A student project on fuzzy inference systems applied to smart grids. Chained FIS (fuzzy inference system) of 10 FIS
This AI - Expert system - choose the best measures among a list in a V2G-G2V smart grid system. Multi Agent system and consensus best decision with fuzzy logic.
Some advantages of this approach is the low computational cost (compared to one super FIS) for complex systems with a lot of variables. The intermediary variables aren't deffuzzyfied, so we don't lose any informations during the process. A drawback is that some important interactitions are maybe lost in the process of chained FIS. Moreover, the entry, with our approach can be fuzzy. It well suited to for example noised sensors or noised environment. Globally the approach og fuzzy logic is well suited to noisy environment because it intresecally take into account these noises making these systems robust from simulation to real world.
For more explanation, you can read the report (in french) explaining the method and presenting some use cases.