The shock radiation temperature of materials is an important physical parameter for the study of equation of state under the action of high temperature and high pressure, which is of great significance to weapon development, scientific research, and industrial manufacturing. Aiming at the characteristics of instantaneous non-contact, complex measurement environmental noise, and unmeasurable emissivity inverted by temperature, a new temperature inversion method is designed to improve the accuracy of temperature acquisition. According to the constrained optimization theory, the multiplier penalty function method and particle swarm optimization (PSO) algorithm are combined to realize the series of the two models, and the PSO-multiplier penalty function algorithm is improved. The results show that the hybrid model solving method fully combines the advantages of the two single algorithms and improves the temperature inversion accuracy and operation efficiency of the test data of impact radiation. It provides a guarantee for studying the real temperature of material under impact radiation.