Merge feat json demand into dev-branch for testing in etwin

Instead of a relative path to a CSV file, now the whole time-series can be loaded from the json file. This enables a better integration fo other services and data sources.

All time series are now saved in this format as well. The option to save the file-path to the CSV (even if loaded) is deactivated completely now. If this should be kept as an option, then a check/flag has to be included to the GUI.

Additionally interpolation method was optimized with help of chatGPT, which eases the integration of constant profiles (onyl start and end entry needed)