A Quadratic Programming Approach to Path Smoothing

This paper presents a method for smoothing a path in an environment with obstacles. Some characteristic nodes of the path are updated in each iteration by solving a quadratic program, which is formulated based on the smoothness constraints and the local environment information. The generated path satisfies the prescribed smoothness constraints, such as bounds on the curvature, and avoids any collision with obstacles. The proposed method is easy to implement and computationally efficient.