Research On The Application And Improvement Of The Data Mining In The Control Of Power Generation Equipment Reliability

This paper is concentrated on the research about the application and improvement of the data mining technology in the control of power generation equipment reliability. This paper firstly provides an overview of research background, content, innovation points and the related research achievements of China and abroad. Then, this paper analyses deeply the current common theory and model of research of power generation equipment reliability, and tries to make the method of data mining technology use in such research. For the realization of the application of this method, we design application process of data mining technology in power generation equipment reliability control and get clear and definite application goal. In addition, this paper chooses back propagation neural network as the main research method from a variety of data mining technologies, which combined with the characteristics of power generation equipment reliability identification. We use the method of multidimensional screening to screen index from many reliability index and construct the index system of this paper according to the results of screening index.For the further improvement on the limitations of the traditional back propagation neural network technology, this paper gives up the use of the small random weight given by network system and tries to confirm the initial weight of each key indicator by using the method of analytic hierarchy process, which can make the network’s initial weight fit the actual weight maximumly. By this way, on the one hand we can avoid the blindness of network training, on the other hand, we can make network training have directivity. Then we can ensure the stability of the system at the same time of reducing the learning time by setting a smaller value to learn factor. After the improvement of neural network, this paper puts forward the division standard of power generation equipment reliability level and designs the realization process of the improved data mining technology in power generation equipment reliability control. Finally, we demonstrate the usability and superiority, high efficiency of the improved data mining technology in power generation equipment reliability control.