Real-Time Combinatorial Optimization for Elevator Group Dispatching

  • Authors:
  • Paul E. Utgoff;Margaret E. Connell

  • Affiliations:
  • -;Wilbraham, MA, USA

  • Venue:
  • IEEE Transactions on Systems, Man, and Cybernetics, Part A: Systems and Humans
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

A new algorithm Minimize Vexation, version 10 (MV10) for managing the elevators of a group is presented. Cars are dispatched in a way that maximizes efficiency for all the individuals who use an elevator in the group of elevators. Much information about the individuals in the group is inferred or estimated, greatly aiding the decision making process. A variety of questions regarding novel components of the algorithm are posed, and the answers are provided by way of deep statistical analysis. Insights are offered into the nature of various objective functions and their effects on system performance. The service time of the MV10 algorithm compares favorably to that of known approaches to the problem, demonstrated again through statistical analysis. For comparison purposes and perspective, the elevator dispatching problem is recast in a separate system in which all individual travel plans are known ahead of time, giving an indication of the best attainable efficiency.