Achieving Optimal Paper Properties: A Layered Multiscale kMC and LSTM-ANN-Based Control Approach for Kraft Pulping

The growing demand for various types of paper highlights the importance optimizing kraft pulping process to achieve desired properties. This work proposes a novel multiscale model optimize and obtain combines mass energy balance equations with layered kinetic Monte Carlo (kMC) algorithm predict degradation wood chips, depolymerization cellulose, spatio-temporal evolution Kappa number cellulose degree polymerization (DP). A surrogate LSTM-ANN is trained on data generated from under different operating conditions, dealing both time-varying time-invariant inputs, an LSTM-ANN-based predictive controller designed set-point values DP while considering constraints. results show that able drive use computationally faster high accuracy low offset.