Control of pulp washing process based on neural network
Mengxiao Wang · China Pulp & Paper Industry · 2011
Pulp washing process is a process of multi-factors,long time delay,nonlinear characteristics.The soft sensor model of the residual soda of washed pulp and the Baume degree of the first stage black liquor were modeled via two-step neural network identification method and least square method.A multi-objective optimization model of pulp washing process was done from three aspects of high quality,high yield,and low consumption.Simulation results and the practical application indicate that the soft sensor model of residual soda and Baume degree have a good esti-mation.Multi-objective optimization not only ensures the quality of pulp washing but also improves the output of washed pulp.