Always on param math changes

[atiselsts]

When optimizing a model with always-on parameters, the parameters are included in the ranges as usual.

The problem: for example, if there are 2 always-on parameters, and the parameter selection strategry is configured as:

"parameters": [{
	"type": "exhaustive",
	"range": [1, 2]
}]

then only one optimization job will be created. If there are 3 always-on parameters, then with the above config no jobs will be created.

This PR changes this behavior to effectively use this dynamic parameter selection range instead:

 range: [1 + number_of_always_on_parameters, 2 + number_of_always_on_parameters]