The Study Of Elevator Optimizing Configuration Method Based On Markov Network Queuing Theory

In elevator systems, passengers' arrival and passengers' transport are very complicated stochastic process; therefore, elevator system is a greatly typical bulk-service queuing system. The randomicity of the elevator queuing system is mainly expressed by the randomicity of its input process: passenger traffic flow. So this paper first studies the analyzing method of peak traffic flow. Elevator cannot refresh or append at any moment, which determine that we need to confirm configuration parameters of reasonable amount, rated capacity and elevator traveling speed. This is also the content which elevator traffic analysis needs to study.This paper first analyzes the situation of the elevator traffic flow. hypothesis testing method is used to estimate what distribution passenger arrival time interval and batch arrival passenger number subject to. G.C.Barney offered method is used to obtain passenger percentage arrival rate and arrival rate curve of passengers based on minute.In this paper, we make use of stochastic server system theory to build elevator traffic flow network queuing model, which divided into three steps. The first step is that we supposes: passengers' arrival is Poisson process in every service station of elevator queuing network system, namely, arrival time interval submits to negative exponential distribution which is mutual independence; the time that every batch passengers' accepting server is negative exponential distribution which is mutual independence; passengers' queuing rule is First-In-First-Out in elevator system; In elevator queuing network system, for carrying passengers to destination floor as soon as fast, we need configure high speed elevators between first service station and second service station as well as between first service station and third service station. The second step is that we build queuing model to describe the elevator server system. The third step is to define the distribution of passengers in need of service and the percentage of up-direction or down-direction in each floor every 5 minutes in the building. Finally, Markov Theory is used to analyze and compute network queuing model of elevator server system, and deduce the distribution of queuing length and waiting time of the system which evaluating indicator, and obtain the relation curve between system evaluating indicator which different service intensity corresponding and system configuration factor. An instance to testify rationality of the model, and the validity of an important conclusion is verified: service intensity of elevator system must less than 80%. Above-mentioned conclusion offered theoretical basis for traffic analysis of elevator system. The software supplies an effective tool for farther traffic flow analysis and the study of dispatching system.After building elevator traffic flow network queuing model, we can use the model to carry through optimal configuration of elevator system. During conventional configuration process of elevator system, user gives traffic flow and a lot of influential factor are ignored. However, the elevator-optimizing configuration based on elevator traffic flow Markov network queuing model, which is according to real traffic flow station and synthetically consider all sorts of factor which influence elevator system configuration. Therefore, according to the method above-mentioned obtained elevator configuration project much more accord with real traffic flow.In sum, by applying elevator traffic flow network queuing model to optimize and configure elevator system, we not only have tested the validity of the model but also gained favorable results so far.