基于遗传算法的电梯群控研究

Research of Elevator Control Group Based on Genetic Algorithm

针对电梯群控系统的特点,设计出一种更适合的梯群控制的遗传算法。本算法中采用了整数编码和可进行种群竞争的双种群机制,设计了以候梯时间、乘梯时间、系统能耗为群控目标的多目标适应度函数,并在选择操作中引入个体最优选择策略,在交叉操作中构造了与遗传代的数目、预交叉个体本身特点相结合的交叉方式,在变异操作中应用了两点对换和位点变异相结合的变异方法,并设计了从最优解集合中选择最优解的评价函数。经过模拟仿真,运行实验结果表明了此方案的可行性和优越性。

An much more suitable genetic algorithm is designed for elevator group control contraposed to its characteristics. In this arithmetic,a real-coded and feasible two-stock competition mechanism are adopted, and a multi-objective fitness function is designed in the genetic algorithm. Moreover, the individual optimal choice strategy is applicated in selection. In crossover, it structures the way of the crossover, which is combined to the number of generation and their own characteristics of the pre-cross individual.The mutation method which is combining the way of one-points changing and the way of two-points swapping are applied.And in this paper, designing the evaluation function, which is used to select the optimal solution from the collection of optimal solution. Then the genetic algorithm is used in the elevator group control. The arithmetic is verified through simulating and emulating , whose results show the feasibility and the superiority of the programmer.