摘要
数据中心温度预测在提高能量利用效率、降低机房空调耗电量方面具有重要意义。RBF神经网络广泛用于温度预测领域,考虑到传统RBF神经网络无法对影响温度的众多因子进行确定和选择,本文提出了一种将云模型和RBF神经网络结合的新模型。通过高维云变换确定RBF隐含层神经元,使RBF神经网络充分表达影响温度各因子的不确定性,进一步优化RBF神经网络结构。实验结果表明,该模型能较好地实现对数据中心温度的预测。
It is important that data center temperature prediction is used in improving the energy efficiency,reducing computer room air-conditioning power consumption,RBF neural network is widely used in prediction of temperature field.Considering the traditional RBF neural and network does not affect the temperature of many factors identified and selective,a cloud model and RBF neural network combined with the new model is presented.Through Gao-wei-yun RBF transformation hidden layer neurons is determined.The RBF neural network makes neural network fully express the influence of temperature factor uncertainty,and further optimize the structure of RBF neural network.Simulation results show that the algorithm can better realize data center temperature prediction.
出处
《沈阳理工大学学报》
CAS
2013年第4期9-14,共6页
Journal of Shenyang Ligong University
关键词
数据中心
云模型
云变换
RBF神经网络
温度预测
data center
cloud model
cloud transform
RBF-neural network
temperature prediction
