Power Equipment Fault Analysis And Diagnosi Based On Big Data

With the continuous expansion of the power system and the increasing complexity of equipment,new challenges have been brought to the safe operation of the power grid.How to improve the reliability of the equipment and ensure the security and stability of the power system has gradually become an important research topic in the power industry.Besides the use of high quality and stable equipment,the operation reliability of equipment can also be achieved by continuously improving the level of equipment health assessment and equipment fault diagnosis.At the same time,because of the complexity of power system,power system data often has a vast amount.It is difficult to get valuable information from the data,so this paper combined health state evaluation and the diagnosis of equipment fault with big data,deep learning and other related content,and the specific research contents are as follows.First,the current situation and development process of data asset management are studied,and the status of asset management in power enterprises.Then,based on the large data and all life cycle management theory,the whole life cycle management process and the overall technical framework of the power equipment assets are proposed.Secondly,based on the large data related technology combined with the operation data of pumped storage power station,the application of the whole life cycle management of the equipment is studied and analyzed.The research mainly focuses on the operation strength,reliability evaluation and equipment health status prediction of pumped storage units,and based on time series analysis,an indirect prediction model of equipment health status is built.Then,the commonly used model methods in the field of power equipment fault diagnosis are studied,and the related contents of deep learning and the application of deep neural network in fault diagnosis are summarized.Then based on the characteristics of deep learning network and equipment fault information,a fault diagnosis model based on LSTM bi-directional recurrent neural network is constructed.Finally,a bi directional recurrent neural network fault diagnosis model based on LSTM is achieved based on TensorFlow,and the experiment is carried out on the basis of the fracture data of wind turbine generator's toothed belt.The results show that the model constructed in this paper can be effectively diagnosed and has high accuracy.