Multiobjective spectral optimization for circadian-effective LED lighting systems

LED lighting systems have evolved beyond basic illumination to support human-centric applications, particularly in regulating circadian rhythms through spectral control. However, existing LED design methods primarily focus on visual requirements while neglecting the physiological impact of light spectra on human biological rhythms, which are governed by wavelength-specific melanopsin responses. We proposed a multiobjective spectral optimization framework for circadian-effective LED lighting systems that simultaneously optimizes color quality and circadian efficacy. The framework integrates a five-channel LED synthesis model (deep red, green, blue, warm white, and cold white) with advanced optimization algorithms including particle swarm optimization and genetic algorithms. For daytime illumination mode, the system achieves a correlated color temperature (CCT) of 5873 K with a color fidelity index (Rf) of 94.9 and a melanopic daylight efficacy ratio (mel-DER) of 0.501, promoting alertness. For nighttime sleep-aid mode, it attains a CCT of 2792 K with an Rf of 93.9 and a mel-DER of 0.165, minimizing circadian disruption. Clinical validation with 11 healthy subjects demonstrated that the optimized lighting significantly improved sleep efficiency (p ⪅0.01) and deep sleep proportion (p ⪅ 0.05) compared to conventional LED lighting and dark conditions. The proposed method provides an effective solution for dynamic spectral control in intelligent lighting systems.