Dispatching approach optimization of elevator group control system with destination floor guidance using Fuzzy Neural Network

Aiming at the problem of uncertainty for information collection in traditional elevator group control system (EGCS), this study proposes a method about dispatching approach optimization for a new type of EGCS with destination floor guidance. The characteristic of destination floor guidance is information admixture of hall call registration and destination selection. EGCS with destination floor guidance can attain nearly complete and reliable information aggregation in advance to guide the passenger flow effectively. In order to realize the accurate prediction of input parameters, all of the hall call registrations are classified to four groups. Based on it, the optimized prediction algorithms of evaluation items are established, including the average waiting time of passengers (AWT), the average riding time of passengers (ART), rate of waiting long time of passengers (LWR), and riding power consumption (RPC). Especially, Fuzzy Neural Network (FNN) is applied to optimize multiple objectives for realizing reasonable dispatching approach. The simulated results show that the algorithm applied in the paper improves the evaluation items by 10% at least as compared with the traditional algorithm and it indicates an excellent application to enhance the holistic capability of EGCS.

I. Introduction

At the present time, the study about EGCS with destination floor guidance is still in the early stage at home and abroad. Foreign scholars have done some systemic researches on the algorithm of single-car elevator system with destination hall call registration [1]–[2]. In their study, it was shown that operation efficiency of a single-car elevator system with destination hall call registration can be improved remarkably by objective function combinatorial optimizing and genetic network programming in the majority of passenger flow pattern. However, it didn't research the dispatching approach of EGCS. Domestic scholars have done the researches on dispatching approach of EGCS with destination floor guidance [3]. In their study, they designed the intelligent model of neural network programming for optimizing evaluation items of EGCS. However, the rationality of control strategy and the integrity of simulation results need further improvement. In this study destination floor guidance has appeared to deal with the information on the destination floor of passengers before getting in elevators instead of putting the cage call button in elevators. Especially the prediction algorithms of evaluation items are established. Through the intelligent control programming of FNN the veracity of evaluation items and functions are enhanced strongly, otherwise the dispatching approach of EGCS can be executed effectively at complex passenger flow.