| The mechanical pulp bleaching process at Irving Paper is manually controlled, which degrades the quality of the produced paper due to the variations in the pulp brightness. The objective of this research was to study the possibility of controlling the pulp bleaching process at the Irving Paper mill. This would improve the pulp quality by minimizing the final pulp brightness variability, and achieve some economical benefits by minimizing the consumption of the bleaching chemicals. The bleaching process was thoroughly studied and modelled in this research. A delay time estimator was designed to tackle the problem of the long variable delay time, which is considered the biggest challenge in controlling this process.; The model predictive control (MPC) strategy is chosen to control the bleaching process taking into account the constraints, which are imposed by the operating conditions of the process itself. In order for the MPC controller to handle such constraints, an interior point optimization method was incorporated.; The designed MPC controller formed a basis for an industrial control prototype. The control technology was designed and implemented in the bleaching process at Irving Paper mill, in order to test and demonstrate the performance of the MPC controller in real time. This, in turn, demonstrated the closed loop behaviour of the bleaching process itself in real time. The implementation results and experiments were very promising, and agreed well with the simulation results. However, two major changes are required for ultimate success: implementing multivariable control of brightness and pH, and imposing constraints on pulp flow through the system, e.g. scheduling it or limiting its rate of change. |
