Research On Belt Conveyor Start Control Strategy

Belt conveyor has been widely used by coal, chemicals, metallurgy,mining, port, power and other industries. With the belt conveyor gradually tothe direction of the long-range, heavy-duty, high-speed, startup problems havebecome increasingly serious. Belt conveyor system will produce distortionduring the boot process. if take direct start or start-up time is too short, theacceleration will be bigger, so that make the visco-elastic deformation of theconveyor belt larger, eventually leading to a dynamic tension, impact directlyon the conveyor system components and shorten their useful life. In addition,the belt conveyor working environment is relatively poor, so it also has certainrequirements on the robustness of the controller. Therefore, the large beltconveyor start-up process is a very important research topic.In response to these problems, the paper does the work of the followingaspects:First of all, use finite element analysis method to establish the dynamicmodel of belt conveyor system. The model focuses considering theviscoelastic properties of the conveyor belt, according to analyze the dynamiccharacteristics of several typical viscoelastic model, finally fix the Kelvinmodel as a dynamic model of the conveyor belt infinitesimal segment, thenorder all of infinitesimal segment dynamics model, eventually forming asimple kinetic model of the entire conveyor system.Second, start processing use the S-type speed controllable curve tocontrol the belt conveyor acceleration. Belt conveyor start processing itself isan unstable process, S-type speed start-up curve can achieve the smooth startof the system, thus it can effectively reduce the shock of the system.Then, focus on the basic theory of the PID control, fuzzy theory, and onthe basis of MATLAB, make a simulation experiment about the start of thebelt conveyor, Laid the foundation for the discovery of the superior controltheory.Finally, for insufficient controller, study RBF neural network. On onehand, use RBF neural network for system identification, and optimize thelearning rules, improve the original PID controller; On the other hand, takefuzzy RBF Neural Network as the controller of the belt conveyor system.Simulation results show that RBF neural network-based controllers have beenquite good control effect.