Study On Optimized Dispatching Algorithm Of Elevator Group With Destination Floor Guidance

The traditional management pattern of elevator traffic information easily causes the "Congestion Phenomenon" and "Over-loading Phenomenon".Meanwhile,it could generate huge of interferences and lavishes to the limited elevator resources,especially,which be serious in the peak traffic condition.Based on the pattern of elevator group dispatching system of traditional buildings,the destination floor guidance mechanism is proposed and the optimal dispatching algorithms to the novel type of elevator group control system is studied. This topic belongs to a research highlight in the field of intelligent control of building automation.Aiming at the problems of uncertainty to information collection and low efficiency to dispatching control in elevator group dispatching systems,based on the traditional elevator group dispatch system,bring forward a new call elevator scheme which is named Destination Floor Guidance System(DFGS).The separate input modes of traditional hall call registration and the destination selection are improved to a single-input mode which makes the acquired ensemble of hall call communication determinate and integrated.Through the lead function of DFGS can efficiently prevent all the passengers from crowding one elevator.At the same time DFGS obviously improve the evacuation to passengers in flood tide and crowed degree of elevator.For optimizing elevator control,the conception of zoning is proposed which means one elevator just serves one concentrated region and the elevator which serves the high-region could do rapid running between the lowest floor of the certain region and the hall.Dynamic zoning of elevator group is solved by dynamic programming algorithm.Be based on the call elevator information which is gathered by DFGS,forecasting the average waiting time,the average ridding time,long waiting percentage,and power consumption.Not only forecast the new call elevator information but also forecast the deferred time of the information which is already dispatched,because of answering the new call elevator information.Neural network technology is applied in the control of elevator group.The evaluation items are used to be the inputs of network for constructing the model of network.Aiming at different traffic pattern, through sample training and learning,the weights are adjusted to fulfill the optimal dispatching policy.This way can efficiently improve the holistic capability of elevator dispatch system. In Windows operating condition,using Visual C++ as object-oriented development tools, the elevator group control system is designed realizing dynamically running and simulating of the visible elevator group dispatch system.Under typical up-peak traffic pattern,through simulating run,analyzing data,and comparing among different algorithms,it's proved that the design of dispatch system can across-the-board ameliorate the aspects of the dispatch quality, efficiency and economization energy and has actual applied value.