Working Condition Recognition of Cement Decomposition Furnace Based on ART-2 Neural Network

Abstract

This paper advances a two-stage ART-2 neural network method on working condition recognition of cement decomposition furnace. The process parameters describing cement decomposition furnace conditions are determined based on the process design requirements and the analysis of field operation experience. The field operating datas through mean filtering are determined as the first-class ART-2 network inputs, and the trend recognitions of the parameters are finished based on the trend recognition function. The results are determined as the second-class ART-2 network inputs,and the real time recognitions of decomposition furnace conditions are completed using the pattern recognition function. The simulation and practical operation show the effectiveness of the method.