Multispectral radiation thermometry approach based on feasible region constraints-divide and conquer optimization

Multispectral radiation thermometry is a widely used non-contact temperature measurement method, particularly in extreme environments. However, accurately retrieving the true temperature remains challenging due to the unknown emissivity of the object. This paper proposes a collaborative optimization approach (feasible region constraints-divide and conquer optimization), which combines a feasible domain constraints method (PCR-PSO) with a divide and conquer optimization algorithm (multi-BFGS). Simulation results show that the proposed approach reduces the temperature inversion error from 0.59% to 0.19% compared to the traditional BFGS algorithm. Experimental results obtained with stainless steel samples indicate an average error of less than 0.2% and an average processing time of 0.2 seconds, highlighting the potential for real-time temperature measurement in engineering applications.