Path Planning for Intelligent Vehicles Lane Change to Avoid Collision in Emergency Based on NSGA-III

Aiming at the requirement that intelligent vehicles should make timely lane change to avoid collisions when facing obstacles in front of them, A multi-objective collaborative optimization path planning algorithm for lane change collision avoidance is proposed, which is safe, comfortable, efficient, and has little impact on road circulation. Using the idea of micro-element method, the lane change collision avoidance trajectory is regarded as the superposition of multiple uniform circles in minimal time, and a multi-objective optimization model of lane change trajectory planning without knowing the end position of lane change in advance is established. The solution is solved by the nondominated sorting genetic algorithm based on reference points (NSGA-III). The path planning of lane change and collision avoidance can be completed in advance within the warning distance, and the value of vehicle warning speed under a certain warning distance could be obtained. Simulation results verify the effectiveness of the proposed method.