Electric Power Forecast Early Warning Method Study

China has more frequent disasters, among them, in the more common natural disasters,about 70% of the disaster comes from meteorological reasons, so that the national economy is not a small loss every year. The weather conditions will be affected by the state power grid when power load at its peak, its peak and the duration will depend on significant changes in the weather, the power grid, the meteorological disasters seriously affect its security.In January 2008, southern China appeared in the network paralysis and larger scale,the causes of disasters is the low-temperature freezing rain and snow, resulting in a great economic loss, totaling up to 151.6 billion. In the past 20 years, due to natural disasters affect grid stability events occur frequently, China Electric Power Research Institute this statistics,the results show that, the equipment fault is the primary factor affecting the stability of power grid,the natural disaster is the second factor. Therefore,with the development of society and economy, the meteorological disasters in various regions have their own unique characteristics. For the power system, it is urgent to analyze the local meteorological disasters.Improve the ability of disaster prevention and mitigation is placed in front of the power system, the most important issue, to do a good job in the basic work, in a certain extent will can reduce the loss caused by meteorological disasters, but also to prevent the role of meteorological disasters and for disaster prevention and mitigation.I worked in Heilongjiang province meteorological service, technical support, chief,undergraduate computer science and technology. Specific units responsible for project management, project construction, system maintenance, information network construction, etc.Hope that through this study, the ability to power, weather, combining the two industries,open up a new train of thought, professional meteorological service for the meteorological forecast warning methods effective for electric power industry, and application in the specific business.By visiting the Heilongjiang electric power sector research, the power sector to clear the demand for meteorological forecasting and early warning services. Should be fully integrated system application of information and communication technology, computer technology and other high technology, of meteorological disasters were real-time monitoring and early warning assessment, for the electrical engineering departments to provide timely, accurate and reliable information, disaster prevention and mitigation have sufficient scientific basis for early warning. At the same time, in the research process, it is found that the most difficult point of the power weather forecast and early warning is the determination of the early warning threshold.Through in-depth study of neural network algorithm, mainly for meteorological data and power data combining optimization method is proposed. And this algorithm is the core of achieve power weather forecasting and early weather warning system design, applied to the actual business. Finally, the experimental results show that the optimized algorithm has a great improvement in efficiency and accuracy, and the power weather forecast and early warning system is demonstrated in detail.