A study of applying genetic network programming with reinforcement learning to elevator group supervisory control system

Jin Zhou, Toru Eguchi, Shingo Mabu, Kotaro Hirasawa, Takayuki Furuzuki, Sandor Markon

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Citations (Scopus)

Abstract

Elevator Group Supervisory Control System (EGSCS) is a very large scale stochastic dynamic optimization problem. Due to its vast state space, significant uncertainty, and numerous resource constraints such as finite car capacities and registered hall/car calls, it is hard to manage EGSCS using conventional control methods. Recently, many solutions for EGSCS using Artificial Intelligence (AI) technologies have been reported. Genetic Network Programming (GNP), which is proposed as a new evolutionary computation method several years ago, is also proved to be efficient when applied to EGSCS problem. In this paper, we propose an extended algorithm for EGSCS by introducing Reinforcement Learning (RL) into GNP framework, and expect to make an improvement of the EGSCS' performances since the efficiency of GNP with RL has been clarified in some other studies like tile-world problem. Simulation tests using traffic flows in a typical office building have been made, and the results show an actual improvement of the EGSCS' performances comparing to the algorithms using original GNP and conventional control methods.

Original languageEnglish
Title of host publication2006 IEEE Congress on Evolutionary Computation, CEC 2006
Pages3035-3041
Number of pages7
Publication statusPublished - 2006
Event2006 IEEE Congress on Evolutionary Computation, CEC 2006 - Vancouver, BC
Duration: 2006 Jul 162006 Jul 21

Other

Other2006 IEEE Congress on Evolutionary Computation, CEC 2006
CityVancouver, BC
Period06/7/1606/7/21

Fingerprint

Network Programming
Genetic Network
Supervisory Control
Elevators
Reinforcement learning
Reinforcement Learning
Genetic Programming
Control System
Control systems
System Performance
Railroad cars
Dynamic Optimization Problems
Resource Constraints
Office buildings
Stochastic Optimization
Stochastic Dynamics
Evolutionary Computation
Tile
Traffic Flow
Evolutionary algorithms

ASJC Scopus subject areas

  • Artificial Intelligence
  • Software
  • Theoretical Computer Science

Cite this

Zhou, J., Eguchi, T., Mabu, S., Hirasawa, K., Furuzuki, T., & Markon, S. (2006). A study of applying genetic network programming with reinforcement learning to elevator group supervisory control system. In 2006 IEEE Congress on Evolutionary Computation, CEC 2006 (pp. 3035-3041). [1688692]

A study of applying genetic network programming with reinforcement learning to elevator group supervisory control system. / Zhou, Jin; Eguchi, Toru; Mabu, Shingo; Hirasawa, Kotaro; Furuzuki, Takayuki; Markon, Sandor.

