WATER CONSUMPTION TECHNIQUES FROM THE FLOW DYNAMICS IN RESIDENCES
Disaggregation, flow, identification, embedded systems.
The research, of the experimental type, aims at the development of technique of separation of water consumption in a residence, which equates simplicity, efficiency and adaptability to low cost installations. Currently, studies and procedures aimed at monitoring the use of water in homes are small, inaccessible and their results can not reach, for example, a low-income consumer. This project consists of the development of a simple technique that can be incorporated into microcontrolled systems, capable of studying, discharging (distinguish) loads and identifying patterns of water use. During development, it was decided to use simple methodologies to acquire the signal that reflects the displacement of fluid inside the pipe towards the outputs, thus analyzing the return pattern of the pulses emitted by a hall effect sensor resulting from this flow; thus being able to identify and translate this shift into digital information. We called this the "digital fingerprint of each point", which will later be applied in real time in an algorithm that was inspired by a second-order polynomial regression, in conjunction with individual flow filters to institute differentiations. Studying and determining the intrinsic characteristic and each type of signal, requiring little information and low processing power; achieving as an end product an extremely simple, replicable and low cost solution, which potentially allows the creation and comercialization of an affordable product for low income consumers.