Hello. I plan to create my own project with a swarm of e-puck robots.
I generate a field on which there are buildings (n pieces). Buildings are marked with flags and are also randomly generated (at some points).
Naturally, this happens on a regular field, on which e-pucks can drive.
I need to make sure that each e-puck selects the next task (the closest one, but this is not important in principle) and goes to it and stops at the point with the task.
Difficulties I can't solve:
1) e-pucks initially only know the coordinates of the tasks (x;y). they don't know anything about the obstacles. How to write the logic of moving to a point in this case ?(the function (pathplanning) of min path search, which is built into coppeliasim, is not offered - it looks for a path for a very long time!!!).
+ e-pucks should disassemble tasks and perform them simultaneously on the same field, of course)
2) it is not clear how e-puck stops exactly at the desired point coordinate (x;y). After all, the movement of robots is calculated through the "program ticks". and the robot can pass the destination point.
I hope I have clearly stated the essence of the question. I really appreciate your help, which is so much needed!!
Swarm of e-pucks / movement to the point
Re: Swarm of e-pucks / movement to the point
maybe you could try potential fields?
Re: Swarm of e-pucks / movement to the point
this method is known to me. but how to implement it for a swarm and not for 1 robot
Re: Swarm of e-pucks / movement to the point
Hello,
have a look at this simple path planning task for a mobile, 2-wheeled robot. This scene will be also part of next release's example scenes (in scenes/pathPlanning/mobileRobotPathPlanning.ttt). What path planning algorithm you use with what parameters (e.g. search timeouts, search range, etc.) is entirely up to you.
Cheers
have a look at this simple path planning task for a mobile, 2-wheeled robot. This scene will be also part of next release's example scenes (in scenes/pathPlanning/mobileRobotPathPlanning.ttt). What path planning algorithm you use with what parameters (e.g. search timeouts, search range, etc.) is entirely up to you.
Cheers
Re: Swarm of e-pucks / movement to the point
Thank you very much for the answer ! I tried about six months ago to use these functions . But it didn't work out very quickly. Maybe I used too much path smoothing... I'll try again then.
Re: Swarm of e-pucks / movement to the point
The algorithm is distributed, the code of each robot is the same. This video may help, good luck!
https://www.youtube.com/watch?v=oAK-0o02dgI
https://www.youtube.com/watch?v=oAK-0o02dgI
Re: Swarm of e-pucks / movement to the point
Hello. Thanks for the answer. The author has a good project. But he does not attach the source code to it (demo version only). Do you have something similar?