2006 IEEE Congress on Evolutionary Computation, CEC 2006. 2006. p. 3035-3041 1688692.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Zhou, J, Eguchi, T, Mabu, S, Hirasawa, K, Furuzuki, T & Markon, S 2006, A study of applying genetic network programming with reinforcement learning to elevator group supervisory control system. in 2006 IEEE Congress on Evolutionary Computation, CEC 2006., 1688692, pp. 3035-3041, 2006 IEEE Congress on Evolutionary Computation, CEC 2006, Vancouver, BC, 06/7/16.
Zhou J, Eguchi T, Mabu S, Hirasawa K, Furuzuki T, Markon S. A study of applying genetic network programming with reinforcement learning to elevator group supervisory control system. In 2006 IEEE Congress on Evolutionary Computation, CEC 2006. 2006. p. 3035-3041. 1688692
Zhou, Jin ; Eguchi, Toru ; Mabu, Shingo ; Hirasawa, Kotaro ; Furuzuki, Takayuki ; Markon, Sandor. / A study of applying genetic network programming with reinforcement learning to elevator group supervisory control system. 2006 IEEE Congress on Evolutionary Computation, CEC 2006. 2006. pp. 3035-3041
@inproceedings{a90102d0114d4a20a376b0d71b9fa720,
title = "A study of applying genetic network programming with reinforcement learning to elevator group supervisory control system",
abstract = "Elevator Group Supervisory Control System (EGSCS) is a very large scale stochastic dynamic optimization problem. Due to its vast state space, significant uncertainty, and numerous resource constraints such as finite car capacities and registered hall/car calls, it is hard to manage EGSCS using conventional control methods. Recently, many solutions for EGSCS using Artificial Intelligence (AI) technologies have been reported. Genetic Network Programming (GNP), which is proposed as a new evolutionary computation method several years ago, is also proved to be efficient when applied to EGSCS problem. In this paper, we propose an extended algorithm for EGSCS by introducing Reinforcement Learning (RL) into GNP framework, and expect to make an improvement of the EGSCS' performances since the efficiency of GNP with RL has been clarified in some other studies like tile-world problem. Simulation tests using traffic flows in a typical office building have been made, and the results show an actual improvement of the EGSCS' performances comparing to the algorithms using original GNP and conventional control methods.",
author = "Jin Zhou and Toru Eguchi and Shingo Mabu and Kotaro Hirasawa and Takayuki Furuzuki and Sandor Markon",
year = "2006",
language = "English",
isbn = "0780394879",
pages = "3035--3041",
booktitle = "2006 IEEE Congress on Evolutionary Computation, CEC 2006",

}

TY - GEN

T1 - A study of applying genetic network programming with reinforcement learning to elevator group supervisory control system

AU - Zhou, Jin

AU - Eguchi, Toru

AU - Mabu, Shingo

AU - Hirasawa, Kotaro

AU - Furuzuki, Takayuki

AU - Markon, Sandor

PY - 2006

Y1 - 2006

N2 - Elevator Group Supervisory Control System (EGSCS) is a very large scale stochastic dynamic optimization problem. Due to its vast state space, significant uncertainty, and numerous resource constraints such as finite car capacities and registered hall/car calls, it is hard to manage EGSCS using conventional control methods. Recently, many solutions for EGSCS using Artificial Intelligence (AI) technologies have been reported. Genetic Network Programming (GNP), which is proposed as a new evolutionary computation method several years ago, is also proved to be efficient when applied to EGSCS problem. In this paper, we propose an extended algorithm for EGSCS by introducing Reinforcement Learning (RL) into GNP framework, and expect to make an improvement of the EGSCS' performances since the efficiency of GNP with RL has been clarified in some other studies like tile-world problem. Simulation tests using traffic flows in a typical office building have been made, and the results show an actual improvement of the EGSCS' performances comparing to the algorithms using original GNP and conventional control methods.

AB - Elevator Group Supervisory Control System (EGSCS) is a very large scale stochastic dynamic optimization problem. Due to its vast state space, significant uncertainty, and numerous resource constraints such as finite car capacities and registered hall/car calls, it is hard to manage EGSCS using conventional control methods. Recently, many solutions for EGSCS using Artificial Intelligence (AI) technologies have been reported. Genetic Network Programming (GNP), which is proposed as a new evolutionary computation method several years ago, is also proved to be efficient when applied to EGSCS problem. In this paper, we propose an extended algorithm for EGSCS by introducing Reinforcement Learning (RL) into GNP framework, and expect to make an improvement of the EGSCS' performances since the efficiency of GNP with RL has been clarified in some other studies like tile-world problem. Simulation tests using traffic flows in a typical office building have been made, and the results show an actual improvement of the EGSCS' performances comparing to the algorithms using original GNP and conventional control methods.

UR - http://www.scopus.com/inward/record.url?scp=34547247623&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34547247623&partnerID=8YFLogxK

M3 - Conference contribution

SN - 0780394879

SN - 9780780394872

SP - 3035

EP - 3041

BT - 2006 IEEE Congress on Evolutionary Computation, CEC 2006

ER -