Section: New Results
Qualitative approaches for building energy management
Participant : Patrick Reignier.
Reducing housing energy costs is a major challenge of the 21st century. In the near future, the main issue for building construction is the thermal insulation, but in the longer term, the issues are those of renewable energy (solar, wind, etc.) and smart buildings. Home automation system basically consists of household appliances linked via a communication network allowing interactions for control purposes. Thanks to this network, a load management mechanism can be carried out: it is called distributed control. An optimal home energy management system is still a goal to aim for, because lots of aspects are still not completely fulfilled. Most of the energy systems respect only the energy needs, but they don't tackle the user needs or satisfaction. Energy systems also have a lack when it comes to the dynamicity of the environments (the system ability to adapt).The problem is similar for the existing HMI (Human User Interface) of those Home Automation Systems where only experts can understand the data coming from the sensors and most important, the energy plan coming from management system (How? and Why?). The goal of this study is to propose a house energy model that can be both used to predict at some level energy evolution and that can be understood by the end user. The house energy model is based on Fuzzy Cognitive Maps representing cause-effects relations. It is first designed by an expert and then automatically tuned to a particular house using machine learning approaches. Preliminary experiments have been done this year using the Predis datasets.