Possibility to define multiple parameter value sets

Describe the feature

For some scenario variation use-cases it is not possible to define a distribution for the parameter values with the given possibilities. E.g. if you want to vary the start position of the ego vehicle, the different coordinate sets containing road id, lane id, s/t offset are dependent on each other and only valid as a complete set. It doesn´t make sense to vary each parameter independently with discrete values in a deterministic distribution.

Describe the solution you would like

Include "parameter value sets" as separate third option additional to DeterministicDistribution and StatisticDistribution in the distribution file (easiest solution). 1...* parameter sets are possible. Each parameter set has two attributes: "parameterName" and "value".

Describe alternatives you have considered

Extend the Deterministic/StatisticDistribution, so that dependencies in terms of parameters sets can be defined (more complex but maybe more flexible to combine with other distributions).

Describe the backwards compatibility

Compatible because new element is added.

Additional context

If you want to measure the KPI of your ADS in terms of accident occurrance / km you would need to drive a lot of km (virtually). Several concrete scenarios on different roads are needed to ensure representative test data. Therefore the road network and the start position of the ego vehicle need to be varied. To avoid having to define multiple concrete scenarios, one logical scenario with a parametrized road network link and ego vehicle start position can be defined. The variation, which is defined in the distribution file then has the above mentioned requirements.