Monitoring and compensation of LED emission shifts in multi-channel LED systems via spectral sensors

The development of light-emitting diodes has revolutionized lighting technology, enabling precise spectral adjustments in luminaires with multiple different LEDs for demanding applications, such as museums, film productions, or color proofing in the printing industry. Over time, LEDs and optical components degrade, causing spectral shifts, which are primarily noticeable in the form of color shifts and reduced luminous flux. The aim of this work is to investigate approaches for processing the signals from mass-manufacturable spectral sensors to monitor lighting parameters and to test their viability in a self-compensating light source. In this context, application-relevant sensor properties are investigated, including the operating parameters as well as the varying spectral sensitivities between devices of the same type. The suitability of the Wiener filter and interpolation approaches for sensor data processing to monitor spectral emission is evaluated by exposing different LED types to a temperature change of 40 K, a halving of the diode current, and accelerated aging through 2000 hours of stress operation. The differences found between spectroradiometer measurements and spectral sensor-based estimates result in maximum errors in the irradiance of 1.7 %, the illuminance of 2.2 %, the chromaticity coordinates of 0.0042 Δu′v′ and an nRMSE of 0.031. The largest errors are observed in connection with a significant change in the radiometric integral. If only the aging experiment and the temperature-induced emission changes are considered, the maximum errors of the spectral estimates are less than 0.9 % with regard to the irradiance, 0.6 % with regard to the illuminance, 0.0011 with regard to the chromaticity distance Δu′v′ and 0.031 with regard to the nRMSE. The effect of these errors on the achievable compensation of current and temperature changes in multi-channel LED light mixtures is evaluated in the last step. For this purpose, test scenarios with diode current changes of 20 % or temperature changes of 35 K in individual LED channels of two RGBALC LED light mixtures with different CCT are realized. All spectral monitoring approaches investigated in this work are capable of achieving compensation results that are on par with those of a laboratory spectroradiometer, even without individual sensor characterization. However, the necessity for a hybrid compensation strategy is identified: To account for scenarios in which an occurring emission change cannot be compensated in the spectral power distribution, a prioritization of chromatic over spectral compensation should be investigated in future work. The insights gained in this thesis can be used directly for the implementation of self-adjustment of multi-channel LED luminaires outside of actual operation with regard to predefined light mixtures. In addition, an application in research on LED degradation is possible, for example, in field tests of entire luminaires with previously unfeasible in situ monitoring of the degradation process. The outlook also presents the possible next steps for the application of spectral sensor feedback in the live operation of a lighting system. The integration of spectral sensors thus has the potential to significantly advance lighting technology in various fields of research and application so that even the most demanding expectations of long-term stability and consistency between light sources can be met in the future.