On the hybrid methods used for solving inverse heat transfer problems
Authors :- Shah S.; Parwani A.K.
Publication :- AIP Conf. Proc. 2960, 040010 (2024)
With the advancement of computational facilities, the hybrid methods gained popularity in last two decades for solving the inverse heat transfer problems (IHTP). The current paper gives a brief review about the hybrid methods used for solving IHTP. A hybrid inverse algorithm can be formed by coupling two algorithms which could be either stochastic od deterministic. Different IHTP like inverse conduction problems, inverse convection-diffusion problems, and inverse radiative heat transfer problems were considered. The hybrid algorithms successfully applied for the estimation of boundary conditions, simultaneous estimation of unknown parameters, physical properties of materials, estimation of surface properties of radiative surfaces as well as estimation of mass transfer characteristics. The hybrid algorithms are found to be more accurate and more computationally economical than the single inverse algorithms. In initial era of hybrid algorithms, the stochastic algorithms are coupled with deterministic algorithms but with increasing the computational power and versatility in stochastic algorithms, more and more stochastic algorithms are coupled to form hybrid algorithms. With proper tuning, structure and training, the hybrid algorithm can be very versatile and accurate for solving IHTP.