超高层建筑电梯交通流量预测方法研究.docxVIP

毕业设计（论文） 超高层建筑电梯交通流量预测方法研究 摘要 本论文主要研究了超高层建筑电梯交通流的预测方法， 良好的电梯交通流分 析能够为电梯群组的调度单元提供有益的、具有预见性的指导，从而方便对楼内 交通进行静态分区，合理选择井道数、电梯台数和型号，提高电梯的运行效率。 论文先简述电梯的国内外发展状况，超高层建筑的相关知识 ，时间序列的特 点和基本概念。将实地考察的平安大厦的交通流数据分为两组，分别作为样本数 据训练模型和检验预测结果。首先应用 ARIMA建模方法和SAS软件，建立电梯 交通流预测模型并仿真出预测结果。然后用样本数据训练神经网络，基于 MATLAB神经网络工具箱建立交通流量预测模型并仿真。 将两种预测结果比较， 后者拟合预测效果要强于前者，于是便选择出较优的预测方法。 最后概述了如何利用预测结果来配置电梯，用 VC++的MFC建立对话框来 说明。 关键词：超高层建筑；电梯交通流；时间序列；预测； MATLAB神经网络 ABSTRACT Predict ion method of the elevator traffic flow of super high-rise buildi ng is studied in this paper. As known, proper an alysis on elevator traffic flow could provide useful and prospective guida nee for dispatch ing unit of elevator groups. Accord in gly, it will help to make a reas on able choice of the nu mber of shafts, the nu mber and types of elevators and enhance the efficie ncy of elevators, it also can make static divisio n of the buildi ng tran sport more easily. The article firstly describes the development of the lift at home and abroad briefly, the knowledge of super high-rise building and the characteristics and basic con cepts of time-series . The traffic flow data which is coun ted at Pingan Buildi ng is devided into 2 groups, one is used to train the model as a sample data and the other to test the results. First of all, the ARIMA modeli ng methods and SAS software are used to establish the elevator traffic flow prediction model and simulate the prediction outcome. The n the sample data is used to train the n eural n etwork and establish and simulate the traffic flow model based on the MATLAB n eural n etwork toolbox. And then by comparing the two prediction results, it is found that the latter effect of predictio n is better tha n the former, so the better predicti on method is selected. Fin ally, the overview of how to con figure the elevator using the predictio n results is summarized, which is described by establish ing dialog boxes with MFC of VC++. Key words: super high-rise building ； elevator traffic flow ； time se