复杂动态环境下基于侧滑力的局部路径规划

研究了复杂动态环境下具有局部感知能力的移动机器人路径规划问题.针对传统势场法避障在拥塞环境下存在局部振荡的问题,提出虚拟侧滑力的方法,障碍物对机器人产生侧滑排斥力,而非传统的反向排斥力,并由力来直接引导机器人运动.静态障碍物的侧滑力计算与障碍物距离、朝向及目标点朝向有关；动态障碍物的侧滑力计算应考虑其速度信息.为解决局部最小问题,对机器人已走路径进行跟踪监督,当机器人路径在一段时间内出现重复时,确认其已处于陷阱状态,继而采用沿墙走的策略来摆脱陷阱.仿真结果验证了算法在复杂动态环境下的实时性和有效性.

Sideslip-force based local path planning in complex and dynamic environment

The local path planning of mobile robots was investigated.A new approach of virtual sideslip-force was proposed to solve the oscillation problem of traditional virtual force in cluttered workspace.Instead of repulsive force,each obstacle casts a sideslip-force on the robot,and the robot's obstacle avoidance motion is determined by the resultant force.The sideslip-force of static obstacle is decided by the robot related distance,bearing and the target's bearing, while for dynamic obstacle the robot velocity is also considered.The approach supervises former path to solve the "local trap" problem.If a repeated path is found during a certain continuous period,the trap situation is reckoned and a wall-following strategy is started to escape the trap.Simulation demonstrates the effectiveness and real-time property of the algorithm in complex dynamic environment.