Dataset of Dual-Fan Air Purifier Control System for Under-Actuated Zones
Description
This dataset contains synchronized multi-sensor recordings collected from a dual-fan air purifier system operating in an under-actuated indoor zone. The system consists of two fans whose behavior is influenced jointly by environmental conditions, occupancy activities, airflow dynamics, and zone characteristics. Since airflow is not independently controlled for each sub-area, the system exhibits dynamic variations that are important for research on intelligent ventilation, HVAC control, and adaptive indoor climate modeling. The dataset includes 4,800 records and 57 features, capturing environmental, occupancy, mechanical, and contextual information used to model and predict dual-fan control increments for intelligent air purifier and HVAC systems. Data Acquisition Location: Universitas Trilogi Library, Jakarta Environment Type: Under-actuated zone with three sub-areas Sampling Interval: 5 minutes Duration: 4 weeks of continuous monitoring Sensors Used: - CO₂, VOC, PM₂.₅ - Temperature, humidity, dew point - Air pressure, airflow volume - Occupancy counts (sitting, standing, walking, etc.) - Mechanical variables (fan speed, ventilation split ratio) All sensor readings were time-synchronized and validated before integration.
Files
Steps to reproduce
1. Download the dataset. The dataset contains both original sensor-based variables and derived features. 2. Inspect dataset structure. The file includes 57 columns and several thousand rows. Each row corresponds to a single sampling interval, representing environmental, airflow, feedback, and safety conditions in an under-actuated zone. 3. Pre-processing: handle missing or blank cells, normalize continuous features, encode categorical features, and optionally compute cyclic time encodings 4. Organize features into five reasoning channels: Humidity and Thermal Conditions (HUM-TH), Indoor Air Quality (IAQ), Occupancy and Activity (OCC), Airflow and Pressure Dynamics (FLOW-PRS), Zone and Temporal Context (ZONE-TEMPORAL) 5. Organize results in tabular form (Excel) for further analysis or machine learning applications.
Institutions
- Universitas Trilogi