基于障碍物势场和模型预测的车辆避障路径规划

Vehicle Path Planning Based on Obstacle Potential Field and Model Predictive Obstacle

传统人工势场法应用于车辆路径规划时存在一些问题:障碍物的圆形斥力势场无法满足车辆在纵、横向上不同的安全距离要求,且该算法未考虑车辆动力学和执行机构约束。针对以上问题,对人工势场法中的障碍物斥力模型进行了改进,将改进后的人工势场法与模型预测控制相结合,提出了一种能够对障碍物进行分类处理的模型预测避障路径规划器。首先,对车辆与障碍物之间的距离根据纵、横向安全距离进行归一化处理,并对3种不同类型的障碍物分别设计了不同的斥力势场函数。然后,设计了模型预测避障路径规划控制器,将障碍物势场加入该控制器的目标函数中,以此来引导车辆躲避障碍物。仿真结果表明:该避障路径规划控制器能够在复杂环境下合理地处理不同类型障碍物,有效避免碰撞,生成路径平滑,保证车辆具有良好稳定性与安全性。

There are some problems when applying traditional artificial potential field method to vehicle path planning.The circular repulsive potential field of obstacles does not meet the requirements of longitudinal and lateral vehicle safety distances.In addition,the algorithm does not consider the constraints of vehicle dynamics and actuators.In this regard,combining the improved artificial potential field method with model predictive control,a model predictive obstacle avoidance path planning controller is proposed,which can classify and process obstacles.First,the distances between vehicles and obstaclesare normalized according to the longitudinal and lateral safety distance.Then,different repulsive potential field functions are designed for three types of obstacles.To guide the vehicle to avoid obstacles,the total obstacle repulsive potential field is added to the objective function of the model predictive controller.Finally,the simulation results show that the proposed controller can reasonably handle different types of obstacles in complex environments and effectively avoid obstacles.