Prediction method of reheater temperature deviation based on attention mechanism spatiotemporal deep neural network
Boiler reheater is a steam superheater that reheats low-pressure steam to a certain temperature.Its stable operation is of great significance to the safe and efficient operation of coal-fired units.During the operation of the boiler,due to the residual rotational momentum in the furnace outlet and the horizontal flue,the outlet temperature on both sides of the reheater is deviated,and if the deviation is too large,a tube burst accident may occur.In order to realize the advance prediction of the temperature deviation of the reheater,this paper designs a spatiotemporal fusion deep neural network model based on the attention mechanism.The convolutional neural network and the gated recurrent neural network are used to extract the spatial and temporal information,and at the same time,the attention is introduced.The mechanism assigns weights to various features to improve prediction accuracy.The model also takes advantage of the convolutional neural network in spatial information processing and the gated recurrent neural network in temporal information processing,and makes full use of the global information of the channel attention mechanism on the feature map,solving the problem of single model feature information extraction.sufficient question.The application results on real data of coal-fired power plants show that the proposed method has higher prediction accuracy compared with other deep learning methods,the root mean square error,coefficient of determination,and prediction accuracy were 0.962,1.342 and 0.985,respectively.
reheater temperature deviation predictionconvolutional neural networkgated recurrent neural networkattention mechanism
NETL
NSTL
万方数据
维普
