RSM-based optimization and SLEC evaluation of a high-frequency PWM LED driver circuit for disaster-relief lighting drones

Purpose This study investigates the optimization of (Unmanned Aerial Vehicle) UAV-based illumination systems using experimental design, Dialux simulations, and image processing methods, with a focus on maximizing lighting performance while minimizing energy consumption. Design/methodology/approach A full factorial experimental design with four factors (circuit status, spot angle, height, and spot efficiency) at multiple levels was employed. Dialux software was used to calculate horizontal and vertical illuminance values and homogeneity indices, while Python-based image processing (OpenCV and polynomial regression) extracted detailed illumination profiles. Response Surface Methodology (RSM) was applied in Minitab to evaluate factor significance and to determine the optimal parameter set. Findings The results showed that spot efficiency was the most dominant factor (p < 0.001) across all responses, while height significantly influenced horizontal average illuminance (p = 0.000). The maximum horizontal illuminance (Eyavg) of 7.10 lux was achieved at 10 m with narrow beam efficiency, whereas the minimum specific lighting energy consumption - Ey (SLEC Ey) was 0.010 W/lux under the same conditions. RSM optimization yielded a composite desirability of 1.000, indicating perfect agreement with the target objectives. Model reliability was confirmed with high coefficients of determination (R2 = 86–99%). Originality/value This study introduces a novel optimization framework for UAV-based illumination systems by integrating experimental lighting performance data with RSM. Unlike previous research that mainly focused on communication or surveillance aspects of drones, our work emphasizes quantitative analysis of lighting efficiency and energy consumption under varying operational factors. The combination of ANOVA modeling, surface plots, and desirability functions provides statistically robust results and identifies the optimum operational parameters with high accuracy.