Predictive Control Implementation

Aggregation MPC (Model Predictive Control):

• Introduced a generalized linear aggregation framework for MPC.
• Developed equivalent aggregation concepts to simplify optimization problems, enabling real-time applicability.
• Optimized the separation between offline (design phase) and online (real-time execution) computations to ensure high efficiency.

Applications:

• Air Separation Systems

  • Controlled fluctuations in process variables during separation.
  • Enabled rapid tracking of operational setpoints under external load changes.
  • Successfully tested embedded (DSP-based) predictive controllers in industrial settings.

• Injection Molding Systems

  • Minimized transient pressure fluctuations during the hold pressure phase, improving product quality.
  • Reduced energy consumption and increased stability in high-volume production environments.

• Photovoltaic Inverter Systems

  • Optimized the duty cycle to minimize harmonic distortions and waveform irregularities.
  • Nominated for the 2014 Chinese Control Conference Guang Zhi Zhi Award for these innovations.

• Boiler Combustion Control System

  • Collaboration between the research team and XinHua/GE Control Company on a dry clinker production project in cement plants, utilizing the CSC Laboratory’s embedded MPC (Model Predictive Control) controller.
  • Energy Savings: Achieved average coal savings of 6%, with a peak efficiency of 8% in relevant processes.