摘要
空调在长期使用过程中,负荷值会随时间量的增加不断提升,在此过程中,受到时间变量与负荷噪声的影响,生成长短期特征不同的波动,在波动作用下负荷特征个体对应的预测网络细胞预测精准度会有所降低,进而增大空调负荷预测记忆网络的预测误差。导致预测方法输出预测量误差同步增大。为了解决误差问题,提升预测值准确度,利用对长短期记忆网络加以改进,并将改进后的长短期记忆网络应用于空调负荷的预测计算,通过对空调负荷的加权灰色关联计算、基于改进后长短期记忆网络预测量确定与空调负荷预测结果输出,完成对空调负荷预测精度优化的计算过程。通过与两种预测模型的数据对比表明:提出预测方法在负荷预测值准确度提升方面具有确切效果,且稳定性较好,达到了提出方法的预期效果。
Air conditioning in the process of long-term use, load value will increase with the increase of time, in the process,under the influence of time variables and load noise, generate long and short-term characteristics of different fluctuations, under the action of fluctuation load characteristics of individual prediction network cell prediction accuracy will be reduced, and increase the air conditioning load prediction memory network prediction error.Increase of the output premeasurement error of the prediction method synchronously.In order to solve the error problem, improve the prediction accuracy, use the short-term memory network to improve, and apply the improved long-short-term memory network to the prediction of air conditioning load, based on the improved short-term memory network prediction and air conditioning load prediction output, complete the calculation process of air conditioning load prediction accuracy optimization.By comparing with the data of the two prediction models, it is shown that the proposed prediction method has an accurate effect in improving the accurate load prediction value, and has a good stability, which achieves the expected effect of the proposed method.
作者
孙春艳
Sun Chunyan(Pingdingshan Branch of China Coal Science and Industry Group Beijing Huayu Engineering Co.,Ltd,Pingdingshan,467000)
出处
《制冷与空调（四川）》
2022年第6期870-875,共6页
Refrigeration and Air Conditioning
关键词
长短期记忆网络
空调
负荷
预测方法
long and short-term memory network
air conditioning
load
forecast method
