Application of the model based on fuzzy behavior charts in the advising control system of rotary cement kiln

Abstract

Article considers the construction of operation model of the cement kiln node “The flue gases temperature” in the form of fuzzy behavior chart. We propose an application of the model constructing method by usage of the fuzzy approach, combined with fuzzy charts of nodes behavior. In addition are shown the ways of adapting the parameters of membership functions with the help of learning algorithms for neural networks. Also, article describes the solution for such several problems as the initial selection of membership functions and its optimal clarification based on the training data. Previously created smoothing algorithm used for solving the problem of reducing the random noise in the signals, coming from the sensors. It basically consists in prediction of the future size of the technological values and their subsequent approximation by k-th order polynomial. The resulting model is necessary for development of the logical structure for kiln control and the implementation of the joint operation of all kiln nodes for development the advising control system for rotary kiln of cement clinker calcination process.