Performance Evaluation of an Option-Based Learning Algorithm in Multi-Car Elevator Systems

In this letter, we present the evaluation of an option-based learning algorithm, developed to perform a conflict-free allocation of calls among cars in a multi-car elevator system. We evaluate its performance in terms of the service time, its flexibility in the task-allocation, and the load balancing.

[1]  Hiroyoshi Miwa,et al.  Algorithm for Controlling Multi-Car Elevator Systems Based on Procedures Estimating Efficiency of Passenger Transport and Call Assignability , 2009, IEICE Trans. Fundam. Electron. Commun. Comput. Sci..

[2]  S. Markon,et al.  Traffic-sensitive controllers for multi-car elevators; design, multi-objective optimization and analysis , 2007, SICE Annual Conference 2007.

[3]  Toshiyuki Miyamoto,et al.  Option-Based Monte Carlo Algorithm with Conditioned Updating to Learn Conflict-Free Task Allocation in Transport Applications , 2011, IEICE Trans. Fundam. Electron. Commun. Comput. Sci..

[4]  Shingo Mabu,et al.  Multi-car elevator group supervisory control system using Genetic Network Programming , 2009, 2009 IEEE Congress on Evolutionary Computation.

[5]  Yuki Kuroda,et al.  An Algorithm to Minimize Average Service Completion Time for the Group Controller of Multi-Car Elevator Systems , 2008, IEICE Trans. Fundam. Electron. Commun. Comput. Sci..

[6]  Ronald A. Howard,et al.  Dynamic Programming and Markov Processes , 1960 .

[7]  Toshiyuki Miyamoto,et al.  MceSim: A Multi-Car Elevator Simulator , 2008, IEICE Trans. Fundam. Electron. Commun. Comput. Sci..

[8]  S. Markon,et al.  Simulation-based optimization of a controller for multi-car elevators using a genetic algorithm for noisy fitness function , 2003, The 2003 Congress on Evolutionary Computation, 2003. CEC '03..

[9]  Doina Precup,et al.  Between MDPs and Semi-MDPs: A Framework for Temporal Abstraction in Reinforcement Learning , 1999, Artif. Intell..