An Efficient Iterative Learning Predictive Functional Control for Nonlinear Batch Processes.

Area of Science:

• Control Engineering
• Nonlinear Systems Theory
• Batch Process Optimization

Background:

• Iterative learning model-predictive control (ILMPC) offers learning and tracking capabilities for batch processes.
• Fast batch processes with nonlinear dynamics pose challenges for ILMPC due to computational efficiency and tracking accuracy trade-offs.

Purpose of the Study:

• To propose an efficient iterative learning predictive functional control (ILPFC) algorithm.
• To address the computational and accuracy challenges of ILMPC in fast, nonlinear batch processes.

Main Methods:

• Linearized the nonlinear system along the reference trajectory to create a 2-D tracking-error predictive model.
• Compensated linearization errors using the Lipschitz condition to bound the objective function.
• Applied predictive functional control (PFC) in the time domain to reduce computational burden.

Main Results:

• Developed a novel ILPFC algorithm that enhances control efficiency.
• Theoretically analyzed the stability and convergence of the ILPFC with terminal constraints.
• Validated the ILPFC's effectiveness through simulations on an unmanned ground vehicle and a batch reactor.

Conclusions:

• The proposed ILPFC algorithm effectively controls fast batch processes with nonlinear dynamics.
• ILPFC achieves a superior balance between computational efficiency and tracking accuracy compared to ILMPC.
• The method demonstrates practical applicability in complex dynamic systems.