基于最优加权的大面积金刚石膜衬底温度数据融合研究

Study on Substrate Temperature Data Fusion in Large Area Diamond Coating Based on Optimal Weighting Algorithm

为保证大面积金刚石镀膜衬底温度均匀,采用多只温度传感器多点测量衬底表面温度。为滤除测量噪声,采用最优加权融合方法对每一时刻多只传感器测量的温度进行数据融合,并分别采用最小二乘、Bayes估计、递推最小二乘法进行第二步数据融合,以进一步提高温度测量的精度。研究结果表明:最优加权与最小二乘结合算法、最优加权与Bayes估计结合算法的温度数据融合结果精度很高,在温度测量受各种扰动时均表现出良好的稳健性;最优加权与递推最小二乘结合算法温度测量数据融合的精度和稳健性相对较差。

The substrate temperature uniformity in large area diamond coating is important to obtain the film with high quality.To filter the measurement noises,multi thermocouples were used to measure the substrate＇s surface temperature from different points and the optimal weighting data fusion algorithm was utilized for dealing with the measured temperature values at the same sampling moment.To further improve the temperature measurement precision,the least square,Bayes estimation and recursive least-square algorithms were adopted for the further data fusion.The considerable research results demonstrates that the temperature fusion results＇ precision is quite high by the combination algorithm of optimal weighting and least square,optimal weighting and Bayes estimation which are with good robustness when different disturbances occur,and the combination algorithm of optimal weighting and recursive least-square is with worse robustness and its temperature data fusion results＇ precision is comparably less.