True-temperature inversion algorithm for a multi-wavelength pyrometer based on fractional-order particle-swarm optimization

Herein, a method of true-temperature inversion for a multi-wavelength pyrometer based on fractional-order particle-swarm optimization is proposed for difficult inversion problems with unknown emissivity. Fractional-order calculus has the inherent advantage of easily jumping out of local extreme values; here, it is introduced into the particle-swarm algorithm to invert the true temperature. An improved adaptive-adjustment mechanism is applied to automatically adjust the current velocity order of the particles and update their velocity and position values, increasing the accuracy of the true temperature values. The results of simulations using the proposed algorithm were compared with three algorithms using typical emissivity models: the internal penalty function algorithm, the optimization function (fmincon) algorithm, and the conventional particle-swarm optimization algorithm. The results show that the proposed algorithm has good accuracy for true-temperature inversion. Actual experimental results from a rocket-motor plume were used to demonstrate that the true-temperature inversion results of this algorithm are in good agreement with the theoretical true-temperature values.