Extreme learning machine evolved by fuzzified hunger games search for energy and individual thermal comfort optimization

Green building with the purpose of increasing personal comfort is a vital problem that requires immediate attention. As a result, it is necessary to measure interior individual thermal comfort (IITC) to design a thermal environment that is both comfortable and ecologically beneficial. We offer a hybrid Extreme Learning Machine-Fuzzified Hunger Games Search model (ELM-Fuzz-HGS) to solve this issue. First, an ELM forecasting model is suggested to evaluate the predicted mean vote (PMV) metric. However, the randomized tuning of ELM parameters results in the final model being unreliable, ill-conditioning, and lacking resilience. To overcome these shortcomings, we propose a novel fuzzified HGS learning technique, dubbed Fuzz-HGS, for maximizing dynamic assessment and energy efficiency. Actually, fuzzy systems were employed to address the HGS's shortcomings in balancing the exploration and exploitation phases. A thorough empirical data-base based on ASHRAE and ISO standards is developed to ensure a fair comparison, which in-cludes six input data and one status. tAlong with Fuzz-HGS, five ELM-based metaheuristics are developed: the ELM-henry gas suitability optimization (ELM-HGSO), the ELM-arithmetic opti-mization algorithm (ELM-AOA), the ELM-chimp optimization algorithm (ELM-ChOA), the ELM -modified grey wolf optimizer (ELM-MGWO), and the marine predator algorithm (ELM-MPA). It has been discovered that from high to low, ELM-Fuzz-HGS, ELM-ChOA, ELM-MPA, ELM-MGWO, ELM-HGSO, ELM-AOA, and conventional ELM are the best performing ELM-based models for forecasting the PMV model from 42 to 36 to 30 to 24 to 18 to 6 such that the most accurate ELM-Fuzz-HGS is recommended for PMV model prediction.

Automated summary

Green building with the purpose of increasing personal comfort is a vital problem that requires immediate attention. This article proposes a hybrid model to measure interior individual thermal comfort and design a comfortable and ecologically beneficial thermal environment. Different models are compared and the most accurate model is recommended for prediction.