遗传算法在AUV局部路径规划中的应用研究

利用遗传算法（GA）实现AUV（自主水下机器人）对于运动目标的自主避碰,根据前视声纳信息探测到的障碍物距离信息,AUV的运动信息以及航迹规划信息.算法采用了二进制编码规则,并把避碰、航迹跟踪等约束条件作为适应度函数的影响因子,将AUV的下一步的航向和速度作为避碰算法的输出结果.仿真表明,AUV对动目标的自主避碰效果良好、避碰和航迹回归路径平滑.

Research on the application of genetic algorithm in localpath planning for AUV

A genetic algorithm （GA） is used to avoid the moving obstacles based on the distance information which is detected by forward looking sonar; the moving information of AUV （Autonomous Underwater Vehicle） and the programming track information for AUV. The algorithm uses binary coding rule, and obstacle avoidance and track tracking as constraint conditions which will affect the function of adaptation. The next course and velocity of AUV are used as the output of GA. It is shown that AUV avoids obstacles well by simulation. The route of avoiding obstacles and returning is smooth too.