Passenger Volume Estimation Based on the Relational Model of Visual Density for Elevator Group-Controlled System

Abstract

The uncertainty of passenger volume is the bottleneck of modern elevator group-controlled system, which influences the optimization process of group control and depresses the control performance. This paper presents a novel passenger volume estimation method. According to the specific conditions of an elevator, background subtraction is used to gain the visual density. And then the relational model of visual density is obtained with the system identification. By considering the status inside elevators, the directions of passenger movement, and the number of waiting passengers, this model offers effective parameters to elevator group-controlled strategy and guarantees the real time of control system. Compared with the traditional Hough transform detection, the method in this paper is an easy and efficient way to learn dynamic traffic by analyzing passenger traffic of elevator group-controlled system.