An elevator intelligent scheduling system and method based on deep learning

Abstract

The invention discloses an elevator intelligent scheduling method and system based on deep learning, and the method comprises the steps: collecting image information, and obtaining all information ofpeople waiting for elevator taking in each elevator; storing and preprocessing video images, collected in real time, of people waiting for elevator taking in real time, which comprise key frame selection of monitoring continuous videos, and the standard of the key frame selection is one frame, most capable of reflecting information of all elevator taking people, in each section of elevator takingpeople waiting video; Finally, obtaining attribute information of the number of elevator passengers, the gender, the shape and the weight estimation of the elevator passengers in a key frame in the monitoring video, and sending the information obtained through algorithm analysis to an elevator intelligent scheduling module; Obtaining the stopping position of the current elevator in real time, andfeeding back an elevator dispatching control center distribution result to an elevator operation monitoring module in real time according to the obtained number of people taking the elevator and attribute information; The elevator operation monitoring module is responsible for elevator safety, elevator passenger safety and convenient maintenance.

Description

technical field [0001] The invention belongs to the field of elevator applications, and in particular relates to an elevator intelligent scheduling system and method based on deep learning. Background technique [0002] In modern life, whether in residential buildings, office buildings or shopping malls, the importance of elevators as an important means of commuting has become increasingly prominent. However, due to factors such as construction costs and building areas, the number of elevators in buildings is limited. For example, during the morning and evening peaks of residential buildings on weekdays and weekends in shopping malls, people taking elevators can only wait on the corresponding floors, which greatly wastes time. [0003] The publication number is CN 108217352A, and the application name is a Chinese patent application for an intelligent elevator dispatching system, which belongs to the field of intelligent elevator control, including a sensor system module, a...

AI Technical Summary

Problems solved by technology

However, in this solution, the AI ​​processing module that mainly analyzes image data mainly analyzes face data. The AI ​​processing module cannot analyze information such as the occupation and body shape of the person taking the elevator through face analysis. To the real-time intelligent dispatching of elevators

Images