Self-Organizing Map Network for the Decision Making in Combined Mode Conduction-Radiation Heat Transfer in Porous Medium

Self-organizing map (SOM) network is used for the first time for decision making in combined mode conduction-radiation heat transfer in porous medium. Variety of variations in the topology of SOM networks are explored for the decision making in porous ceramic matrix (PCM) operation. The heat transfer problem considered a PCM with heat generation zone and air flowing through it. The governing equations of the model are solved with the help of finite volume method. Pairs of solid and gas phase temperature profiles are computed for various values of extinction coefficient β and convective coupling P 2 , with the help of numerical model. Based on the values of β and P 2 , the PCM operates in four regimes. The pairs of temperature profiles computed by numerical model correspond to different regimes of operation and are used to train the SOM. Once training of the SOM is completed, four new pairs of temperature profiles of gas and solid phase are fed to the SOM, and the SOM correctly identifies the regimes of operation of the PCM. The SOM model is also trained with 1% noise in the input data and is able to correctly identify the regimes of operation of PCM, in that case as well. Present approach intends to simplify the operation and design of the PCM-based equipment like burners.