Real-time self-tuning approach for intelligent elevator group control system

This study considers a real-time self-tuning intelligent elevator group control system with fuzzy logic in which the fuzzy membership functions are tuned based on the historical traffic data so as to maximize service quality and minimize the energy consumption. However, in terms of car-assignment function, most studies focus on car-assignment algorithms which are considered before group elevator control system is constructed, and rarely stress on passengers traffic which alters as elevators operate in real time. Accordingly, the parameter setting in advance may not fully meet the needs of instantaneity, flexibility, and energy saving towards elevator system design. In this research, a set of real-time self-tuning for intelligent elevator group control system is proposed and applied to the hall calls in practice. First, elevator records are used to list out the variables as possible elevator calls. Second, the optimization car-assignment is established on a basis of real-time self-tuning fuzzy rule to improve the car assignment efficiency including reducing average completion time and power consumption of elevators. The numerical results indicate the proposed method achieves better elevator assignment decision than other approaches in literature for the elevator group control problems more efficiency and energy saving.