Research On Fault Early Warning Method Of Wind Turbine Driven By Big Data

As a clean energy,wind energy plays an increasingly important role in optimizing and improving the energy structure of our country and even the world.In recent years,the scale of wind power generation is becoming larger and larger,and the technology of wind power generation is becoming more and more mature.The requirements for the operation and maintenance of large-scale wind turbines are also getting higher and higher.But because wind turbines are usually located in the Gobi Beach,the environment is very bad,and they are always facing the risk of damage or even destruction.The cost of wind turbines is high and maintenance is difficult.It is particularly important to monitor and warn the faults of wind turbines in real time.It is expected that the wind turbines will be able to predict their damage trend in time before serious faults occur,and the operation of wind turbines will be adjusted accordingly.Avoid further expansion of damage and even the risk of wind turbine destruction.Scholars at home and abroad have done some research on the fault diagnosis and early warning of wind turbines,providing a model and scheme based on a small amount of data and non-real-time calculation in laboratory environment for the fault diagnosis and early warning of wind turbines.However,with the deepening integration of informationization and industrialization,all kinds of sensor technologies are becoming more and more mature,the amount of data needed to be calculated is increasing,and the real-time requirement of calculation is getting higher and higher.The single-machine serial computing mode and the fault early warning model obtained in the single-machine environment can not meet the requirements of large data and fast processing.Increasingly mature and abundant Internet of Things technology,big data technology and artificial intelligence technology provide new means and methods for the real-time summary,distributed storage,intelligent calculation,Apriori early warning,monitoring of operation parameters and fault prediction of wind farm fan units.In this context,this paper mainly carries out the following research:(1)Taking the wind turbine gearbox as an example,the fault forms and equipment of the wind turbine gearbox are analyzed.The cause of volume formation and the characteristic parameters used to distinguish the fault form are reasonably extracted.(2)A unified wind turbine structure coding and fault coding rule is proposed,and the rules are stipulated.Standard data format.(3)A wind turbine fault early warning architecture based on big data technology is proposed and designed.The improved ensuing forest algorithm is implemented in a distributed and parallel way based on Spark platform.(4)Based on the improved Successive Forest algorithm,a gear fault early warning model is established.The evaluation method system of the computational model is proposed.Finally,the performance of the algorithm is tested on the large data platform.The experimental results show that the improved random forest classifier has high accuracy,good acceleration ratio in parallel environment and good comprehensive calculation performance.(5)The software architecture of large data-driven wind turbine fault monitoring and warning platform is proposed.The functional modules of the platform are designed in detail.The workflow of the platform is described in detail.The software and hardware environment is built to develop and test the platform completely.It has been proved that the platform runs accurately and efficiently,and the real-time monitoring and fault warning of wind turbine gearbox gears have been successfully realized.Through this research,large-scale real-time acquisition,transmission and storage of wind turbine operation parameters are realized,and efficient early warning of wind turbine components failures is realized.The purpose of improving the real-time operation and maintenance of wind turbines,reducing the cost of wind turbine operation and maintenance is achieved,and the efficiency and accuracy of wind turbine failure operation and maintenance are greatly improved.