Maximum luminance prediction method for multichannel LED light sources under chromaticity constraints for tunable...

Multichannel light-emitting diode (LED) light sources (MLLSs) offer substantial flexibility for tunable lighting, enabling dynamic adjustment in both correlated color temperature (CCT) and luminance. However, determining the maximum luminance achievable by an MLLS at a given chromaticity remains a challenge. This is because simple linear scaling of the input control signals often induces undesirable color shifts, a consequence of the nonlinear light output and chromaticity variations inherent to each LED channel. To address this challenge, we propose a prediction method to estimate the maximum luminance of an MLLS at a specified chromaticity. The method integrates a color prediction model with a luminance optimization model, employing a constrained optimization algorithm based on differential evolution. The method effectively accounts for nonlinear light output and chromaticity variation across channels. We validated the method on two distinct MLLSs: a four-channel system (MLLS1) dimmed via 10-bit amplitude modulation and an eight-channel system (MLLS2) dimmed via 10-bit pulse-width modulation. The validation was conducted at ten representative white-light chromaticities, with CCTs ranging from 2200 K to 6500 K. The results demonstrate excellent prediction accuracy, with mean relative luminance errors of 1.11% and 0.88%, and mean Commission Internationale de l'Eclairage 1976 chromaticity differences of 1.07 × 10^-3 and 0.57 × 10^-3 for MLLS1 and MLLS2, respectively. These performances are comparable to those of the underlying color prediction model.