Inverse Problem Approach for Estimating Operating Conditions and Uncertainty Quantification in Actual Forced Convection System

Abstract

In this work, we apply the inverse problem method to infer operating conditions and simultaneously quantify uncertainties in steady-state turbulent forced convection systems. This approach dealt with reduced variable dimensions using prior knowledge of the system to avoid underdetermined problems. The methodology encompasses modeling the heat transfer system with the energy equation, employing Newton's method to solve the inverse problem, and quantifying total uncertainties arising from various sources. Significantly, this study extends the inverse heat transfer problem's scope to the actual thermal-fluid problem, overcoming the challenge of minimal sensor usage while ensuring the accuracy of estimation. This work not only demonstrates the application of inverse problem-solving in thermal-fluid systems but also contributes to the further development of sparse sensing methods. The results of this study enhance thermal system monitoring and control efficiency with broader industrial and environmental applications. They also lay the groundwork for advanced sparse sensing and inverse problem-solving, transforming thermal system analysis and optimization methods.