Abstract: Local avoidance planning of intelligent vehicle is targeted at unknown environment or part of unknown environment. 2D lidar scanner and camera sensors were equipped on the intelligent vehicle. The information map was provided by the real-time environmental information for the danger to avoid planning. Three algorithms were proved correctly by real vehicle experiment under different ordinary weight, particles attract, and bionics path selection for reaction tentacle to avoid in time, optimal and smoothness factor for pure obstacle avoidance environment. In the Future Challenge 2010 (intelligent vehicle challenge), single pure obstacle avoidance was applied in the environment of the three algorithms in different environmental conditions have got good results.