Novel path curvature optimization algorithm for intelligent wheelchair to smoothly pass a narrow space

This paper presents a novel algorithm to address the smooth narrow pass traversing issue, which is based on optimizing the curvature of the wheelchair path. Being aware of the fact that the path smoothness is determined by the path curvature and its change rate, after calculating the position of the narrow pass relative to the base frame of the wheelchair from perception sensor data, the algorithm takes the curvature and its change rate of Bezier curve as the optimal objective, and the wheelchair heading and the condition that the Bezier curve polygon should be convex polygon as constraints, and plans a smooth and optimal path for the controlled wheelchair to follow. This process is iterated dynamically to enable the intelligent wheelchair to traverse the narrow pass smoothly. Simulation is firstly conducted to compare the performances of our method and the A*-based path planning navigation algorithm, which shows that the proposed algorithm is able to achieve more smooth path with smaller curvature when the wheelchair traverses narrow path. Furthermore, the algorithm can control the wheelchair to traverse narrow pass smoothly even without any global map and localization. Real experiment with detailed explanation of algorithm implementation is also given to verify the effectiveness of the proposed algorithm.