Non-Dominant Genetic Algorithm for Multi-Objective Optimization Design of Unmanned Aerial Vehicle Shell Process

Area of Science:

• Materials Science
• Manufacturing Engineering
• Optimization Theory

Background:

• UAV housing parts require high quality, with warpage and mold index as critical defect parameters.
• Optimizing injection molding process parameters is crucial for minimizing these defects.

Purpose of the Study:

• To develop a multi-objective optimization system for enhancing UAV housing part quality.
• To identify key injection molding parameters influencing part defects and optimize them.

Main Methods:

• Utilized Pareto-optimized frames and a three-stage optimization system.
• Employed mold flow analysis, Kriging function prediction, and a non-dominant rank genetic algorithm II (NSGA-II).
• Experimental verification of Pareto optimal frontier points.

Main Results:

• Injection pressure and time were identified as the most influential parameters for mold index and warpage.
• Achieved up to a 96.2% optimization rate for the die index.
• Average optimization rate across four nodes was 91.2% with a low error rate of 8.48%.

Conclusions:

• The proposed optimization system effectively improves UAV housing part quality.
• Injection molding parameters, particularly pressure and time, can be precisely controlled to minimize defects.
• The method is suitable for actual production needs, enhancing UAV component manufacturing.