In another question I presented my developed methodology to add noise to a time series of electricity loads and exchange prices. With these noisy time series I want to test how well my system can cope with predicted time series (Function for testing system stability, which receives predicted time series as input).
RobertBaron pointed out that my methodology only adds "white noise" (i.e. normally distributed noise) to a time series.
Hence the question: What other types of noise are typically added when predicting electricity loads and exchange prices?
