Experimentation, Taguchi and machine learning-based optimisation of parameters concerning heat losses from the trapezoidal box-type solar cooker

Badri N. Panigrahi,Jayashree Nayak,Jagadish Bhatra,Sudhansu S. Sahoo,Debasis Gountia,Balaji K Choudhury
DOI: https://doi.org/10.1080/15567036.2024.2416084
2024-10-17
Energy Sources Part A Recovery Utilization and Environmental Effects
Abstract:The optimal design parameter combinations for a trapezoidal solar cooker that yield the lowest heat loss coefficient have been determined using a thorough Taguchi and Machine learning-based optimization approach that has been established for this study. Data for the previously described optimization have been generated through the use of an experimental approach in a laboratory environment. Numerous influencing factors are taken into account, including the temperature of the bottom plate, the depth of the cavity, the emissivity of the bottom cover, and the wind-induced heat transfer coefficient on the top glass cover. For the current investigation, the ideal factor combination is found to be a parametric combination plate temperature of 350K with 0.4 plate emissivity, 15W/m 2 convective heat transfer coefficient, and 0.25 m height. To determine the percentage contribution of each control element to the heat loss coefficient, analysis of variance (ANOVA) has been used for the collected data and signal-to-noise ratios. Plate temperature at 74.72% and emissivity at 23.84% are the primary contributing variables found. Likewise, the least contributing factors like height (0.57%) and convective heat transfer coefficient (0.32%) are discovered. A regression equation is used to represent the connection between the response and the control factor. This formula is suggested as a forecast formula for calculating the heat loss coefficient.
energy & fuels,engineering, chemical
What problem does this paper attempt to address?