srijan - Stardance

Ship Changes requested

@srijan on 3d Graph Search · 23 days ago

I wrote an algo to try and travers a 3d graph space as quickly and efficently as possible with 3 UAVs. The simulation was provided by Autonomous Teaming Solutions with a baseline of a random search. I implemented node weights, frontier nodes, Dijkstra, zoning, and communication between searchers to ensure that we don't have redundant movement

1 devlog
5h

Try project → See source code →

Open comments for this post

@srijan on 3d Graph Search · 28 days ago

5h 7m 43s logged

This is a simulated search optimization algo. It consists of two phases: observation & surveillance. We have 3 simulated UAVs, our “searchers”. They follow a greedy algo. First, the 3D Graph is split into zones based on which drone is where. Next, the searchers identifiy “Frontier Nodes”, or the nodes bordering observed & unobserved nodes and utilize a Dijkstra algo to reach them. This repeats until we have a threshhold of 90%, where the algo then knows the total map and must cover it once more as efficiently as possible. This is the surveillance phase.