基于改进基因算法的海洋平台巡检机器人路径规划

传统遗传算法最优路径搜索效率相对较低，容易产生无实际意义个体。为此，在遗传算法选择操作中引入邻域搜索算法，提高算法的局部搜索能力，调整可变长度染色体邻接点交叉算子进化操作，避免生成间断路径。同时，在变异操作中引入多样性约束与改进的A*算法，提高遗传算法前期搜索效率。最后，在适应度函数中考虑路径长度、安全性和移动代价，生成的路径远离障碍物并在一定程度上降低转弯次数。实验证明，改进后的遗传算法在多障碍物环境下的路径规划过程中提高了搜索效率，更有利于找到实际应用中的最优解。

Path Planning of Offshore Platform Inspection Robot Based on Improved Genetic Algorithm

The optimal path search efficiency of traditional genetic algorithm is relatively low, and it is easy to generate meaningless individuals. Therefore, the idea of simulated annealing is introduced into the selection operation of the genetic algorithm to improve the local search ability of the algorithm, and the evolution operation of the variable-length chromosome adjacent point crossover operator is adjusted to avoid generating discontinuous paths. At the same time, diversity constraints and an improved A* algorithm are introduced into the mutation operation to improve the early search efficiency of the genetic algorithm. Finally, the path length, safety and movement cost are considered in the fitness function, and the generated path is far away from obstacles and reduces the number of turns to a certain extent. Experiments show that the improved genetic algorithm improves the search efficiency in the process of path planning in a multi-obstacle environment, and is more conducive to finding the optimal solution in practical applications.