TMP Predictive Quality Controller

The Advance Quality Control solution is a multivariable model predictive control (MPC) application specifically designed to handle the complex requirements of TMP quality control.


  • Control multiple main lines, screens and reject lines with common control objectives (total power, total production, inventory, average quality).
  • Controls quality based on real-time continuous Ponderosa™ quality predictions, with no intermediate control levels or heuristic control decisions.
  • Truly multivariable dynamic control with no one manipulated variable dedicated to one control variable (unless desired).
  • Adaptable to different refiner types. The TMP Predictive Quality Controller has been used to control double disc refiners, single disc refiners, twin disc refiners, as well as refiners with plate gap measurements, plate position measurements, hydraulic pressure measurements, and mechanically loaded refiners.
  • Contains an off-line tuning simulation that displays the effect of tuning changes, optimization changes, and setpoint/limit changes on control performance.
  • Automatically decouples ALL manipulated variable to control variable relationships including primary to secondary refiner effects.
  • Flexible architecture that allows for addition of manipulated, controlled and constraint variables.
  • Flexible LP and QP optimization for high level control objectives (production maximization, specific energy minimization, refiner consistency optimization)
  • Option to control inventory including main line latency chest, unrefined rejects, and pulp storage, while accounting for throughput effects on quality.
  • Option to hide relationships from control decisions while using the true dynamic relationships as a feedforward component of the solution.
  • Option to make automatic on-line changes to the gain and/or deadtime of the MPC dynamic models based on known changes in process relationships. This can be used to account for non-linearities in the process models or to hide relationships in certain operating conditions.


  • Sites that have implemented the TMP Predictive Quality Controller have achieved 30% to 80% reduction in variability and 4.5% to 15% increase in throughput