水质油污染原位传感器稳定性设计

摘要

Due to the instability of the light source,ambient light interference,the non-linearity of the standard curve and the influence of contamination,the stability of the sensor was poor. So,the article based on the principle of UV fluorescence measurement,and had optimized design in structure,craft and algorithm. Optical path design introduced a reference optical path,ZEMAX software simulation results show that"cross"optical path was feasible. PTFE,optical coating and brush design could effectively prevent oil and biological adhesion. Avoid light mask crea-ted a detection dark cavity for the sensor,which effectively solved the problem of ambient light interference. Experi-ments showed that the variation coefficient of the fluorescence signal was less than 0.3% during the sensor measure-ment. The coefficient of variation of fluorescence signal was 1.28% in light intensity decay experiment with 30 μg/L as an example. When the sample concentration was 62.4 μg/L,the expansion uncertainty was 0.256 μg/L(k=2). The stability design of the sensor had achieved satisfactory results,but couldn't actually application,because that the main factors such as turbidity and temperature were not introduced into the algorithm correction model.%由于受光源不稳、环境光干扰、标准曲线非线性及沾污的影响,传感器的稳定性表现不佳.为此,基于紫外荧光测量原理,分别从结构、工艺及算法3个方面做了优化设计.光路设计中引入了参比检测,ZEMAX软件仿真结果显示"十"字光路可行.聚四氟乙烯、光学镀膜的采用及电刷设计可有效防止油污粘附及生物附着.遮光罩为传感器营造一个检测暗腔,有效解决了环境光干扰问题.实验表明:传感器测量样品时,荧光信号的变异系数<0.3%.以浓度30 μg/L为例,模拟光强衰减时荧光信号的变异系数为1.28%.样品浓度为62.4 μg/L时,扩展不确定度为0.256 μg/L(k=2).传感器的稳定性设计取得了较理想的成果,但距离实际应用还有差距,主要是浊度、温度等影响因子未引入到算法修正模型中.