Q: What is a local minimum and how is it handled?

A local minimum is a 'valley' where the potential is lower than all surrounding points but is not the goal, causing the robot to become stuck. It is the main weakness of the potential field method. Solutions include: (1) random walk to escape, (2) wavefront planning (stronger repulsion near obstacles only), (3) combining with RRT or A* search, and (4) using harmonic potential functions.

Q: When should I use RRT vs. the potential field method?

The potential field method is fast and suited for real-time path modification in environments with few obstacles or little risk of local minima. RRT explores the full workspace via random sampling, so it avoids local minima and can find paths in narrow passages and complex environments. However, it is more computationally expensive and post-processing is needed for smooth paths. In practice, global planners (A* or RRT) are combined with local planners (potential field).

Q: How is the potential field method used in CAE simulations?

Main applications include: (1) path computation for autonomous transport robots (AGV/AMR) in factory lines, (2) initial design of space debris avoidance trajectories, (3) material removal direction in structural topology optimization (analogous sensitivity field), and (4) intermolecular interaction models in molecular dynamics (similar mathematical framework). The computational cost is far lower than FEM/CFD, making real-time control integration feasible.

Question 1

What is the principle of potential field path planning?

Accepted Answer

In the potential field method, an attractive potential U_att=½k_att×d_goal² pulls the robot toward the goal, and a repulsive potential U_rep=½k_rep×(1/d_obs−1/d₀)² (when d_obs<d₀) pushes it away from obstacles. The robot moves according to the resultant force F=−∇U (negative gradient) and naturally follows the path of steepest descent toward the goal. The method is computationally light and simple to implement, but can trap the robot in local minima.

Question 2

What is a local minimum and how is it handled?

Accepted Answer

A local minimum is a 'valley' where the potential is lower than all surrounding points but is not the goal, causing the robot to become stuck. It is the main weakness of the potential field method. Solutions include: (1) random walk to escape, (2) wavefront planning (stronger repulsion near obstacles only), (3) combining with RRT or A* search, and (4) using harmonic potential functions.

Question 3

When should I use RRT vs. the potential field method?

Accepted Answer

The potential field method is fast and suited for real-time path modification in environments with few obstacles or little risk of local minima. RRT explores the full workspace via random sampling, so it avoids local minima and can find paths in narrow passages and complex environments. However, it is more computationally expensive and post-processing is needed for smooth paths. In practice, global planners (A* or RRT) are combined with local planners (potential field).

Question 4

How is the potential field method used in CAE simulations?

Accepted Answer

Main applications include: (1) path computation for autonomous transport robots (AGV/AMR) in factory lines, (2) initial design of space debris avoidance trajectories, (3) material removal direction in structural topology optimization (analogous sensitivity field), and (4) intermolecular interaction models in molecular dynamics (similar mathematical framework). The computational cost is far lower than FEM/CFD, making real-time control integration feasible.

Robot Path Planning · Potential Field Method Simulator

Potential Field Equations