混合电梯群控系统建模及新型优化调度策略  

Hybrid elevator group control system modeling and new scheduling strategy

作　　者：赵小翠[1] 罗飞[1] 许玉格[1] 

机构地区：[1]华南理工大学自动化科学与工程学院

出　　处：《计算机应用研究》2010年第9期3290-3293,共4页Application Research of Computers

基　　金：国家自然科学基金资助项目(60774032);国家教育部高等学校博士学科点专项科研基金(新教师基金课题)资助项目(20070561006)

摘　　要：为了简化电梯系统建模问题的复杂性,国内外学者一直将电梯系统作为离散事件动态系统来分析,然而实质上电梯群控系统是一种典型的混合系统,包含了连续状态和离散状态以及复杂的人为逻辑策略。针对这一情况,打破此传统分析方法,将电梯作为混合动态系统进行分析,以细胞自动机理论为基础,加入多种连续速度曲线,建立了完整的混合电梯群控模型。仿真结果表明该模型结构简单,具有较强适应能力和可扩展性。同时,针对粒子群算法容易陷入局部最小值的缺陷,对其加以改进,提出一种新型的调度策略,并应用到混合电梯群控系统模型中,进行优化派梯。仿真结果表明该策略具有一定的可行性与优越性。本派梯策略为电梯群控算法提供了新思路与新方法,并扩展了粒子群算法的应用范围。To simplify the complexity of elevator system modeling,analyzed the elevator group control system（ EGCS） as a discrete event dynamic system. In fact,EGCS was a kind of typical hybrid dynamic system including continuous states,discrete states and complex human logic strategy. This paper broke the traditional method,researched EGCS as the hybrid system. Based on the cellular automata theory and several speed curve added in,built up the complete hybrid model of elevator system. This model is simple,has strong adaptability and scalability,can be applied in many different fields. This paper also proposed a new scheduling strategy by introducing simulated annealing to PSO,which overcame PSO’s weakness trapping in the local optimum easily,applied the new scheduling strategy to the hybrid elevator system. The simulation results demonstrate its feasibility and superiority in optimizing scheduling. This paper added the new scheduling algorithm,provided new ideas and new ways for EGCS,and expanded the application of genetic algorithm and particle swarm optimization.

关 键 词：电梯群控 混合系统 细胞自动机 粒子群算法 模拟退火算法 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]