Study On Intelligent And Network-based Technology In Steam Turbine-Generator Group Online Monitor And Diagnosis

The intelligent, network-based and automatic technology of device status on-line monitoring and diagnosis for status of devices was researched. Based on the steam turbine, boiler and main accessorial equipment of thermal power plant, the network and intelligence of the on-line monitoring and diagnosis system were studied. The concretely works were listed as follow: 1) With the knowledge of intelligent system, various expressive methods and characteristics of the intelligent system were analyzed and comprehensively the main learning method of the knowledge of intelligent system and its algorithm was discussed. Based on the expressive methods, the expressive modes based on the network topological structures and the communal knowledge framework which is fit to combine the intelligent reasoning were produced so as to visually and flexibly express and increase the applied efficiency of knowledge base; By studying on the learning method and algorithm of knowledge of intelligent system, we develop the learning method and algorithm to acquire mixed type knowledge which combine layer clustering with fluctuant structure. And we can acquire and study knowledge on-line or off-line, which is effective for graphic-network topological structure and combined reasoning system. 2) The reasoning system and algorithm of intelligent system were explored and analyzed systematically, then the reformative reasoning system and algorithm for the expert system, fuzzy logic reasoning and neural network reasoning were introduced. The parallel combined reasoning, optimizing decision and algorithm can make fully use of the communal knowledge based on the graphic-network topological structure for reasoning and decision. The parallel combined intelligent reasoning can be closer to human thinking model and improve the intelligent level of device status on-line monitoring and diagnosis. 3) The network-based structural system of device status on-line monitoring and diagnosis were studied. The method of device status on-line monitoring and diagnosis was produced by means of distributing Web Service network system, which can effectively achieve resource sharing, information issuance and long-distance transmission. So a new approach was provided for intelligent and network-based device status on-line monitoring and diagnosis. 4) A system was developed for network-based and intelligent on-line monitoring and diagnosis of steam turbine-generator group. The system is successful because it not only breaks through the mode for specially equipment status monitoring and diagnosis, but also explores an effective approach of monitoring and diagnosis for multiparameter complex system. So it can be helpful to improve the level of network-based, intelligent and automatic device status on-line monitoring and diagnosis. The innovations in this dissertation were: 1) The communal knowledge was merged from the expression, utilization, acquisition and learning of knowledge based on graphic-network topological structure. And the characteristics were agility, intuition and easy modification; so it enlarged the applicable range of knowledge base and improved service efficiency. 2) An intelligent reasoning strategy was put forward, in which expert system applied inaccurate reasoning based on cases, fuzzy logic applied deduction reasoning based on regulation, and neural network applied no-monotonously reasoning based on model. The three kinds of reasoning can parallel combine manually or automatically, using communal knowledge base. Meanwhile, the result of reasoning can be optimized, which can be closer to human thinking mode and improve the intelligent level of reasoning. 3) Using Web Service network technology resource sharing, long-distance transmission and monitoring and diagnosis can be achieved, which is accurate, effective and brevity for intelligent and network-based information management, information sharing, information issuance and long-distance transmission of the device status on-line monitoring and diagnosis. 4) The network-based and intelligent on-line monitoring and diagnosis system was developed by applying technology of network-based, intelligent and automatic of device status on-line monitoring and diagnosis to multiparameter complex turboset, which achieved successfully the running status monitoring and diagnosis of steam turbine, boiler and main accessorial equipment, So an effective approach was explored for multiparameter complex monitoring and diagnosis.