基于MIV-BP神经网络的成品烟丝质量预测模型构建

目的：构建卷烟制丝过程成品烟丝质量模拟预测模型。方法：使用平均影响值法（the Mean Impact Value, MIV）对制丝加工过程工艺参数进行筛选，然后通过反向传播（Back-Propagation，BP）神经系统构建起制丝关键工艺参数和成品烟丝质量的模拟模型。结果：通过模拟数据与实测数据比较，填充值的模拟预测平均相对误差为3.16%；整丝率的模拟预测平均相对误差为0.67%；碎丝率的模拟预测平均相对误差为5.33%。结论：该模型预测值与实测值之间相对误差较小，精确性高，该模型适用于卷烟制丝生产过程工艺参数仿真优化。

Prediction model building for finished tobacco quality based on MIV-BP neural network

Objective: The simulation and prediction model between the process of silk making and the quality of finished tobacco was established. Methods: The average influence value method was used to screen the process parameters in the process of making silk, and then a simulation model of the key process parameters of the silk and the quality of the final tobacco was constructed through the Back-Propagation neural network. Results: Comparing the simulated data with the measured data, the average relative error of the simulated prediction of the filling value was 3.16%; the average relative error of the simulated prediction of the whole cut rate was 0.67%; the average relative error of the simulated prediction of the broken cut rate was 5.33%. Conclusions: The relative error between the data predicted by this model and the real data is small, and the accuracy is high, which provides a theoretical basis and simulation method for the optimization of process parameters in the tobacco process.