Optimization Study of RGBW LED Light Mixing Based on NSGA-MA-II

Abstract In this study, based on the RGBW LED mixing model with PWM dimming characteristics, the three-objective problems of luminous flux, CRI maximization and color coordinate deviation minimization are transformed into an underdetermined system, and an improved multi-objective genetic algorithm NSGA-MA-II is proposed, which is able to effectively achieve single-objective optimization of luminous flux and CRI. For the three-objective optimization problem of luminous flux, CRI and color coordinate deviation, the interval normalized comprehensive evaluation function is constructed, and the luminous flux, CRI and color coordinate accuracy are comprehensively ranked to realize the optimal duty cycle automatic selection. The experimental results show that the method can control the color coordinate deviation within ±0.005 and ΔCCT ＜100K within the color temperature range of 3571-6520K, while taking into account the dynamic balance of luminous flux and CRI, which verifies the effectiveness and engineering applicability of the method.