摘要
采用VC语言编写基于神经网络技术的城市燃气短期负荷预测模型,经实例验证可以较精确地预测出城市燃气短期负荷。预测模型在权值修正项中引入动量项以加速收敛,在数据输入时引入噪声,以提高网络的泛化推广能力。
A mathematical model for short-term city gas load forecast based on neural network programmed with Visual C++ has been established. It is verified experimentally that short-term city gas load can be forecasted accurately. The momentum term is introduced into weight correction term of the forecast model to accelerate convergence, and noise is introduced when inputting data to enhance the network ability for wide application.
出处
《煤气与热力》
2005年第12期10-14,共5页
Gas & Heat
基金
科技部十五科技攻关计划"小城镇基础设施建设关键技术"的子课题"小城镇燃气供给系统建设关键技术研究"(子课题编号:2003BA818A15-3-1)