Arrowstar Posted February 26 Author Share Posted February 26 32 minutes ago, arbsoup said: Why does the genetic algorithm's "Best" value sometimes (and quite randomly/erratically) increase again? Was it wrong about what the best solution was? It's very disheartening to see MSFS appear to converge on a good value, only to go wildly off and report something with gigantic delta-v cost because "Best" just shot back up. Yeah, I've seen this before. I don't completely know why GA does it but it is odd. Unfortunately I don't have a solution either. Quote Link to comment Share on other sites More sharing options...
Awesomesauce1337 Posted March 1 Share Posted March 1 Using the Porkchop Plot Generator gives a wildly different estimated dV and computed dV. In my case, doing a transfer from Rhode (homeworld, 3rd body around Sun) orbit to Hydrus (2nd body around Sun) gives an estimated departure burn of 945m/s while the calculated burn is 1704m/s. This seems to be coming from a massive extra burn in the normal and radial directions. I am using the Beyond Home planet pack with a 2.5x rescale. The bodies.ini file was generated in the same save as the issues Relevant Images: https://imgur.com/a/zu7gHIz bodies.ini File: https://drive.google.com/file/d/1hiN7sINuSLS24-8Q76xjQX7XSLULROHi/view?usp=sharing Quote Link to comment Share on other sites More sharing options...
Awesomesauce1337 Posted March 1 Share Posted March 1 Further info: the homeworld in Beyond Home (Rhode) does have a rather large moon (Lua) in a relatively close orbit. Could the departure burn calculator be trying to avoid a collision with said moon? Quote Link to comment Share on other sites More sharing options...
arbsoup Posted March 1 Share Posted March 1 (edited) 6 minutes ago, Awesomesauce1337 said: Further info: the homeworld in Beyond Home (Rhode) does have a rather large moon (Lua) in a relatively close orbit. Could the departure burn calculator be trying to avoid a collision with said moon? Probably not. The porkchop tool's departure calculator seems to always be off from the estimate, even for an Earth-Mars transfer with the default bodiesSolarSystem config loaded. MFMS (must've had the flight simulator on the brain last post) will usually produce a better-looking maneuver for the situation (I got a 1.1 km/s departure burn, matching your results with your config), but that's assuming the genetic algorithm doesn't just randomly decide to forget its "best" solution. Edited March 1 by arbsoup Quote Link to comment Share on other sites More sharing options...
arbsoup Posted March 2 Share Posted March 2 (edited) -removed- Edited March 2 by arbsoup had the wrong idea Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.