基于分离障碍物的地下车库车位排布算法

Parking space arrangement algorithm of underground garage based on separated obstacles

地下车库中的障碍物是车位排布中除不规则外轮廓边界之外最复杂的约束条件,为此本文提出了一种根据障碍物分离处理的启发式算法。根据障碍物分布,将整个车库区域分为带障碍物和不带障碍物两部分,对其分别使用不同的方法计算。其中,包含障碍物部分通过像素分割构建用于描述空旷程度的矩阵,并计算最优排布方案;不包含障碍物的部分通过遗传算法处理,同时考虑两部分结合部位的车位损失,并合理规划承重柱,保证连通性。通过对不同形状、内部结构的地下车库的排布,该算法可以在有限运算时间内得到排布结果,为人工设计师提供参考辅助。

The obstacle in the underground garage is the most complex constraint condition in parking space arrangement except for irregular outer contour boundary, so this paper proposes a heuristic algorithm based on obstacle separation. In this paper, the whole garage area is divided into two parts with and without obstacles according to the distribution of obstacles, which are calculated by different methods. The part containing obstacles is divided by pixels to construct a matrix to describe the degree of emptiness and calculate the optimal layout scheme. The part without obstacles is processed by genetic algorithm. Meanwhile, the parking loss of the part combining the two parts is considered, and load-bearing columns are planned reasonably to ensure connectivity. Through the layout of underground garages with different shapes and internal structures, the algorithm can obtain the layout results in limited operation time, which can provide reference for human designers.