A new sensitivity-based adaptive control vector parameterization approach for dynamic optimization of bioprocesses.

Area of Science:

• Bioprocess Engineering
• Computational Optimization
• Chemical Engineering

Background:

• Dynamic optimization enhances bioprocess profitability and productivity.
• Control Vector Parameterization (CVP) requires optimal discretization for balancing computational cost and solution quality.
• Traditional refinement methods can be computationally expensive for bioprocess optimization.

Purpose of the Study:

• To propose a novel sensitivity-based adaptive refinement method for CVP.
• To develop an embedded optimization technique for efficiently solving bioprocess problems with sensitive performance indices.
• To achieve economic and effective discretization grids in dynamic bioprocess optimization.

Main Methods:

• A sensitivity-based adaptive grid refinement strategy is introduced.
• New time grid points are adaptively inserted, and unnecessary points are eliminated.
• An optimization technique is embedded within the CVP approach to handle sensitive performance indices.

Main Results:

• The proposed method generates economic and effective discretization grids.
• The approach efficiently solves bioprocess optimization problems, even those with high sensitivity.
• Effectiveness demonstrated on two well-known bioprocess optimization problems.

Conclusions:

• The novel sensitivity-based adaptive refinement method improves discretization efficiency in CVP.
• The integrated optimization technique enhances the solution of complex bioprocess problems.
• The methods offer a more cost-effective and efficient approach to dynamic bioprocess optimization.