Idea: Use only transition probabilities instead of load profiles for appliances, which do not have particular load profiles (e.g. light, TV...)

[GJawana]

Problem: with the current methodology we assume that every appliance remaines switched on for a given periode and is then switched off (at least for 15 mins)
Idea: use only transition probabilities for those appliances, which do not have particular load profiles
but still inhomogeneouse markov chain (time dependent probabilities)
--> transition probabilities would be needed