Genetic Algorithm and Grasshopper Optimization Algorithm with Metaoptimization and RL-Based Parameter Fine-Tuning and Their Comparison for Optimal Thermal Performance Analysis of Buildings in Tropical Climate

Abstract

Energy consumption by heating, ventilation, and air-conditioning systems has driven the need for energy-efficient building designs. This research presents a methodology for optimizing the design features of intermittently active air-conditioned commercial buildings in India’s tropical climates to enhance thermal performance. Utilizing the Building Energy Simulation Optimization (BESO) approach, the study integrated the admittance method–based simulation framework with a recently developed metaheuristic optimization algorithm, the grasshopper optimization algorithm (GOA), and assessed its efficacy. Expanding upon a foundational study that initially applied GOA for building thermal performance enhancement, this research introduces a more comprehensive set of key decision variables. A rigorous comparative analysis of GA and GOA was conducted across three major Indian climatic zones, maintaining consistent population size and generations for robust comparison. The results indicate that GOA outperforms GA in minimizing the annual thermal load of air-conditioned buildings by 2.32%–12.37% and reducing computational time by 4.18%–37.11%. Furthermore, the study focused on refining algorithm parameters to enhance performance, employing advanced techniques such as metaoptimization (using auxiliary algorithms for calibration) and dynamic parameter adjustment based on reinforcement learning (RL) principles. These advanced techniques, previously unexplored in the BESO domain, address complex optimization challenges effectively. Fine-tuning parameters with metaoptimization resulted in performance improvements of 0.04%–2.86%, whereas RL-based fine-tuning resulted in improvements of 1.03%–5.97%. Although the efficacy of the optimization framework was demonstrated using case studies in India’s tropical climate, the approach provides a robust guideline for integrating advanced RL techniques and metaheuristic optimization algorithms to design energy-efficient buildings in various other climatic conditions. The holistic consideration of diverse decision variables is identified as a critical factor in sustainable building design. This research has practical implications for the design and implementation of energy-efficient building systems in the real world, with enhanced thermal performance, and contributing to the goal of creating sustainable building solutions. This study introduces a comprehensive building energy simulation optimization approach to optimize building design parameters for enhanced thermal performance in commercial buildings within India’s tropical climate. By utilizing advanced optimization algorithms such as the grasshopper optimization algorithm, significant improvements in minimizing annual thermal loads are achieved, along with computational time savings. Moreover, innovative methodologies for algorithm optimization, including metaoptimization and reinforcement learning–based fine-tuning, lead to additional enhancements in optimization outcomes. These findings offer practical solutions for architects, engineers, and policymakers to design energy-efficient buildings, addressing the urgent need for sustainability in the built environment. This research contributes valuable insights into optimizing building performance and reducing energy consumption, which are crucial steps toward creating environmentally responsible and resilient urban spaces.