• International Journal of Control, Automation, and Systems
• /
• v.5 no.6
• /
• pp.674-680
• /
• 2007

The probabilistic roadmap (PRM) method, which is a popular path planning scheme, for a manipulator, can find a collision-free path by connecting the start and goal poses through a roadmap constructed by drawing random nodes in the free configuration space. PRM exhibits robust performance for static environments, but its performance is poor for dynamic environments. On the other hand, reinforcement learning, a behavior-based control technique, can deal with uncertainties in the environment. The reinforcement learning agent can establish a policy that maximizes the sum of rewards by selecting the optimal actions in any state through iterative interactions with the environment. In this paper, we propose efficient real-time path planning by combining PRM and reinforcement learning to deal with uncertain dynamic environments and similar environments. A series of experiments demonstrate that the proposed hybrid path planner can generate a collision-free path even for dynamic environments in which objects block the pre-planned global path. It is also shown that the hybrid path planner can adapt to the similar, previously learned environments without significant additional learning.

• Journal of Institute of Control, Robotics and Systems
• /
• v.17 no.4
• /
• pp.362-367
• /
• 2011

The collision-free path of a manipulator should be regenerated in the real time to achieve collision safety when obstacles or humans come into the workspace of the manipulator. A probabilistic roadmap (PRM) method, one of the popular path planning schemes for a manipulator, can find a collision-free path by connecting the start and goal poses through the roadmap constructed by drawing random nodes in the free configuration space. The path planning method based on the configuration space shows robust performance for static environments which can be converted into the off-line processing. However, since this method spends considerable time on converting dynamic obstacles into the configuration space, it is not appropriate for real-time generation of a collision-free path. On the other hand, the method based on the workspace can provide fast response even for dynamic environments because it does not need the conversion into the configuration space. In this paper, we propose an efficient real-time path planning by combining the PRM and the potential field methods to cope with static and dynamic environments. The PRM can generate a collision-free path and the potential field method can determine the configuration of the manipulator. A series of experiments show that the proposed path planning method can provide robust performance for various obstacles.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.50 no.3
• /
• pp.157-164
• /
• 2022

In this paper, we describe how to implement the mountainous terrain, radar, and air defense network for UAV path planning in a 3-D environment, and perform path planning and re-planning using the PRM algorithm, a sampling-based path planning algorithm. In the case of the original PRM algorithm, the calculation to check whether there is an obstacle between the nodes is performed 1:1 between nodes and is performed continuously, so the amount of calculation is greatly affected by the number of nodes or the linked distance between nodes. To improve this part, the proposed LineGridMask method simplifies the method of checking whether obstacles exist, and reduces the calculation time of the path planning through parallelization. Finally, comparing performance with existing PRM algorithms confirmed that computational time was reduced by up to 88% in path planning and up to 94% in re-planning.