Optimization of 3D printed drone performance using synergistic multi algorithms
Pedklah Kamonsukyunyong, Tossapon Katongtung, Thongchai Srinophakun, Somboon Sukpancharoen
Abstract
This research aims to develop and optimize a cost-effective, high-performance 3D-printed drone through the integration of advanced computational techniques while maintaining manufacturing accessibility and operational efficiency.. We combined Box-Behnken design (BBD) within response surface methodology (RSM), artificial neural networks (ANN), and the Osprey Optimization Algorithm (OOA) to enhance drone performance while maintaining cost-effectiveness. The optimization focused on frame length (20-30 cm), motor efficiency (1250-1750 kV), and flight duration (3-5 minutes). Results showed a 62.20% increase in stability and a 14.27% reduction in battery power loss compared to conventional designs. The RSM model (R² = 0.991) slightly outperformed the ANN model (R² = 0.929) in predictive capabilities. OOA integration yielded optimal parameters: 20.00 cm frame length, 1456.47 kV motor efficiency, and 3.69 minutes flight duration. This study demonstrates the effectiveness of combining 3D printing with advanced optimization techniques, paving the way for affordable, high-performance drones suitable for various applications from precision agriculture to urban delivery. Future research should address scalability for larger drones and investigate long-term durability of 3D-printed components in real-world conditions.