Abstract: The traditional liquid-level detection method for storage tank by using thermal infrared imager is mainly digital image processing,so the measurement accuracy is not high.In this paper,a mathematical model of heat transfer in storage tank is built and the temperature distribution of external surface of the tank is solved by finite volume method.An inverse heat transfer algorithm for quantitative identification of liquid-level in storage tank is proposed based on the temperature distribution of external surface of the tank and the effects of initial guess,measurement error(σ)and maximum surface temperature difference on the level detection results are investigated.The results show that within the range of acceptable measurement error in engineering applications(σ=2℃),the inversion liquid-level has a high accuracy,with relative error less than 2%.It provides a new thought and a way for the liquid-level detection for storage tank.

Key words: inverse heat transfer problem, liquid-level detection, Levenberg-Marquardt method, infrared thermography, storage tank

摘要： 传统的利用红外热像仪进行储罐液位的检测主要是基于对红外图像进行数字处理,因此测量精度不高。通过建立储罐的传热模型,利用有限体积法,对筒体外表面的温度分布进行了数值模拟,并提出了基于筒体外表面温度分布对筒内液位进行定量识别的传热反问题方法,同时分析了初始假设、测量误差(σ)和最大温差对液位识别精度的影响。结果表明:在工程允许的测量误差范围内(σ=2℃),利用红外热像仪检测储罐外表面温度反演估计储罐液位和罐内流体温度有较高的识别精度,误差在2%以内。该方法为储罐液位和内部温度识别提供了一种新的思路和途径。

关键词: 传热反问题, 液位检测, Levenberg-Marquardt法, 红外热像仪, 储罐