Tianhong Zheng

Thermoelectric (TE) heating, ventilation, and air-conditioning (HVAC) systems represent a sustainable alternative to conventional vapor-compression technologies; however, their adoption is limited by relatively low coefficients of performance (COP) and challenges in performance optimization. This study presents an integrated experimental, modelling, and deployment framework for predicting and evaluating TE HVAC performance. A custom-built test rig was developed to generate experimental data under controlled laboratory conditions, capturing key thermodynamic variables. A total of 487 samples were used to benchmark multiple regression and machine learning models. Among the evaluated approaches, Linear Regression achieved the best performance, with a root mean squared error (RMSE) of 3.28 and a coefficient of determination (R²) of 0.885, outperforming more complex models. Beyond statistical evaluation, model predictions were validated through recomputation of heat transfer rate and COP, demonstrating strong thermodynamic consistency and physical interpretability. The validated model was subsequently implemented within a Shiny-based graphical interface, enabling interactive scenario analysis and real-time performance evaluation. The results demonstrate that, for moderate-scale experimental datasets, simple and interpretable models can outperform more complex approaches while maintaining physical consistency. The proposed framework provides a practical pathway for integrating experimental data, predictive modelling, and decision-support tools in thermoelectric HVAC applications.

With the increasing global demand for cooling and the rising environmental concerns associated with conventional cooling systems, solar-powered cooling technologies have emerged as a sustainable alternative. This study provides a comprehensive review of recent advancements in solar-powered cooling systems, highlighting the latest technological developments, challenges, and future directions. The paper classifies solar cooling technologies into two primary categories: Solar electric cooling and solar thermal cooling. It discusses various configurations, including PV-assisted vapour compression systems, solar absorption and adsorption cooling system, and hybrid approaches that integrate thermoelectric and thermal cooling mechanisms. The findings indicate that while solar cooling systems offer significant energy savings and carbon emission reductions, several technical, economic, and operational challenges remain. Key issues include the intermittency of solar energy, efficiency limitations of current photovoltaic and thermal conversion technologies, and high initial investment costs. Moreover, the effectiveness of different solar cooling methods varies depending on climatic conditions and building types, necessitating further research on system optimisation and scalability. Future research should focus on improving photovoltaic efficiency through advanced materials such as perovskite-silicon tandem cells, enhancing thermal storage systems, and integrating artificial intelligence-based optimisation for real-time energy management. Besides, hybrid cooling approaches that combine passive and active strategies, along with emerging energy storage solutions, could further enhance the viability of solar cooling technologies. By addressing these challenges, solar-powered cooling systems can play a crucial role in reducing energy consumption in the building sector and contributing to global sustainability goals.

This paper presents an analysis of an off-grid solar cooling system designed for rural healthcare facilities in Malaysia, where the electricity infrastructure is often unreliable, affecting the delivery of essential healthcare services. Aiming to minimize the reliance on electrical battery storage, the system utilizes a DC-powered vapor compression cooling unit directly driven by photovoltaic (PV) solar panels with energy storage using thermal energy storage. Meanwhile, electrical battery storage was designed sufficiently for auxiliary electrical equipment. In addition, the system was designed to maintain indoor air temperatures within a comfortable range as per MS 1525– 2014 and DOSH 2010 standard. In conducting the feasibility study of the proposed idea, we have designed and installed a 5-kW cooling system at a test room facility in Malaysia. The performance of the system was monitored over various conditions, and the results show that indoor temperatures were kept between 22 °C to 26 °C, even when external temperatures reached 35 °C. Additionally, levelized cost of cooling analysis, and a simple payback period indicates that cost per kWh of cooling is only 0.033 USD and less than 2 years respectively. This feasibility study not only demonstrates the technical viability of solar-powered cooling in rural healthcare settings but also highlights its potential for broader application in similar off-grid regions, contributing to sustainable energy solutions in the healthcare sector.

Poor indoor air quality (IAQ) and inadequate thermal comfort continue to challenge healthcare facilities in Malaysia, especially in rural areas where poor ventilation and high humidity are prevalent. These environmental stressors negatively impact patient outcomes, staff well-being, and operational efficiency. Existing solutions often overlook the integration of subjective human perception with objective sensor data, resulting in limited adaptability and responsiveness. This study introduces a datadriven, human-centric framework that combines environmental sensor measurements with user-reported comfort feedback to develop predictive models for IAQ and thermal comfort. A comprehensive machine learning pipeline was implemented using four algorithms—Random Forest (RF), XGBoost, Artificial Neural Network (ANN), and Support Vector Machine (SVM)—to model both continuous IAQ indicators and categorical thermal preferences. Experimental results show that Random Forest achieved the best overall performance, with the lowest root mean squared error (RMSE = 14.35) in regression and the highest classification accuracy (87.5%) in predicting thermal preference. Statistical validation confirmed that Random Forest and XGBoost performed similarly in regression, while Random Forest and SVM showed no significant differences in classification accuracy. These findings validate Random Forest as a robust and consistent model across both tasks. This study contributes a validated, AI-enhanced framework for intelligent environmental monitoring in healthcare settings, emphasizing the integration of subjective and objective data streams. The approach supports personalized, data-driven interventions and offers practical insights for facility managers, clinicians, and policymakers aiming to optimize indoor conditions. Future work will focus on scaling the dataset across diverse facilities and climates, enabling real-time deployment, and incorporating explainable AI techniques to enhance model transparency and stakeholder trust.