Order of Addition in Mixture-Amount Experiments.

Area of Science:

• Statistics
• Experimental Design
• Chemical Engineering

Background:

• Mixture experiments focus on component proportions.
• Mixture-amount experiments consider both proportions and total amount.
• Order-of-Addition (OofA) experiments investigate the impact of component addition sequence.

Purpose of the Study:

• To develop efficient designs for Order-of-Addition (OofA) mixture-amount experiments.
• To optimize designs for maximizing G-optimality and minimizing experimental runs.
• To enable precise estimation of mixture-component and OofA effects.

Main Methods:

• Construction of full mixture OofA designs.
• Application of the Threshold Accepting (TA) Algorithm for subset selection.
• Utilizing G-optimality and G-efficiency criteria for design assessment.
• Employing Fraction of Design Space (FDS) plots for visual evaluation.

Main Results:

• The TA Algorithm selects optimal subsets from full OofA designs.
• Achieved designs balance G-optimality with reduced experimental runs.
• Designs effectively estimate mixture-component model parameters and OofA effects.
• FDS plots provide visual insights into prediction capabilities.

Conclusions:

• The proposed method offers efficient OofA mixture-amount experimental designs.
• The TA Algorithm is effective in optimizing design selection.
• The developed designs support accurate parameter estimation and prediction assessment.