基于数值天气预报的母线负荷综合预测模型

Integrated Forecasting Model of Bus Load Based on Numerical Weather Prediction

母线负荷预测是电网安全校核和编制发电调度计划的基础。基于母线负荷历史数据和数值气象数据,通过相关性分析将母线负荷进行分类,提出了一种基于数值天气预报的BP神经网络母线负荷预测方法。进而以倍比平滑法和基于数值气象数据的BP神经网络法为基础,建立了分时段变权重的综合预测模型。对烟台地区的母线负荷预测结果表明,所提出的母线负荷综合预测模型有效地提高了预测准确率,对含大规模风电接入电网的输电设备安全校核具有重要的实用价值。

Bus load forecasting is the basis for power grid security checking and for power generation dispatch plan. Based on the historical data of the bus load as well as numerical weather data, bus load was classified through correlation analysis. A BP neural network bus load prediction method was proposed on the basis of numerical weather prediction. Furthermore, a composite prediction model with time-variable weight was set up in the multiple proportions smoothing method as well as the BP neural network approach based on numerical weather data. The results of prediction of bus load in Yantai area indicated that the proposed composite bus load prediction model could greatly improve prediction accuracy and have great practical value in security check for the transmission equipment in the power grid containing large- scale wind power integration.