Study On The Life Evaluation And Operation Analysis Expert System For Boilers' High Temperature Bearing Elements

With the development of science and technology, the equipment becomes more and more complex. The investment of the maintenance to the equipment also increases much more. It's necessary to push the advanced Condition Based Maintenance (CBM) system. Aiming at the superheater and reheater of the boilers in the power plant, this paper emphatically researches the life expectancy technology which relates to CBM closely. Using the result of life expectancy, and adopting the achievements of the fault diagnosis technology, it develops an expert system for operation analysis of the superheater. This expert system can bring forward reasonable opinions to the problems in the operation process of the superheater. Elevated temperature creep is the main factor of life-expense of the boiler抯 superheater and reheater. In this paper, time fractional methodology based on formula of Larson-Miller is discussed primarily. Then it introduces calculation method of ultimate creep strain based on 0 projection, idea of equivalent operation time and method of evaluation on creep strain. At last, it brings forward time fractional methodology based on 0 projection. Because insider and outside corruption and abrasion lead to decrease of pipe thickness, pipe thickness must endure larger stress, which needs to be included in calculation. At present, there is no effective means to measure the temperature of the pipe inside the furnace. We consider that there is a complicated non-linear function between highest temperature of each surperheater or reheater pipe in inner furnace and temperature of pipe and steam in the entrance and exit, and steam flux. The idea that B-P NN is applied to approach this function has been incarnated in this paper. Improve on arithmetic for NN has been provided in this paper, which can heighten velocity of constringency. This paper considers that operation personnel should be involved in the maintenance work. Thus, it designs an expert system for operation analyses that aims at the security of the superheater, using the idea of fault diagnosis expert system. The process of inference adopts two methods that are fuzzy rule inference and fuzzy inference that bases on fuzzy relation matrix. This paper sums up the problems in operation and maintenance of the superheater. In view of the life expenditure, It brings forward fuzzy expression method, which inflects the harm to the superheater more accurately. Some fuzzy rules and fuzzy relation matrixes is set up, which can be used for operation analysis and fault diagnosis for superheater and composes the primary content of the expert system repository. Expert system and neural network are two fundamental approaches applied in the field of artificial intelligence, but they both have obvious predominance as well as its deficiency. The Artificial Neural Network Expert System (ANNES), a combination of neural network and expert system, can utilize their respective advantages, to overcome disadvantages of the other side. This paper introduces the two primary strategies to integrate expert system and neural network, puts emphasis on the knowledge base and inference mechanism of ANNES. and finally. it comes up with a method of converting traditional expert system into ANNES. Operation analysis and fault diagnoses are based on mass reliable real-time data. This paper, INFI-90 as a example, discusses the approach of acquiring real-time data from distributed control system, and designs a distributed fault diagnosis data, based on computer network and adopting client...