The Research Of Fault Diagnosis Method Based On BP Neural Network For Municipal Solid Waste Incineration

The elementary technologies of Municipal Solid Waste (MSW) Incineration and the characteristic of the neural network are described briefly in this thesis. Based on the characteristic of the three typical faults diagnoses for MSW Incineration, the modeling variables are selected from the operating data of the Incineration systems, which are employed to the model training and test for the BP Neural Network algorithm.Many BP neural network algorithms are applied to the three typical fault statuses respectively. According to the effect of training and test, the optimum optimum BP neural network algorithm is obtained to train the fault diagnosis model. Although this thesis is focus on the small vertical incinerator, the methodology of fault diagnosis is easy to apply to other type incinerators. Based on this work, it is possible to establish a more generalized system for early fault diagnosis in MSW incineration, which could avoid further sevious operaton failure, personal and equipment losts.