Multi-objective optimization trade-offs for color rendition, energy efficiency, and circadian metrics

Abstract

The spectral power distribution of light sources can be optimized to achieve optimal target goals, such as energy efficiency, color quality, and circadian synchronization. Although several researchers investigated the optimization of light sources for circadian and color rendition metrics, the complex relationship and trade-offs between three distinct parameters (circadian, color rendition, and energy efficiency metrics) have not been widely investigated. Here, multilevel analyses of four clusters of metrics (radiometric, photometric, colorimetric, and circadian) have been conducted. Metric correlation analysis was conducted for 308 theoretical and commercially available light sources, and results suggest that melanopic irradiance correlates highly with irradiance but does not correlate well with correlated color temperature. There was a negative but weak correlation between circadian metrics and energy efficiency measures. Energy efficiency and color quality were inversely correlated. In the following analysis, the spectra of an 18-channel LED system were optimized using a multi-objective genetic algorithm in several combinations, and the effect of the number and the order of the optimization parameters on metric relationships were analyzed. The number of optimization parameters significantly affected the optimal results. The order of optimization parameters only changed the relationship between color fidelity index and luminous efficacy of radiation but did not have an impact on the color fidelity index and melanopic radiance relationship. There was an inverse relationship between melanopic irradiance and the luminous efficacy of radiation. Trade-off analysis highlights the importance of the order and number of target parameters in the optimization process.