function; excellent individuals are selected for mutation; crossover and selection to produce a new generation of individuals; evolving completely at a time. Li Shaobo and song Qisong (2020) points out that in the past five years; genetic algorithm has the largest number of published articles in EI and SCI among other algorithms; and it is widely used in practical application because of its fast convergence speed and less calculation time. To solve the AUV path planning problem in a wide range of known marine environment; a hierarchical genetic algorithm (Peng Yan; Gu Guochang; 2003) is proposed. It only searches the path in the useful area; which effectively improves the search efficiency and has certain practicability.
