A multi-mode prediction method for elevator traffic flow based on classification off-line

In order to achieve the on-line traffic flow prediction, a novel multi-mode prediction method based on mode classification off-line is proposed. Firstly, the elevator traffic flow is classified off-line into patterns by the two-stage Clustering Algorithm, Artificial Immune C-Means Clustering Algorithm (AI C-Means CA). Then Gaussian Mixture Model (GMM) is used to model the multi-mode elevator traffic flow. And the EM algorithm is utilized to estimate the parameters of GMM to predict elevator traffic flow on line. Finally, the effectivity of this prediction method is validated by comparing simulation results with other prediction methods.