Elevator – industrial lift truck – or stationary lift for vehicle – With call registration means – Shared by plural load supports
Reexamination Certificate
2003-12-23
2004-08-17
Salata, Jonathan (Department: 2837)
Elevator, industrial lift truck, or stationary lift for vehicle
With call registration means
Shared by plural load supports
C187S247000, C187S391000
Reexamination Certificate
active
06776264
ABSTRACT:
SUBJECT OF THE INVENTION
The present invention relates to a method and an apparatus for allocating calls given via the landing call devices of elevators comprised in an elevator group so that each call is served by one elevator in the elevator group. In particular, the invention relates to a method for stabilizing repeatedly defined allocation decisions in an allocation system based on a genetic algorithm.
PRIOR ART
When a passenger wants a ride on an elevator, he calls an elevator by pressing a landing call button on the floor. The elevator group control system receives the call and tries to decide which one of the elevators in the elevator group will be best suited to serve the call. This activity is call allocation. The allocation function is a kind of minimization problem in which, as a solution to the function, the aim is to find an elevator that minimizes the value of the function on the basis of one or more selected cost factors.
When the number of calls and elevators serving them increases, the minimization problem becomes significantly more complicated. This naturally increases the amount of computation capacity required. The quantity to be minimized in the allocation process may be passenger waiting time, passenger traveling time, number of stops of the elevators, or a combination of several cost factors weighted in different ways.
Traditionally, when a suitable elevator is to be found for a call, the reasoning is performed on a case-specific basis using complicated conditional structures. The final objective of this reasoning is to minimize a given cost factor describing the operation of the elevator group, typically the average passenger waiting time. Since the possible combinations of states of the elevator group are complex, the conditional structures are also complex and apt to have gaps, in other words, there occur situations in which the control does not work in an optima way. Likewise, it is difficult to take the entire elevator group into account as a whole. A typical example of this is the traditional collective control, in which a landing call is allocated to the elevator nearest to it and traveling in the direction of the call.
However, this simple optimization principle leads to a grouping of the elevators so that the elevators travel in a front in the same direction, and therefore to a deterioration of the performance of the elevator group as a whole.
When the cost factors for all possible route alternatives are to be determined, the required amount of computing work easily grows beyond the capacity of the processors. If the number of calls to be served is C and the building has L elevators, then the number of different route alternatives is N=L
C
. As the number of route alternatives increases exponentially with increase of the number of calls, systematic consideration of all route alternatives is impossible even in small elevator groups. This has restricted the practical application of route optimization.
Finnish patent application FI 951925 discloses a method for allocating landing calls in an elevator group. This method is based on forming a plurality of allocation options, each of which contains call data and elevator data for each landing call, said data together defining the elevator which is to serve the landing call. After this, the value of a cost function is computed for each allocation option and one or more allocation options are altered in respect of one or more data items comprised in it, whereupon the values of the cost functions of the new allocation options thus obtained are computed. In this way, on the basis of the cost functions, the best allocation option is selected and the currently active landing calls are allocated accordingly to the elevators of the elevator group.
The solution according to FI 951925 substantially reduces the required computation work as compared with calculating all possible route alternatives. In a method based on a genetic algorithm like this, the elevator group is treated as an entity, the cost function being optimized at the level of the entire elevator group. The optimization can be performed without considering individual situations and ways of coping with them. By modifying the cost function, a desired mode of operation can be achieved. It is possible to optimize e.g. passenger waiting time, call time, number of starts, traveling time, energy consumption, rope wear, operation of an individual elevator if the use of a given elevator is expensive, uniform use of the elevators and so on, or a desired combination of these.
If there are a large number of active landing calls, a single solution or allocation option that is clearly better than other options will not necessarily be found. To this situation, there are several solutions that are almost equally good. This is a case of so-called multi-modality.
As the above-described genetic algorithm is a search method of a stochastic nature, it does not guarantee that a search repeated over and over again will always produce the same solution or allocation option, even if the same or almost the same starting values are used. In the case of continuous allocation, this appears as an instability of allocation decisions. The calls allocated to the elevators in the elevator group change from one time of allocation to another. However, if it is known that there are several approximately equally good allocation options for the elevator group, then it does not make a difference which one of these allocation options is used. By contrast, it is important that the allocation option should not be altered each time the allocation option is selected again, but that a solution once selected should be retained as long as the starting parameters of allocation remain identical or sufficiently identical.
In practice, this means that, if the situation in the elevator group remains unchanged, i.e. no more calls are received and no elevators are added or removed from the group, then the elevator routes selected for a given set of calls have to be retained until even the last call has been served. If this is not the case, then an elevator in the elevator group may receive new control data every half second, which is cumbersome for the elevator to implement and leads to instability of the elevator group and the system.
The object of the invention is to disclose a new type of allocation method.
A specific object of the invention is to eliminate the above-mentioned drawbacks of prior-art allocation methods or at least to provide a substantial improvement to them.
A further alternative object of the invention is to disclose an allocation method that is substantially more stable and reliable than prior-art methods.
A further alternative object of the invention is to make it possible to stabilize the solution of an allocation problem and elevator group control so that an unchanged or almost unchanged set of calls is served by an unchanged allocation option as long as the allocation option in question produces a solution that is one of the best solutions obtained by a stochastic selection process.
As for the features characteristic of the invention, reference is made to the claims.
BRIEF DESCRIPTION OF THE INVENTION
The object of the invention is to achieve a new solution for allocating landing calls to the elevators in an elevator group in such a way that the control call received by each elevator changes as seldom as possible.
In the genetic method of the invention, to allocate calls to the elevators in the elevator group, several allocation options, i.e. chromosomes are generated, each of which contains for each active landing call a call data item and an elevator data item, and these data items or genes together determine the elevator that is to serve each landing call. For each chromosome generated, a fitness function value is determined. Next, one or more of the chromosomes are altered in respect of at least one gene, and fitness function values are determined for the chromosomes thus obtained. The search, i.e. the generation of new chromosomes is continued until an agreed term
Tyni Tapio
Ylinen Jari
Birch & Stewart Kolasch & Birch, LLP
Kone Corporation
Salata Jonathan
LandOfFree
Method for allocating landing calls does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for allocating landing calls, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for allocating landing calls will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3344882