Research And Development On The Wind Power Prediction

China's wind power has entered the rapid development. Wind power prediction is a very effective way to increase the wind power penetration and improve the security and economy of the power system.This paper analyzes the research situation of the wind power prediction technology, the prediction method and basic principle; On this basis, the design of the wind power prediction system framework, established the wind power prediction system. The system is based on numerical weather forecast, fully considered the air humidity, temperature and pressure on wind farm output power influence; Based on the error Back Propagation algorithm of multilayer feed-forward networks (BP Neural Network) for wind power are predicted.The system uses layered structure, which is beneficial to the system function expansion and maintenance; the forecast system software according to different need to adopt windows operating system platform and Unix/Linux operating system platform, has the good human-machine interface, and has realized seamless connection with Energy Management System (EMS).The system has been applied in one provincial power grid dispatch center, the results of the test data indicate that the prediction system reliable work, the error of the prediction result meet the engineering requirements.