Research On Power Grid Load Forecasting Of Xinyu City

One of the key indicators of electricity production and development of electric power load forecasting. No method could guarantee satisfactory results in any cases. Combination forecasting takes advantage of information from various forecasting models and get combination forecasting models in appropriate forms In order to enhance their fitting capability and improve the forecasting accuracy. his paper further studies the application principles and methods of ant colony algorithm and combination forecasting in power load forecasting on the basis of previous researches, and in which middle-term and long-term combination forecasting models of power load forecasting are constructed.Combination forecasting model, the selected single forecasting model weights and select a single prediction model is the most critical. During the construction process, the most important things are the screening of single forecasting model and the identification of its weight. This eight methods including gray prediction method, multiple linear regression,linear regression analysis, quadratic and cubic exponential smoothing method, the second moving average, electricity elasticity coefficient method, load density method. Was selected as the single prediction models according to the characters of power load forecasting. In order to obtain the weighting coefficient of combination forecasting model that also incorporates the relevant information in the past few years, the establishment of the objective function with the least squares criterion, so that the objective function is minimized in the constraint conditions; and then select the weighting factor based on a single prediction model, except where the weights are too small cast, so the model is more easily and accurately achieve qualitative and quantitative perfect unity, so that the model in line with scientific.The constructed objective function is a non-liner problem and there is no common method could solve non-liner problems. While ant colony algorithm mainly deals with discrete combination optimization problem, this paper improves this method and by which the author try to get weighting coefficients.Ant colony algorithm and the combination forecasting are adopted in this paper to ascertain the weighting coefficients which are difficult to ensure before and select single forecasting models in a more reasonable way. In this way, the power load combination forecasting would be more scientific. And the practical applicability and veracity of thestudy have been proved in the practice in Xinyu, which means the result of this study is worth spreading and applying to some extent.The process design including: the total plane civil heating ventilation and cooling water supply control power environmental protection design of process design according to the requirements of the design process and the proposed data design make whole design can do high efficiency energy saving and environmental protection, and to ensure food safety accord with a country and industry to the related standards.