基于矩形安全邻域的智能车移动仿真研究

Simulations on movement of intelligent vehicle based on rectangular safe neighborhood

为了利用仿真技术研究智能车在行进时的实时测控过程,先由法线法构建近似等距曲线作为仿真道路的边界;再利用非线性规划给出了道路内满意矩形安全邻域的寻找方法;通过道路边界变异,构建了有障碍道路的仿真模型。整个过程形成了一套基于矩形安全邻域的智能车移动仿真方法,并对智能车实时搜索矩形安全邻域及据之决策行进的过程进行了仿真。结果表明,该方法可以实时规划智能车的行驶路线,有效地应对突发行驶障碍等问题。

To simulate the monitoring process of an intelligent vehicle in automatic driving, firstly, the authors suggested a road model whose boundary consists of two approximate isometric curves. Secondly, this paper proposed a way for searching satisfying rectangular safe neighborhoods （RSN） in the road with a nonlinear programming method. Moreover, it costructed a simulation model with some obstacles in road by mutation of road boundary. Finally, it simulated that searching process of RSN and driving process for the intelligent vehicle based on the RSN in different roads. Simulation results show that, the method can plan driving route in unknown environments in real time, and deal with unexpected obstacles and other problems effectively.