A Smart Positioning System for Personalized Energy Management in buildings

Abstract

Energy consumption in official buildings contributes to 42% of total energy generation in India. The key features in commercial buildings are usage of high energy consuming devices, long duration of usage of electrical equipments, large population density and large equipment density compared with the floor area usage in houses. Hence this problem has motivated to perform research on energy management in official buildings. The individuals in these buildings mostly have unique authority on most of the equipments they handle, and they have their own comfort level requirement based on the context and the equipment availability. Therefore, to devise an effective energy management solution it is required to consider personal requirements with highest priority than the community requirements. Hence in this research work we design and develop systems & solutions needed for Personalized Energy Management (PEM). Our proposed system is developed to capture the spatio-temporal data of context and electrical usage pattern for each individual with bare minimum sensors. To address this challenge, we proposed a smart positioning system (SPS) for personalized energy management. In SPS, we have developed an Real time Smart Positioning System (RSPS) algorithm for integrating electrical map and sensing coverage of electrical appliances inside a building to position the individual in real-time with respect to each of the electrical appliances. Using SPS, the current position of an individual inside the building is determined along with the position of nearby electrical appliances to automate the appliance usage. This is performed using the proposed RSPS algorithm where real-time mapping of electrical map, sensing coverage of nearby equipments, signal strength, and pattern of individual requirements are used to control usage of equipments related to individual's choice. Experimental analysis of the RSPS algorithm on our prototype has been performed and the results showed that it requires a minimum of 2 coverage and it is not required to have 3 coverage as in other localization algorithms. Under the above condition of 2 coverage this algorithm was able to achieve an accuracy of 90%.