Legacy documentation

Operation modes

The legacy, MatLab-based routine offers two operation modes:

  1. Markovian chain-based sampling

    Having a distinct household with several appliances, the time-discrete simulation proceeds in time and switches on different appliances based on the switch-on probabilities found in Markovian chains. If a device is switched on, a pre-defined time series for the appliance is replayed. The totality of time-series is then aggregated afterwards.

  2. Probability weighted aggregation

    Once again, this mode is as well time-discrete. Based on the given switch-on probability, the rated apprent power of an appliance is weighted with this probability and the totality of consumed power is accumulated.

Refactoring note

For the refactoring, we will foremost focus on operation mode 1 and maybe add 2 later.