question

W.Ali avatar image
1 Like"
W.Ali asked Jason Lightfoot edited

query about Autonomous vehicle model

Hi,

We have a FlexSim license for educational purposes. It is an impressive tool to create a simulation model.

I am a Ph.D. student, and My research topic is Autonomous Vehicle cyber security. Recently we decided to simulate the V2X network and use machine learning techniques to detect intrusions and attacks on V2X.

I am not sure if we can do this work on FlexSim, so I decided to write here may an expert can confirm that we can do it and suggest some references or examples which we can make as a base for our plan.

I am grateful for any help in this regard.

Thank you

FlexSim 21.2.4
machine learningartificial intelligenceautonom vehicle
· 2
5 |100000

Up to 12 attachments (including images) can be used with a maximum of 23.8 MiB each and 47.7 MiB total.

Jason Lightfoot avatar image Jason Lightfoot ♦♦ commented ·
Just to clarify, are you hoping to simulate the communication network (abstract space) or the vehicle operation and location (2/3D space) in relation to cell towers and other infrastructure or both?
1 Like 1 ·
W.Ali avatar image W.Ali Jason Lightfoot ♦♦ commented ·

Thank you for your reply,

I want to simulate both first vehicle operation and location with other infrastructure also the communication network.


Is there any way to do it?

0 Likes 0 ·

1 Answer

Jordan Johnson avatar image
0 Likes"
Jordan Johnson answered Jason Lightfoot edited

I lean towards this being a possibility. I think FlexSim's Agent module might be a good way to go, for simulating cars that can interact with eachother, and with physical objects. I don't know that anyone has used the Agent Module for this purpose, so I don't know what issues you'd run into. But here's some related documentation:

https://docs.flexsim.com/en/22.0/Reference/Tools/AgentSystem/AgentSystemOverview/

You may be able to use an AStar grid to define a road network, and use Agent to detect or handle collisions. Often, AStar allows for fairly open travel, with a few barriers. I suspect that to model traffic that way, that you'd need to have the reverse; lots of barriers and one-way areas, to restrict path finding. But it could work:
https://docs.flexsim.com/en/22.0/WorkingWithTasks/Travel/WorkingWithAStar/

For the network logic itself, I suspect you could use the Process Flow tool:

https://docs.flexsim.com/en/22.0/ModelLogic/ProcessFlowBasics/OverviewProcessFlowInterface/

Process Flow allows you to lay out logic as a flow chart, and you can watch that logic being executed. It makes it fairly straightforward to visualize how the logic happens.

There are other ways to specify paths in FlexSim, like AGV or NetworkNodes. But I think you'd end up fighting more with them than with AStar. They either prohibit or ignore collisions on their own.

All that being said, this is brand new territory, as far as I'm aware. So you may need to try a few approaches before finding one that works.

· 2
5 |100000

Up to 12 attachments (including images) can be used with a maximum of 23.8 MiB each and 47.7 MiB total.

W.Ali avatar image W.Ali commented ·

Thanks a lot for this useful information and I will try with your suggestions.

What about the machine learning techniques to detect attacks on the network? do you know any examples applied on Flexsim?

0 Likes 0 ·
Jason Lightfoot avatar image Jason Lightfoot ♦♦ W.Ali commented ·

I think this is a large subject for which there is a large volume of material (like this) published on the internet as I'm sure you're aware. ML is relatively new for FlexSim and that fact coupled with us rarely hearing of network simulation being done in FlexSim (at least I don't hear of it) means examples of your application almost certainly don't exist. However we're keen to see your approach and will try to help in any way we can.

1 Like 1 ·