I’m wondering how the weighted graph will work in the multi-agent system. Is that from the software/algorithm side to compute from the server or by the agent itself? Thanks.
Hi Jiarong, thanks for your question! The graph here describes the topology of the information exchange network within the system and helps to determine whether it is possible to propose a controller for the system. For now, we would consider it static, or pre-defined by us, so it is not calculated by algorithms.
We may consider a varying network in the further progress of the project but haven’t touched on it by now. What I can tell so far is the graph cannot be calculated by the agents as they are not able to gain a full view of the system, thus some centralized algorithm “above” the agents have to take care of it.
Hi Binghong,
I’m wondering how the weighted graph will work in the multi-agent system. Is that from the software/algorithm side to compute from the server or by the agent itself? Thanks.
Hi Jiarong, thanks for your question! The graph here describes the topology of the information exchange network within the system and helps to determine whether it is possible to propose a controller for the system. For now, we would consider it static, or pre-defined by us, so it is not calculated by algorithms.
We may consider a varying network in the further progress of the project but haven’t touched on it by now. What I can tell so far is the graph cannot be calculated by the agents as they are not able to gain a full view of the system, thus some centralized algorithm “above” the agents have to take care of it.
Hope this can answer your question!
https://drive.google.com/file/d/1N4EwQEmpgyI3L3zEB_KFBbrO6it2qQ2Q/view?usp=sharing
Here is the link for an optimized version of this poster. You may have a look at it if possible.