Abstract: Considering the parking guidance in the management systems of large-scale parking lots，in this study，a novel approach based on particle swarm optimization is presented.The crossover and mutation operators are introduced in the PSO.The information of environment constrains and path length is integrated in the fitness function which is constructed by neural network，the advanced algorithm overcomes the limitation of particle’s “prematurity” in the later phase of convergence.Simulation results are provided to verify the effectiveness and practicability of this approach.

Key words: parking guidance, path planning, particle swarm optimization, neural network

摘要： 针对现有停车场管理系统中存在的车位诱导问题，提出了一种新型的粒子群算法。该算法对粒子群算法加入交叉、变异算子，用神经网络构造适应度函数，该适应度函数描述了环境约束及路径的距离信息，该算法克服了粒子群算法在后期出现的粒子“早熟”现象。仿真结果表明了该方法的正确性和有效性。

关键词: 车位诱导, 路径规划, 粒子群算法, 神经网络