A finite state based residential demand generator for scarce statistical data scenarios

In many countries detailed mappings of various related issues such as appliance time of use, usage patterns, and human occupancy are not available making the process of generating residential demand profiles for demand management purposes very challenging. In this paper an abstracted residential demand generator (ARDG) based on behaviour finite state diagrams is introduced. The generator takes into consideration seasonal variations and also the effects of socio-economic factors on the emulated demand. The resulting residential demand is abstracted because it is both appliance and appliance-mode independent and owing to a relatively high temporal granularity mimics residential demand with good accuracy without generating a huge volume of superfluous data. The generated demand is used to verify demand side management strategies.