Chameleon swarm algorithm for data processing of a light-field multi-wavelength pyrometer

It is recognized that unknown emissivity and ill-posed radiation equations present significant challenges to light-field multi-wavelength pyrometry. Furthermore, emissivity range and choice of initial value also have a significant impact upon the measurement results. This paper demonstrates that a novel chameleon swarm algorithm approach could be used to ascertain temperature information from light-field multi-wavelength data at a higher accuracy level without prior emissivity knowledge. The performance of chameleon swarm algorithm was experimentally tested and compared with the traditional internal penalty function and generalized inverse matrix-exterior penalty function algorithms. Comparisons of calculation error, time, and emissivity values for each channel show that the chameleon swarm algorithm is superior in terms of both measurement accuracy and computational efficiency.