Application And Research Of Wind Field Intelligent Forecast Management System

The high-speed operation of the fan in the wind farm will cause frictional heat generation,resulting in excessive temperature of parts,causing fan damage and serious economic losses,and even causing fire.The traditional automated wind farm solution focuses on the numerical monitoring of sensor temperature,and the manpower to determine the danger of the wind turbine components is time-consuming,and it is impossible to make protection measures for the wind turbine parts in time,that is,when the abnormal temperature of the sensor is detected,it has caused a certain degree of damage to the fan parts.In order to accurately predict the operating state of the wind turbine and reduce the damage degree of the wind turbine parts,this project designs a reverse propagation neural network temperature prediction model based on genetic algorithm optimization through the study of machine learning and virtual reality technology,and combines the three-dimensional visualization technology to complete the design of the wind farm intelligent prediction management system,so as to realize the real-time status visualization of the wind turbine,historical data visualization,and Warning information visualization and other functions achieve the purpose of real-time monitoring of the temperature of fan parts and early warning of overtemperature of fan parts.Results achieved include:Firstly,in the research of intelligent prediction technology,the multiple linear regression temperature prediction model and the single BP neural network temperature prediction model are selected,due to the unpredictability of the dependent variable of the multiple linear regression temperature prediction model and the grid model learning ability of the single BP neural network is not guaranteed.In this paper,a BP neural network based on genetic algorithm optimization is proposed,and the genetic algorithm(GA)determines the most suitable individual through selection,crossover and mutation operations,so as to optimize the initial weight and threshold of the BP neural network for better sample prediction.Experimental results show that the GA-BP neural network model has a better fitting effect on data and has strong generalization ability.Secondly,in view of the fact that traditional data visualization can only display simple sensor data in a table,and cannot establish a more intuitive connection with the real wind turbine,this paper chooses to use Unity3 D engine,combined with modeling technology and three-dimensional visualization technology to describe and manage the data information of the real wind farm,enhance the realism of the virtual wind farm,and completely solve the problems of traditional visualization.Finally,based on the GA-BP neural network model,combined with 3D visualization technology,this paper designs and develops the intelligent prediction management system of the wind farm by using the Unity3 D development engine,which realizes the functions of real-time status of the wind turbine,warning information and statistical report visualization,so as to achieve the purpose of real-time monitoring of the health status of the wind turbine and predicting the overtemperature risk of the wind turbine components in advance.