Fault Diagnosis And Process Control Of Small Vertical Municipal Solid Waste Incinerator

Incineration is an important method for Municipal solid waste (MSW) treatment. At present, fault diagnosis and process control of small MSW incinerator in China is mainly rely on operator's experience. Fault diagnosis and trouble shooting can not be realized automatically in most facilities. The research on fault diagnosis and process control will be helpful for improving the performance of incineration process..This thesis is focus on the small vertical incinerator. MSW incinerator process fault diagnosis models and process control model are established, which are based on expert system and artificial neural network by analysis of the real MSW incinerator data and operation record. Fault tree analysis is used to obtain knowledge for expert system. Optimal stopping rule and neural network ensemble are used for improving generalization ability.The test on real-time data shows that the accuracy of expert system fault diagnosis model is higher than 87.12%. The accuracy of artificial neural network fault diagnosis model is higher than 92.35%. The accuracy of artificial neural network process control model is higher than 92.39%. And optimal stopping rule and neural network ensemble can improve thegeneralization ability remarkably.Three models are programmed by Matlab software for achieving on-line fault diagnosis and process control of MSW incinerator.