基于改进NSGA-Ⅱ算法的多光谱测温数据处理

Data Processing of multi-wavelength thermometry based on improved NSGA-Ⅱ algorithm

多波长测温能够对高速运动物体的温度进行测量,通过算法对其数据进行处理有助于提升测量的精度。通过对四光谱高温计获得温度的数据进行处理,计算得到目标的真实温度,测温范围为700~900℃。文中对NSGA-Ⅱ算法进行了修改,使其交叉变异概率以及交叉策略具有动态的调整的能力,并采用了与差分进化算法相融合对变异的方向进行了优化。利用改进后的算法从求解的稳定性、计算的精度、收敛速度以及对于经典发射率模型物体温度的求解精度上与目前测温所采用的遗传算法进行比较,结果表明应用改进的NSGA-Ⅱ算法求得的目标温度精度较高,稳定性较强。

The multi-wavelength thermometry can be used to measure the temperature ol moving objects with high speed, and the data acquired can be improved with higher precision by processing the data with the algorithms. In this article the temperature data acquired by the four spectroscopic pyrometers was processed, the real temperature of the target was calculated within the scope of 700？900 In this paper, the NSGA-II algorithm was modified to make the crossover and mutation probability and crossover strategy own the ability of dynamic adjustment, and the direction of variation was optimized by the fusion of differential evolution algorithm. The improved algorithm was compared with the genetic algorithm in aspects of stability of the solution, precision of calculation, convergence speed and the precision of temperature for classical emissivity models. The results show that the improved NSGA-n algorithm has higher accuracy and stronger stability in calculating the target temperature.