The energy and environment problems are the key issues that affecting the nationaleconomy’s sustainable development. Our country attaches great importance to energysaving measures now. The improvement of the efficiency of boiler operation and thereduction of emissions is the two main tasks that the cogeneration enterprise faced.Making full use of large number of boiler operating data the Distribute Control System(DCS) recorded, using data mining technology to improve boiler efficiency has becomean inevitable trend.Data warehouse with the theme of boiler optimization needs to be established tomeet the requirements of vast amounts of data in data mining. Since boiler DCS systemhas various operating parameters archived, it is feasible to use the data structures whichWinCC can provides to design and implement the system using the archived data storedin the data warehouse.According to the complex nonlinear relationship between Boiler operatingparameters, this thesis adopted Artificial Neural Network theory and data miningalgorithms of Association Rules method to analysis the real-time operation data of theboiler and established boiler operation model. Combining Data Mining techniques withboiler operation, we established boiler operation model based on Artificial NeuralNetworks and Association Rules, developed boiler optimization data mining platformusing the Microsoft dot Net development technology.This system maximums the use of existing resources, improves boiler operationefficiency using data mining method without any hardware added. The designedmining platform supports many other methods included into the system, so that thiswork is helpful for further system optimization. |