摘要
针对神经网络在非数学模型预测中所面临的3个主要问题,提出了一种基于BP_Adaboost算法的预测模型对燃气负荷进行短期预测。预测结果表明,该模型与BP神经网络相比,不但提高了预测精度和泛化能力,而且更能满足具有非线性、时变性和不确定性的负荷预测的需要,具有较好的应用前景。
Aiming at the three main problems faced by BP Neural Network in non- mathematical model forecasting, the new forecasting model based on BP_Adaboost was put forward and forecasted the gas short-term load.Prediction results show that not only prediction accuracy and generalization ability were ？Improved but also more satisfied the request of the load forecasting with nonlinear, time variation and uncertainty compared to BP neural network. Which has a good application prospect.
出处
《科技通报》
北大核心
2013年第10期55-57,共3页
Bulletin of Science and Technology
基金
河南省教育厅高等学校青年骨干教师资助计划基金(2011GGJS-252)
