Dataset of Energy Consumption and Performance Indicators of Split-Type Air Conditioners under Single and Multiple Fault Conditions in Tropical Island Climates
Description
This dataset contains simulated operational data of split-type air-conditioning systems operating under fault-free, single-fault, and multiple-fault conditions in tropical island climates. The faults considered include vapor-compression system faults and sensor-related faults. The dataset was developed to support research on the impacts of air-conditioning faults on tertiary buildings, as well as fault detection and diagnosis (FDD) of simultaneous faults under different severity levels. The data were generated through building energy performance simulations using OpenStudio and EnergyPlus, with automated simulation workflows implemented in Python. Key variables include climate type, outdoor air temperature, electrical energy consumption, relative energy consumption variation and coefficient of performance (COP) variation. The target variables describe multiple fault types and fault severity levels, categorized as fault-free (0), low degradation (1), medium degradation (2), and severe degradation (3). The dataset enables the analysis and comparison of fault impacts on energy performance under both tropical coastal and tropical highland climate conditions. Researchers may use the dataset to evaluate the impacts of faults on air-conditioning system performance, develop supervised classification models for fault detection and diagnosis (FDD), perform feature selection, apply explainable artificial intelligence (XAI) techniques, and conduct comparative evaluations of FDD methodologies. The dataset is particularly valuable because publicly available HVAC fault datasets representative of tropical island climates remain limited.