Smart Spaces: Occupancy Detection using Adaptive Back-Propagation Neural Network

In recent years, the concept of smart spaces and occupancy detection is crucial for creating more efficient and sustainable buildings that prioritize both energy efficiency and occupant comfort. By leveraging advanced technologies and real-time data analysis, building managers can optimize energy consumption, reduce operating costs, and improve the overall user experience of the space. One important aspect of this optimization process is the detection of occupancy in smart buildings, which can help to control various systems like Heating, Ventilation, Air Conditioning (HVAC) and lighting. Accurate occupancy detection is a key factor in reducing energy consumption and cost optimization for building control systems, particularly in the areas of HVAC and lighting. To this end, Adaptive Back-Propagation Neural Network (ABPNN) algorithms is applied to a dataset generated by CO2, humidity, temperature, and light sensors to enhance the accuracy of occupancy detection in smart spaces.In this study, the ABPNN based Bayesian Regularization (BR) algorithm is found to have the highest precision rate, with 70% (5701 samples) of the dataset used for training, 15% (1221 samples) for validation, and 15% (1221 samples) for testing. Furthermore, the BR algorithm demonstrated an impressive accuracy rate of 99.8% when compared to other algorithms. Smart spaces with advanced sensors and BR algorithms, such as the BR algorithm, accurately detects occupancy and optimize systems like HVAC and lighting to improve energy efficiency and occupant comfort. This reduces energy consumption and operating costs, while providing insights for optimizing space usage and improving overall building performance.