| The deposition and slagging of coal burning boiler are always the problems that power industry faces and have not been solved perfectly yet, and with the magnification of boiler capacity the problems become even worse and will influence security and economic operation of station boiler. So, predicting and determining the state of deposition and slagging to take action accordingly is very important to prevent boilers from heavy slagging and deposition and optimize operation.These years, many experts and scholars at home and abroad have studied the characteristics of slagging deeply and presented lots of predicting methods, however, most of their accuracy are not satisfied. For this reason, this thesis begins with analyzing the reasons that lead to lowering of prediction veracity and contributes an open comment on research findings and new advances both at home and abroad for nearly 20 years. Among them, nearly 20 single norms and their bounds defined are concluded, which are divided two kinds: static and dynamic, thus will benefit the unprofessional to get to know them and benefit the professional to master them and explore new norms according to the architecture. At the same time, this paper has also analyzed merits and defects of these methods existing and presented advice and suggestions. Then, two dynamic norms are applied in the Fuzzing-Mathematics Model, Pattern- Recognition Model, Fuzzy-Neural-Network Model, through experiment it proves that these models are higher in accuracy than the traditional ones accordingly.The traditional fuzzy sets only touch upon positive membership grade but the practical ones often relate to not only the positive but also the negative and sometimes hesitation between them is mentioned. For this reason, Vague sets is applied in predicting model. Through experiment, it proves that this method excels in accuracy. Because the usual complex methods are all single model, multilevel fuzzy model are presented to improve accuracy further and avoid misoperation.Delphi is called the fourth-generation computer language, which is simple, highly active, powerful function, etc and is popular in programmers. So, the predicting application software in this paper is developed by the advanced Object-Oriented Program language. What's more, database-digging and CPU technology is widely used to develop database management system.Besides, deposition in tail heating surface has also been studied: acquiring elementary regularities of distribution between coal powder and fly ash; using microwave to measure carbon-in-ash online in the duct etc. |
