Study On Model Order Redction Of Belt Conveyor

With the rapid development of modern industrial technology,there are highrequirements for the delivery of bulk materials in tne areas of mining, buildingmaterials, ports, chemicals, electricity, food, metallurgy, coal. Belt conveyorbecome the important equipment to transport the bulk materials continuously.The order of the belt conveyor system model is very high. It’sdifficult to makereal-time control and design controller. The system model with high order need toreduce the order.The paper analyses the viscoelastic properties of belt and establishes dynamicmodel. The conveyor belt is divided into30short sections. The paper obtains thestate space description of the belt conveyor system of SIMO （single input multipleoutput）.In this paper, the reduced-order methods are reviewed.This paper simplifies theoriginal system with the order of60into reduced-order system with the order of10appling five kinds of reduced-order algorithm, such as the Arnoldi algorithm,balanced truncation algorithm, Laguerre-SVD algorithm, the Lanczos algorithmand the padé algorithm.This paper adoptsH₂norm to assess the similarity of the original model andreduced order model quantitatively. The order of reduced-order system is decidedbyH2norm. This paper analyzes the degree of approximation of the originalmodel and reduced-order model’s logarithmic frequency characteristics curve toassess the similarity of the original model and reduced order model qualitatively.This paper analyzes the comparison chart of the five kinds of reduced-ordermethod, considering the overall reduced-order effect of the five kinds ofreduced-order algorithm. The Arnoldi reduced-order algorithm is closer to theoriginal system model, the Arnoldi reduced-order algorithm is more suitable for themodel order redction of the belt conveyor system, compared with other four kindsof reduced-order algorithm.