Title: Planning and Control Models for Elevators in High-Rise Buildings

Author: Marja-Liisa Siikonen

Date: August, 1998

Status: Systems Analysis Laboratory Research Reports, A68, October 1997


Keywords: elevator planning, group control, call allocation, simulation, optimization, fuzzy logic, forecast

The passenger service level in an elevator system depends on the planning of the elevator group and on the control of the elevators. In this thesis the contribution of passenger and elevator traffic for an elevator group with multiple shafts is studied by simulations and theoretical analysis.

During this work an elevator control system, the Traffic Master System 9000 (TMS9000), was developed. The TMS9000 group control system optimizes passenger waiting times instead of hall call times, and the control actions can be started in advance, which is not the case with conventional controls. With the TMS9000 control system, average passenger waiting times became shorter and more balanced floor by floor compared to conventional controls. The functional properties of the TMS9000 group control system were tested with the Advanced Lift Traffic Simulator (ALTS), which is described in the thesis. ALTS has also been applied in elevator planning to predict hall call times in new buildings, or to estimate improvement potential in hall call times for existing elevator groups. Simulation has revealed new correlation between the passenger and elevator traffic. In this thesis, new equations and planning criteria for the mean passenger journey time and ride time inside the car during an up-peak are introduced, and the results are verified with simulations.