On Line Monitoring And Diagnosing For Air Leak State Of Power Station Boiler

With the development of the thermal power station towards large capacity, high parameter, integration and the continuous improvement of data-gathered system in operation, it will be an important task of engineering practicality to build an energy-saving potential diagnosis system with the function of on-line monitoring, energy-loss diagnosis and operation guidance, which can find the main source of economy-loss in operation adjustment and the equipment drawback management, thereby instructs the running adjustment, renovation and improve the operation economics of the units. The existence and quantity of air leak affects the boiler's thermal efficiency greatly. So on-line monitoring and diagnosing for the boiler's air leak is an important item of energy-saving potential diagnosis in power station, there isn't an effective method to solve this problem and few researches are made on this task at the present time.As a part of the project of energy-saving potential diagnosis based on artificial intelligence in thermal power station, this paper elementarily studied the important factor which affects boiler's operation economy-loss-boiler's air leak. According to the boiler's practical condition, two feasible methods to determine the diagnosing characteristic parameters are brought out, one is based on the system's energy-distribution principle and the other is based on the system's energy-balance relation, on the basis of the above research, in order to carry out on-line monitoring and diagnosing for air leak state in power station boiler, the artificial neural network is adopted as the intelligence diagnosis tool, the air leak diagnosis scheme based on the principles of system energy-distribution and system energy-balance relation is presented firstly and then the different diagnosis models of artificial neural network are built respectively.Simulation tests shows that the diagnosis system can carry out on-line monitoring and diagnosing for the system's air leak by making use of the existing routine measure points and the operating parameters to be easily measured directly; because of the generalization characteristic and reasoning function of the artificial neural network, the diagnosis system can reach better diagnosis result even under the circumstance of insufficient information, obvious measuring error and yawp shape change; the above research result plays an important instruction meaning in improving operating economy of power station boiler.