18-12-2012, 05:28 PM
Design and Implementation of a Fuzzy Elevator Group Control System
1Design and Implementation.pdf (Size: 336.42 KB / Downloads: 96)
Abstract
Elevator group control systems (EGCS’s) are the
control systems that systematically manage three or more elevators
in order to efficiently transport passengers. Most EGCS’s
have used the hall call assignment method to assign elevators
in response to passengers’ calls. This paper proposes a control
strategy generation method, a hall call assignment method based
on the fuzzy theory, and then the fuzzy elevator group control
system (FEGCS). The control strategy of FEGCS is made using
the classification of the passenger traffic and system manager’s
requirements, and the hall calls are assigned to suitable elevators
by the generated control strategy. The system is operated using
the given control strategy which is defined by the system manager.
The proposed system shows better results than the conventional
methods in simulations and is under commercialization by an
industrial company.
INTRODUCTION
ELEVATOR group control systems (EGCS’s) are control
systems that manage multiple elevators in a building in
order to efficiently transport the passengers. The performance
of EGCS’s is measured by several criteria such as the average
waiting time of passengers, the percentage of passengers waiting
more than 60 s, and power consumption [14], [15], [18],
[19]. EGCS’s manage elevators to minimize the evaluation
criteria; it is, however, difficult to satisfy all criteria at the
same time. Therefore, the EGCS is designed to satisfy each
criterion at certain levels. Nowadays, system managers want
to define the control strategy of EGCS’s, i.e., some managers
want to reduce the average waiting time while others may want
to reduce the power consumption.
ELEVATOR GROUP CONTROL SYSTEM
In this section, the general structure of EGCS will be
discussed. There are two hall call (up, down) buttons on a
floor, and multiple elevators as shown in Fig. 1. The EGCS
selects an elevator for the passenger who has pressed a hall
call button. The selected elevator moves to the floor where
the hall call occurred. To understand the EGCS, consider an
example of the elevator group control process.
1) A passenger who is going to the 15th floor from the
second floor presses the up hall call button.
2) The hall call signal is transmitted to the EGCS.
3) The EGCS selects an elevator to service the passenger.
4) The EGCS sends a message to the selected elevator.
5) The selected elevator moves to the second floor and the
passenger boards.
System Overview
Fig. 3 shows the structure of FEGCS. In Fig. 3, the FEGCS
manages eight elevators and the status of the FEGCS is
monitored via a terminal. The FEGCS consists of the traffic
data generation, control strategy management, hall call assignment,
data management, elevator management, and terminal
management parts. The control strategy generation and the
hall call assignment are the most important parts and have the
most effect on the performance of FEGCS. They are new parts
added to the conventional EGCS in this paper and the other
parts are somewhat modified.
Traffic data management part manages traffic data of passengers
by collecting, learning, and prediction. Traffic data, the
number of passengers who get on/off elevators on each floor,
is collected and learned periodically. Traffic data is predicted
for the next unit time to help the hall call assignment.
Hall Call Assignment
The hall call assignment part assigns hall calls to the
suitable elevators whenever new hall calls occur. Three fuzzy
inferences are implemented to test the suitability of each
evaluation criterion, and an ordered weighted average (OWA)
[20] operation is employed to get the overall suitability of
each elevator. Finally, the elevator having the largest overall
suitability is selected to service the new hall call.
EXPERIMENTAL RESULTS AND ANALYSIS
We have implemented a simulation environment to evaluate
the proposed elevator group control system’s performance.
The simulation environment consists of four parts as shown in
Fig. 10. The first part is a real elevator group controller and the
second part is a central processing unit (CPU) emulator which
is used for convenience of programming and debugging. The
third part is a car emulator which generates hall calls and car
calls like real elevators. It collects and displays the statistics
of the simulation. The car emulator is developed on an IBM
PC. The final part is a front end terminal. We use another PC
to program and debug our elevator group control system. In
the simulation environment, we used real hardware from an
elevator group control system and developed a car emulator
which simulates the elevator’s movement and operations.
CONCLUSIONS
In this study, the FEGCS is designed and implemented
to increase the performance of elevator systems. The control
strategy generation and hall call assignment parts, which are
the most important parts of the FEGCS, are developed and the
FEGCS is tested by computer simulation.
In the control strategy generation part, the passenger traffic
patterns are classified, and the membership functions used at
the hall call assignment are made by the classified traffic mode
and importance degrees of the evaluation criteria. In the hall
call assignment part, the hall calls are assigned to the suitable
elevators to service passengers. The suitabilities of elevators
for a hall call is given by the fuzzy inference and the system
selects an elevator by the rank of the overall suitability.