Research Of Elevator Group Control Experimental System Based On Intelligent Algorithm

One of the basic problems of the elevator group control is to predict the traffic flow of the elevator. Elevator traffic flow has the characteristics of nonlinearity time series and small sample size. Support vector machine is applied to predict elevator traffic flow of time series. As for the traits of elavator triffic requirement, which is characterized by degeneration and uncertainty, iterative learning predictive models based on LS-SVM has been set up. The model has the ability to dynamically track changes in traffic demand. Through the use of support vector machine technology, we get continuous traffic demand prediction function. Upon examination, it’s obvious that the model has good predictive effect.As for various elevator traffic model, it’s suggest that we should carry out the coculation according to corresponding control algorithm, and therefor it may greatly improve the performance of elevator group control system. Using differential evolution neural networks for pattern recognition of elevator traffic, DE algorithm has strong global and local search capability.The design of the laboratory elevator group control hardware system and its simulation has already been Completed. It includes CAN bus elevator group control test bench hardware design and Elevator Group Control Simulation software design system. Elevator Group Control test bench is designed into three subsystems:bus communication, call signal and group control and monitoring subsystem. It stands in line with the actual structure of the elevator group control system.