Research On Soft Sensor For The Calcination Rate Of Preheater In NSP Cement Production

New Suspension Preheater dry process (NSP) cement production use suspension preheater and calcinations out of kiln as the core, basing on Distributed Control System. it makes the use of Homogenizing raw material and fuel, precalcining and preheating technology, energy-saving technology and equipment, to achieve automation, efficiency, quality, environmental protection under the process of cement production. The precalcining technology is that made the raw material preheated and decomposed outside the preheater and calciner to reduce the heat load of firing zone of the rotary kiln and greatly improve the production efficiency of the kiln system. Therefore, the calcination rate of preheater and its variation become the one of the important target for controlling the calciner and kiln.Currently, There is no measurement method to detect the calcination rate of preheater, the calcination rate of calciner can be detected off-line by sampling and chemical analysis, but it has great lag relate to real-time control, So the reference value on controlling the calciner and kiln is decreasing. The soft sensor of the calcination rate of preheater can obtaine the situation of preheating in advance and provide the necessary information for calciner and kiln in time. Therefore, the achievement of the calcination rate of preheater has a great significance.Cement production enterprises has no specific measure to detect the calcination rate of preheater currently, it causes that cannot use historical data directly to build the model for predicting the preheater decomposition rate, also lack the corresponding forecast inspection way. the calcination rate of preheater and calciner has the same variation through in-depth analysis of the working mechanism, and mainly depends on its internal temperature and pressure, more importantly, the calcination rate of calciner can be obtained through manual inspection. This paper build the relationship-rules of the calcination rate of calciner, temperature and pressure by analysis data and experience combined with on-site workers which can category the corresponding rules about preheater. According to the rules of the preheater and forecasting model, we can achieve the soft sensor of the calcination rate of preheater calciner.This paper builds the model of several important parameters including the temperature and pressure in the level of C3, C4 preheater combine with the production process characteristics of preheating systems in NSP. Because the preheater is strong coupling, great delay, nonlinear, multi-factors and time-varying, it is difficult to establish exact mathematical model. it use the method of more input and one output to build the modeling of prediction variables one by one, From simple to complex modeling principles, This paper models the temperature and pressure basing on Least Squares Support Vector Machine (LS-SVM) and BP neural network which select the auxiliary variables by correlation analysis and preprocess the original data by Mean Minimum Distance(MMD), The results of the comparison of two experiments show that the model basing on LS-SVM has favorable learning ability, generalization performance, and satisfactory prediction accuracy. Therefore, we choose the LS-SVM to build the model of the calcination rate of preheater.Analysis the mechanism of raw material, we establish the variation relational table of the calcination rate of calciner with temperature and pressure, combined historical data of t the calcination rate of calciner by off-line detection and the experience of the field operations,and This table also applies to the preheater. We can achieve the soft sensor of the calcination rate of preheater combine with this table and the parameters predicted.In this paper, we take ABB's Freelance 2000 DCS system as an example, and gives NSP cement preheater decomposition rate of soft measurement scheme basing on the OPC interface and SQL Server with database technology, Using vc++6.0 to develop the soft measurement interface of the calcination rate of preheater, and have been applied to the NSP cement production site, obtaining a good results.