Reducing electricity costs in a dynamic pricing environment

Smart Grid technologies are becoming increasingly dynamic, so the use of computational intelligence is becoming more and more common to support the grid to automatically and intelligently respond to certain requests (e.g., reducing electricity costs giving a pricing history). In this work, we propose the use of a particular computational intelligence approach, denominated Distributed W-Learning, that aims to reduce electricity costs in a dynamic environment (e.g., changing prices over a period of time) by turning electric devices on (i.e., clothes dryer, electric vehicle) at residential level, at times when the electricity price is the lowest, while also, balancing the use of energy by avoiding turning on the devices at the same time. We make this problem as realistic as possible, by considering the use of real-world constraints (e.g., time to complete a task, boundary times within which a device can be used). Our results clearly indicate that the use of computational intelligence can be beneficial in this type of dynamic and complex problems