在电弧等离子体的光谱诊断中, 标准温度法测温原理与目前先进的图像传感技术相结合, 通过特征谱图像完成电弧全场温度信息采集, 因其良好的时、 空分辨率而被广泛应用于电弧温度测量。 但是谱线的发射系数与等离子体温度不是单调变化关系, 传统标准温度法选取一条ArⅠ谱线完成对电弧等离子体的测量, 在电弧内部的高温电离区域产生谱线辐射强度降低的现象, 需要人为判定电弧不同位置所处的温度区间才能完成温度的计算, 整个过程无法通过软件自主完成。 针对此问题, 根据电弧等离子体的局部热力学平衡条件, 探索一种基于双特征谱线的标准温度法测温原理, 通过融合电弧在外层低温区域聚集的Ar原子发出的ArⅠ谱线发射系数场, 和在高温区域的Ar一次电离离子所发出的ArⅡ特征谱线发射系数场, 将达到ArⅠ谱线标准温度的位置处的ArⅡ谱线发射系数作为电弧不同温度区域的判定依据, 完成电弧等离子体高温区域的自动判别, 继而应用ArⅠ谱线发射系数与温度对应关系在电弧高、 低温区域分别计算电弧温度, 消除单一的ArⅠ谱线发射系数场暗区给计算带来的不利影响; 设计并搭建了一种镜前分幅采集系统, 其中分光镜将弧光等能量分成两束, 利用两组反射镜和窄带滤光片建立起两路光学通道, 使CMOS在一次曝光中完成两组电弧特征谱图像的采集, 并且两幅图像的采集时刻、 焦距、 光圈等拍摄参数完全一致, 达到良好的时间、 空间一致性, 从而减小谱线融合时误差的输出, 满足了原位获取两组电弧特征谱图像的需求; 为验证测量系统可行性以及后期的电弧图像提取, 以黑白棋盘为标靶, 用Harris算子对系统采集的图像进行扫描, 根据角点坐标证明系统所采集的两幅图像具有良好的一致性, 并且据此将两幅图像做归一化处理, 以便后期的电弧特征谱图像的提取; 通过假设所测电弧等离子具有轴对称属性, 以CMOS所采集的特征谱图像亮度信息作为电弧发射系数场在不同角度下的投影依据, 经过中值滤波降噪后, 利用ML-EM迭代重建算法求解电弧的三维发射系数分布。 实验中, 选择受自吸收效应影响较小的ArⅠ696.5 nm谱线和ArⅡ480.6 nm谱线为测量目标, 并且在696.5 nm谱线的光通路中加入OD0.4的中性减光片, 使两幅特征谱图像的最高亮度值保持一致。 选取150A焊接等离子弧为测量对象, 经ML-EM法三维还原后, 将两条谱线发射系数场等像素融合, 在ArⅠ谱线发射系数达到最大值的像素点位置处, ArⅡ谱线发射系数达到 ε_rp, 判定ArⅡ谱线发射系数大于 ε_rp的像素点位置为电弧高温区域, 其余位置为低温区域, 最终在不同温度区域自动完成焊接等离子弧的温度计算。 实验结果表明696.5 nm谱线和480.6 nm谱线发射系数场融合后可以自动识别电弧高温区域, 继而完成电弧等离子体的自动测量, 为电弧温度实时监测的实现提供更多可能。

In the research field of arc plasma spectrum diagnosis, combined with the advanced image sensing technology, the Fowler-Milne method uses spectral image to obtain temperature information of arc plasma. Because of its high time and space resolution, the Fowler-Milne method is widely used in arc plasma temperature measurement. However, the relationship between line emission coefficient and temperature is not monotonous, and traditional Fowler-Milne method selects one ArⅠ spectral line to complete the measurement, which leads to the decrease of line intensity, and the process of measurement needs researchers to determine the temperature range of different locations to complete the calculation of temperature. The whole process can’t be automatically completed by software. In view of this problem, based on PLTE conditions of arc plasma, applying the partial LTE model of arc plasma, modified Fowler-Milne method based on two spectrum line, which combines the ArⅠ spectral line emitted by Ar atoms in an outer low temperature region of arc, and the ArⅡ spectral line emitted by Ar ionization ion in high-temperature area of arc, to determine the different temperature range of the arc plasma, and then the whole temperature is calculated by the temperature corresponding to the intensity of the ArⅠ spectral line in low-temperature area and high-temperature area, eliminating the adverse effects of single ArⅠ spectral line emission coefficient field. A light splitting system was also designed and built, dividing the light of arc into two beams by a spectroscope. Two sets of reflectors and narrow-band filters were used to collect the image of two sets of arc spectral images though one exposure, of which parameters such as the focal length aperture are exactly the same, achieving good time and spatial consistency and reducing the error of emission coefficient fusion. In order to verify the feasibility of measurement system and arc image extraction, black and white chessboard was used as a target, and the extraction of corners extracting of two image proved the system satisfies the demand of collecting two groups of arc spectral images, and was also used to normalize two images for the extraction of arc image in the later stage. Based on the assumption that the plasma arc has axisymmetric properties, with the brightness information of spectral image CMOS collected as the integration of emission coefficient under different angle projection, after the median filter noise reduction processing, ML-EM method was used to reconstruct the 3D emission coefficient distribution from the 2D luminance distribution. In the experiment, ArⅠ696.5 nm spectral line and ArⅡ480.6 nm spectral line with little self-absorption effect were selected. The OD0.4 ND filter was added in the pathway of 696.5 nm spectral line, to make the maximum brightness value of the two spectral images consistent. 150 A welding plasma arc was measured in the experiment. After three - dimensional reduction of ML-EM method, the two spectral line emission coefficient fields were fused. At the pixel point location where the ArⅠ spectral line reaches the maximum value, ArⅡ spectral line reaches ε_rp, which were used to determine the high-temperature zone or low-temperature zone. The measurement of the plasma arc of 150 A showed that 696.5 and 480.6 nm spectral line can automatically identify the high temperature zone in welding arc plasma, making it more possible for the arc temperature real-time monitoring to be realized.