You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
One main objective of uesgraphs is to generate modelica code based on an uesgraphs and another is to visualize the results or even analyze them properly. While uesgraphs is a graph representation of for example an heating network, a modelica simulation is an object-oriented approach with pipes. A pipe has furthermore ports, where other pipes, or more precise the ports of other pipes, are connected with. Some sizes are constant over a pipe like the mass flow (extensive sizes) and some are constant on ports (intensive) like pressure. In a graph representation one would assign the mass flow to an edge and the pressure to a node. So while its easy to assign a pipe to an edge its mor complicate to assign a port to a node in a bidirectional heating network. The picture below show the difference when plotting:
Graph representation in uesgraphs
Graph representation in modelica
To map the simulation results properly onto an uesgraphs the first transformation from uesgraph to modelica has to be understood.
Key findings
Before the modelica code is generated a new uesgraphs object is created where nodes and edges are switched. Every edge of the former uesgrpahs becomes a node, what later gets a pipe. Every node is then an edge connecting two nodes (later pipes). The new edges are kind of the blue connection lines in the upper picture
In a uesgraph representation from above that would be:
Activity