hierachical parameter mapping

[masoud-najafi]

Hi,
In SampleSystemSubSystemDictionaryInline.ssd example, parameters of gains1 are defined in two levels:

1. in the main system level (SampleSystem), i.e. subsystem.gains1.gain1.k=8
2. in the component level, i.e. gain1.k=2*Param1+3 where Parm1=7 is defined via the component parameter bidnings.

My question is whether it is allowed to provide the value of Param1 in the higher level overriding the component level value Param1=7?
For example, in the SampleSystem level can we have a parameter bindings like this:.

prefix="subsystem."
Param1=230
gains2.gain2.k=18

Thanks,
Masoud

[pmai]

Hi Masoud,

as you say, you can provide parameter bindings at a higher level, which override parameter bindings provided for the lower level; however there is no mechanism to override parameter values independently; i.e. if you provide a parameter in a different parameter binding, the parameter mapping taking place in the lower level will not access that parameter source as part of its parameter binding.

So in your example, the Param1 parameter value you provide at top-level will not bind to anything, and will definitely not influence the lower-level parameter binding, i.e. it will not have any effect.

Regs, Pierre.

[masoud-najafi]

In many applications, having a single parameter to control the internal parameters of lower-level FMUs can be highly beneficial. Take, for example, a battery pack composed of 1000 cells, where each cell is modeled by the same FMU. If we want to adjust a parameter uniformly across all cells at the battery pack level, the standard approach in SSP would require defining 1000 individual parameter sets in the parameter bindings which is an impractical and cumbersome process (Unless there is another way in SSP that I am not aware of).

In our inhouse modeling application, we introduce a parameter 𝑃1, which is bound to the corresponding parameters in each internal FMU. This allows us to modify P1 at the system level, effectively propagating the change across all 1000 FMUs. This method greatly simplifies parameter management and enhances model scalability.