垃圾焚烧炉蒸发量预测可行性研究

Feasibility Study on Evaporation Prediction of Waste Incinerator

本文建立基于梯度下降算法的4层神经网络,选取一次风、炉膛温度等14个参数作为输入变量,蒸发量作为输出变量,进行蒸发量预测的数据训练及预测误差分析。分别进行10、13、16、19min蒸发量预测,相关系数R2分别为:0.72、0.71、0.45、0.07、13min内预测误差相对较小,对工程提前控制相关参数,保持蒸发量平稳有较大意义。

In this paper, a four layer neural network based on gradient descent algorithm is established. 14 parameters such as primary air and furnace temperature are selected as input variables, and evaporation capacity is selected as output variables. Data training and prediction error analysis of evaporation capacity are carried out. The correlation coefficients R2 were 0.72, 0.71, 0.45 and 0.07 respectively. The prediction error within 13 minutes is relatively small, which is of great significance for controlling relevant parameters in advance and keeping evaporation stable. At the same time, whether to use the air pressure parameter under the grate as the input variable is compared.