关键词: 能耗预测, 阳极氧化, 成膜系数

Abstract:

The energy consumption calculation of aluminum production surface oxidation is very complicated due to many influencing factors,such as multi-type product and inconsistent calculation standard (size,weight,and thickness),etc.In order to solve these problems,this paper proposes an aluminum oxide energy consumption computing method based on neural network.We first train the neural network by using historical data and calculate the influence of the oxidation factor on the coefficient of oxidation.Then the coefficients of oxidation of the batches of oxidation and production energy consumption are calculated according to the trained neural network.Finally,we verify the effectiveness of proposed method by comparing the calculated production energy consumption with the measured data.The experimental result shows that our method can find abnormal energy loss,improper production parameters in the production.

Key words: energy consumption prediction;anodic oxidation;oxidation factor