17-03-2014, 06:51 PM
In this paper, the new elevator group-supervisory control method to which
the expert system and the fuzzy rule are applied will be suggested. This
method determines the optimum car to answer a hall call by using the
knowledge bases, the production rule and the fuzzy rule, extracted from the
knowledge of experts in the elevator group-supervisory control.It has been
verified through simulation t h a t the new method can bring about considerable
improvement, reducing the average waiting time by 15-20%, the rate of long
waits (not less than 60 seconds) 3040% in comparison with the conventional
method in which a fixed evaluation function is used.
Furthermore, the knowledge-based systems suitable to the online and realtime
control for the implementation of this new method on 32-bit microcomputers
will be suggested. The result of the simulation has demonstrated
t h a t the suggested knowledge-based systems satisfy the required response time
100-150 msec in the group-supervisory control system with an immediate
prediction function.
INTRODUCTION
With the advancement of intelligent computerized builidings
in recent years, demands for the elevator group-supervisory control
with more sophisticated and diversified functions than before have
increased. The main function of the elevator group-supervisory
control is to determine which elevator is the optimum one in the
group to answer the hall call for shortening the waiting time as a
whole. In many cases, however, functions r e l a t i n g to
comfortableness, safety and economy are also desired in addition to
the main function.
A "Call Assignment Method" which aims to determine an
elevator to answer for each hall call has been adopted to most of the
group-supervisory systems at presnt. In this method, the car
selected for an assigned car is decided by the evaluation function
which includes the waiting time as an element.
It has been realized lately that besides utilizing specific
evaluation functions, the experts knowledge and experiential
rules, which have not been used because they make the evaluation
functions extremely complicated, need to be incorporated in the
group-supervisory control system in order to further impove the
control performance. The experts here mean engineers who have
been engaged in design, development and maintenance of elevators
under the group-supervisory control for long years.
This paper suggests a new elevator group-supervisory control
method where the group-supervisory control is basically exercised
by means of a knowledge-based system housing the knowledge and
experiential rules of experts in group-supervisory control.
First of all, Chapter I1 describes the system configuration,
functions and performance evaluation standards of the elevator
group-supervisory control system and clarifies the target where the
expert system will be introduced. In Chapter 111, an outline of
major group-supervisory control methods adopted in the past to
achieve the goal of the group-supervisory control and the
background which has necessitated the introduction of the expert
system will be described. Chapter IV deals with the detail of the
group-supervisory control method by means of the knowledgebased
systems suggested in this paper. Chapter V demonstrates the
effect brought about by the group-supervisory control method
based on the knowledge-based systems suggested by using a
simulator. In conclusion, Chapter VI shows how the new method
suggested will actually be adopted as well as the problems for
further discussion.
the expert system and the fuzzy rule are applied will be suggested. This
method determines the optimum car to answer a hall call by using the
knowledge bases, the production rule and the fuzzy rule, extracted from the
knowledge of experts in the elevator group-supervisory control.It has been
verified through simulation t h a t the new method can bring about considerable
improvement, reducing the average waiting time by 15-20%, the rate of long
waits (not less than 60 seconds) 3040% in comparison with the conventional
method in which a fixed evaluation function is used.
Furthermore, the knowledge-based systems suitable to the online and realtime
control for the implementation of this new method on 32-bit microcomputers
will be suggested. The result of the simulation has demonstrated
t h a t the suggested knowledge-based systems satisfy the required response time
100-150 msec in the group-supervisory control system with an immediate
prediction function.
INTRODUCTION
With the advancement of intelligent computerized builidings
in recent years, demands for the elevator group-supervisory control
with more sophisticated and diversified functions than before have
increased. The main function of the elevator group-supervisory
control is to determine which elevator is the optimum one in the
group to answer the hall call for shortening the waiting time as a
whole. In many cases, however, functions r e l a t i n g to
comfortableness, safety and economy are also desired in addition to
the main function.
A "Call Assignment Method" which aims to determine an
elevator to answer for each hall call has been adopted to most of the
group-supervisory systems at presnt. In this method, the car
selected for an assigned car is decided by the evaluation function
which includes the waiting time as an element.
It has been realized lately that besides utilizing specific
evaluation functions, the experts knowledge and experiential
rules, which have not been used because they make the evaluation
functions extremely complicated, need to be incorporated in the
group-supervisory control system in order to further impove the
control performance. The experts here mean engineers who have
been engaged in design, development and maintenance of elevators
under the group-supervisory control for long years.
This paper suggests a new elevator group-supervisory control
method where the group-supervisory control is basically exercised
by means of a knowledge-based system housing the knowledge and
experiential rules of experts in group-supervisory control.
First of all, Chapter I1 describes the system configuration,
functions and performance evaluation standards of the elevator
group-supervisory control system and clarifies the target where the
expert system will be introduced. In Chapter 111, an outline of
major group-supervisory control methods adopted in the past to
achieve the goal of the group-supervisory control and the
background which has necessitated the introduction of the expert
system will be described. Chapter IV deals with the detail of the
group-supervisory control method by means of the knowledgebased
systems suggested in this paper. Chapter V demonstrates the
effect brought about by the group-supervisory control method
based on the knowledge-based systems suggested by using a
simulator. In conclusion, Chapter VI shows how the new method
suggested will actually be adopted as well as the problems for
further discussion.