Date: Sep. 12, 2017, 8:15am-10am Pacific times

The purpose of the meeting was to discuss the FMI extensions required for QSS efficiency.

LBNL proposal (https://lbl-srg.github.io/soep/softwareArchitecture.html#proposal-of-lbnl) was discussed.

Few points needed clarification. These points will be discussed below

Higher Order Derivatives

Changes to the element Derivatives will be highlighted.

State Events

• The event indicator handler will be removed. Instead, the dependencies list of state derivative, will be extended to include event indicator variables.

• The dependencies list of event indicator variables will include

  • input variables
  • continuous time state variables
  • event indicator variables

• The new proposed smooth token data type will be removed. The fmiSetSmoothToken and fmi2GetSmoothToken will be replaced by fmi2SetRealInputDerivatives and fmi2GetRealOutputDerivatives. These two functions exist in the FMI for co-simulation API and will be used for QSS as well. Discussion remains to clarify whether FMU-QSS should be an FMU for model exchange which uses co-simulation functions or an FMU for co-simulation. LBNL will look at the Hybrid co-simulation proposal to see whether this API could be a better fit for FMU-QSS.

  Hubertus mentions that the hybrid co-simulation proposal will most likely not be in the next version of the standard.

  Michael mentions that it would be good to compare our proposal with the hybrid co-simulation proposal to make sure we are not building an incompatibility that makes merging them later hard.

• Thierry and all to look at the hybrid co-simulation proposal
• Thierry to revise LBNL proposal and include the proposed changes