Group Control Of Elevator Traffic Flow Forecasting And Scheduling Strategy Research

With the rapid development of urbanization, the construction of intelligent building and high-rise building is the national comprehensive national power needs. High-rise buildings of vertical transport within group control elevator research appears very urgent.Based on the elevator group control as the research object, in order to improve the intelligent level of the system as main point, on the one hand, of traffic flow for real-time sampling and statistical analysis, and on the basis of a time period ahead of the characteristic value of traffic flow forecast, to avoid the real-time sampling of traffic flow lag brought about by the traffic patterns of falsely accused of serious problems. On the other hand, in order to design principle of human nature, and to reduce the passengers waiting time and take the ladder ladder time as the core, to reduce the system operation consumption as auxiliary purpose, leading the elevator group control system research direction, and create a different traffic model based on recognition of the objective function optimization scheduling strategy for group of ladder system of intelligent development provides theory basis and practical significance. The paper mainly work are as follows:(1) Introducing objective stratum of call pie ladder guide mechanism, design the elevator group control with RS-485serial bus would elevator group controller, the design of the building method of working from bus structure system, realized the signal of the exchange and transmission.(2) Based on the real-time change traffic flow prediction problem, choose the BP neural network structure, and through the initial learning and periodically for learning obtains network weights and new training sample, makes neural network for traffic flow changes can timely tracking forecast, especially for peak time forecast, the prediction accuracy is the highest.(3) To study the traffic pattern recognition problem, according to the forecast results of using regularization fuzzy neural network traffic patterns to recognize. Network test shows that this method of traffic flow of time-varying has good recognition ability.(4) Designed a population control elevator multi-objective optimization algorithm. Try to layer and outbound signal the same passengers arrange the same lift, to reduce the average passenger syndrome, reduce the average passenger elevator time by ladder, and reduce time elevator group control system operation energy losses as main optimization goal, using combinatorial optimization method will this several target function and structure of weighted combination of group control system, draws the evaluation function, realization of elevator YouPa ladder scheme multi-objective optimization scheduling. The simulation results show that the average service time in fluctuation peak mode raised more than15%, on average by ladder time, average time and stop syndrome ladder times, long hou ladder rate and long by ladder rates in interlayer pattern also greatly decrease.