Study And Application Of Intelligent Fault Diagnosis In Thermal Power Plants

In order to meet with the diagnosis requirements of the complex system, the monitor technology turn into the intelligent fault diagnosis phase, which is focus on knowledge disposure, combining with signal disposure, the disposure of model construction and the knowledge disposure, with the speedy development of computer technology and artificial intelligence. Intelligent fault diagnosis provides a new way for enhanced the reliability and the security of thermal power plants.Firstly, the thesis totally introduces the necessity of thermal power plants to carry on the intelligent fault diagnosis in chapter 1. Then the source and background of the task were briefly introduced, moreover the paper further summarize the research of diagnosis technology in both domestic and overseas. Finally, the mostly works of this paper were simply presentation.In chapter 2, the author particularly introduces the study methods about intelligent fault diagnosis, discusses the further progress trends and the common structure of the intelligent fault diagnosis system, and also introduces the current popular methods used in intelligent fault diagnosis system.In chapter 3, there is a detailed introduction of two related softwares platforms, one is the DCS simulation system of XiaGuan thermal power plant; the other is INTEMOR which is the platform of intelligent fault diagnoses.In chapter 4, based on the DCS simulation system of XiaGuan thermal power plant intelligent fault diagnosis system, on the platform of INTEMOR, the paper particularly introduce the software structure of system , and emphasize the construction of knowledge base using expert system and the research & development of on-line prediction module about one equipment.In chapter 5, the paper introduces the supplement of INTEMOR in detail in order to solve the problems meet in the project. Finally uses the conjugate gradient algorithm which is the improvement of BP net netwok to realize the condenser fault diagnosis, the simulation result indicats that, this method is feasible.In final chapter, the author sums up the full paper and and put forward the further revision proposal to the system.