基于灰色预测理论的电梯交通流组合预测  

Combined forecast of elevator traffic flow based on gray theory

作　　者：李晨 徐克辉 姜文刚[1] LI Chen;XU Keihui;JIANG Wengang(School of Electronic and Information,Jiangsu University of Science and Technology,Zhenjiang 212003,China;Department of Planning and Development,China Shipbuilding Industry Company Limited,Beijing 100097,China)

机构地区：[1]江苏科技大学电子信息学院,镇江212003 [2]中国船舶重工集团有限公司规划发展部,北京100097

出　　处：《江苏科技大学学报（自然科学版）》2019年第1期57-65,共9页Journal of Jiangsu University of Science and Technology:Natural Science Edition

摘　　要：电梯乘客交通流预测是电梯智能群控调度的前提,有效提高电梯乘客交通流的预测精度能够使电梯智能群控系统更加合理派梯,从而缩短乘客候梯时间,减少电梯不必要的停靠次数,节约电梯运行的能源消耗.针对电梯乘客交通流时间序列小样本特征,根据等维新息原则,提出一种改进的灰色GM(1,1)模型对电梯交通流分别进行周期特性及发展变化趋势预测.综合考虑电梯交通流的周期特性及发展变化趋势,将两组预测数据进行最优加权组合预测.通过交通流数据最优加权组合预测与小波神经网络预测模型相比较表明,在实际获取的乘客交通流数据量较少的情况下,最优加权组合预测的预测误差更小.将两种预测方法的结果进行群控调度,仿真结果表明,组合预测方法使智能群控系统更加合理派梯,缩短了乘客乘候梯时间,节约了电梯能耗.Elevator passenger traffic flow forecast is the premise of elevator intelligent group control, the effective improvement of the accuracy of the elevator traffic flow can make the intelligent group control system more reasonable to send the ladder, thus shortening the passengers waiting for the elevator time and reducing unnecessary parking of the elevator, saving elevator running energy consumption. According to the principle of equal dimension information, an improved gray GM(1,1)model is proposed to forecast the periodic characteristics and development trend of elevator traffic flow. Considering the cycle characteristics and development trend of elevator traffic flow, the two sets of forecast data to optimize the optimal combination of forecasting. Compared with the wavelet neural network prediction model, it is shown that the prediction error of the weighted combination forecasting model is smaller in the case of the actual amount of elevator traffic flow. The results of the two prediction methods are dispatched by group control. The simulation results show that the combined forecasting method makes the intelligent group control system more reasonable to dispatch the ladder, shortens the passenger waiting time and saves the elevator energy consumption.

关 键 词：电梯交通流 灰色模型 组合预测 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]