基于Fast R-CNN的动态分区多轿厢电梯调度研究  

Multi-car Elevator Systems Using Dynamic Zoning Based on Fast R-CNN

作　　者：刘剑[1] 赵悦[1] 徐萌 常玲[2] LlU Jian;ZHAO Yue;XU Meng;CHANG Lin(College of Information and Control Engineering, Shenyang Jianzhu University, Shenyang 110168, China;Department ofInformation and Control Engineering, Shenyang Urban Construction University, Shenyang 110168, China)

机构地区：[1]沈阳建筑大学信息与控制工程学院,沈阳110168 [2]沈阳城市建设学院信息与控制工程系,沈阳110167

出　　处：《控制工程》2019年第2期208-214,共7页Control Engineering of China

基　　金：国家自然科学基金(61272253);辽宁省自然科学基金(201602616);辽宁省教育厅科学研究项目(L2015443);住建部项目(2015-K2-015)

摘　　要：为了在有限的空间内有效提高垂直交通系统的运载效率,在原有的一条井道内安装多个电梯轿厢,即"一井多梯"或称多轿厢电梯应运而生,它是提高运行效率、解决垂直交通拥挤的最佳选择,也是全世界垂直交通运输领域研究的前沿问题。针对多轿厢电梯的调度问题,笔者提出了一种基于Fast R-CNN的动态分区多轿厢电梯调度方法,首先通过FastR-CNN模型检测厅前和轿厢内人数;然后运用检测结果进行合理派梯;最后根据派梯任务划分轿厢的运行区域,实现合理调度。通过实验仿真表明,该方法适用于电梯的各种交通模式,具有较高的运行效率和灵活性。In order to improve the efficiency of the vertical transportation system effectively in limited space,multiple elevator cars are installed in a single elevator shaft, which is called "one shaft, multi-cars" elevator or multi-car elevator, which is the best choice to raise the transportation efficiency and solve the congestion in the vertical traffic. For the multi-car elevator scheduling, this paper puts forward a multi-car elevator scheduling method using dynamic zoning based on fast region-based convolutional network(Fast R-CNN). Firstly, fast R-CNN model is used to detect the number of people in the front of the hall and in the car;Then, elevators are reasonably dispatched according to the detection results;Finally, a zone is allocated to a car according to the assignment of calling to the car. The experimental results show that the proposed method is applicable to all kinds of elevator traffic patterns, and has higher operation ef?ciency and flexibility.

关 键 词：多轿厢电梯 FAST R-CNN 模型 动态分区 电梯调度 

分 类 号：TP202.7[自动化与计算机技术—检测技术与自动化装置]