基于改进灰色模型的电力系统负荷预测

Power System Load Forecasting Based on Improvement of Grey Model

传统的灰色预测模型因其所需历史数据少、计算快、对平稳地区的负荷预测有较高精度等优点,曾被广泛应用。但传统的灰色预测模型对于历史数据要求较高,最好为指数形式,并且在数据波动较大的情况下,其预测误差可能变得较大,不符合实际需要。为了减小预测误差,本文在传统灰色模型的基础上,首先对部分历史数据进行平滑处理,以确保其光滑性,同时对历史数据进行等维处理,不断的剔除旧数据,增加新数据,最后进行灰色循环残差修正,在原始数据和预测模型两个方面进行了修正,提高了电力负荷预测精度。

The general gray model needs a fewer history data, calculates faster and makes high precision to load forecast of areas with stable load. So it is widely used. The general grey forecasting model demands higher quality of historical data, the best form is the index. It has the distinct deviation under the circumstances of considerable data fluctuations. This is against the actual condition. In order to reduce forecasting errors, Based on traditional gray model in this paper, historical data is smoothed to ensure smoothness, processes with equal dimension which continuously removes the old data and adds new data. We establish an RRGM（1,1） model and make a revise precision. in the raw data and prediction model with high forecasting precision.