Research On Improving The Accuracy Of Soft Sensor For Kappa Number Estimation In Pulp Cooking Process

Cooking process is an important stage in pulp and papermaking industrial. It isalso a very complicated physical and chemical process. Kappa number is the mostimportant quality index of cooking process. Good control of Kappa number is the keyto stabilize the quality of paper pulp. The steady Kappa number is also helpful todecrease the consumption of stream and chemical products, to decrease theenvironment pollution and enhance production efficiency. In order to control theKappa number of pulp, it must be measured or estimated online. But until now theKappa number online measurement instrument, which is precision, dependable, cheapand easy-to-maintain has not been developed throughout inland and overseas.Therefore, it is significant in theory and application to develop soft sensingtechnology of Kappa number in cooking process. In the former research, we havedone a deeply work on the soft sensing field of Kappa number. The results we gothave achieved certain efficiency in real applications, but farther research is needed toimprove the prediction precision.This dissertation began the research work from the shortcomins of formerresearchs or the angle that is never concerned. By adding process information, usingsimple and effective data preprocessing method, analyzing temperature-rising curvesin complexy production condition, using new modeling method with goodgeneralization ability and building a hybrid model with better performance, some newmethods were put forward to improve prediction precision of Kappa number. In orderto reach the goal, different kinds of theorys and methods should be integrated to digout useful information from the original data.This dissertation concentrated on the research work listed below and achievedsome creative results:1) Based on the technical analysis and the condition of actual product process of thebatch pulp cooking, the dissertation points out the limits of single model for the wholecooking process, since the process of delignification is linearization for differentphase. A new subsection model is presented based on the simplified Hatton model.2) After analyzing the composing of prediction error of soft sensing model, a methodof abnormal data discovery for data processing of Kappa number soft sensing ispresented. The new data processing method digs out incompatible data based on dataclustering and mechanism analysis, as well as finds out the outlier data by regressionanalyzing and statistical analysis. It also can explain the impact of abnormal data on...