摘要
结合鞍山市燃气日负荷预测,论述了历史燃气负荷数据的处理方法。运用GCAQBP神经网络对鞍山市2002年燃气日负荷进行了预测,预测的平均相对误差为4.1%。研究表明,这种预测方法可行。
Based on the forecast of daily gas load in Anshan City, the method for processing history data of gas load is discussed. The daily gas load in 2002 in Anshan City is forecasted by GCAQBP neural network with average relative error of 4.1%. The study shows that this forecast method is feasible.
出处
《煤气与热力》
2006年第10期16-19,共4页
Gas & Heat
