Boiler Intelligent Fault Diagnosis System

"Four pipe" fracture is the main problem in the boiler operation which affect the safe operation of the boiler equipments Intelligent fault diagnosis plays an important role in power plant operation as to how to timely find and prevent the"four" burst.. This paper introduced two applications of intelligent fault diagnosis in the boiler, which are multi-dimensional BP neural network in fault diagnosis of the boiler and the combination of data mining and BP Neural Network Fault Diagnosis. The applications of BP neural network is the way of fault diagnosis modeling targeted at the feature of thermal characteristics of the boiler with the adoption of multi-dimensional method of BP neural network. The input layer used the fuzzy mathematic method to quantify the operating parameters and then established the Multi-dimensional structure of BP neural network through the relations between parameters and patterns of the fault. Taking the "Four pipe" fracture for instance, the BP neural network model was built and the results of the simulation experiment proved that this method can give a rapid, efficient result with high accuracy. Combining the advantage of data mining and BP Neural Network Fault Diagnosis in Boiler, integrating the rule mining and BP neural Network of organically, the intelligent fault diagnosis system was established which can demonstrate a more exact process of the human thought, and a large number of fuzzy and random data was also extracted from the history of the operation so as to make sure that the aimed data can be input the BP network effectively. Experimental results showed that the two method have their own merits respectively and adopting which depends on the features of the boiler. So an optimum method should be take while modeling according to the characteristic of the boiler.