基于混合遗传算法层间交通模式下的电梯群控派梯新方法

A Control Algorithm of Elevator Group Control System Based on Hybrid Genetic Algorithm with Inter-floor Traffic Pattern

提出一种基于混合遗传算法的层间交通模式下的电梯群控派梯算法。以乘客的平均等待时间最小为目标寻求最优派梯方案,根据电梯的运行状态和各层站的外呼信号,构造了适应度函数。在进化初期阶段,采用本标准遗传算法在解空间中全局搜索,当种群收敛到最优解附近时,引入自适应正交局部搜索算子,以提高算法的局部搜索能力,仿真结果表明,该算法克服了标准遗传算法后期收敛速度慢的缺点,同时较好地满足了电梯群控系统的实际要求。

A control algorithm of elevator group control system with inter-floor traffic pattern based on hybrid genetic algorithm is proposed in this paper. In this algorithm, we will search an optimal assigning method with the goal of passenger＇s average waiting time reaching the minimum and construct the adaptive function according to elevators＇ running state and the hall-call signal among layers. At the early stage of the evolution, we use standard genetic algorithm to do the global search in the result series. If the population is convergent around the optimal result, then we can put the adaptive orthogonal local search into the algorithm to improve the local search ability of it. The result of simulation demonstrated that it conquers the disadvantage of the standard genetic algorithm which is very slow convergence at the late stage, thereby satisfying the real time control of elevator group control system.