HU-ICT-LAB / RobotWars

Robomaster S1 autonomous lasergaming for HU
12 stars 1 forks source link

Make the simulation multi agent #163

Closed RichardDev01 closed 2 years ago

RichardDev01 commented 2 years ago

Description

As developers, we want this so the policy is trained to compete with ever better getting opponents. This way it will learn to handle more complex strategies of opponents and not exploit a simple never changing opponent.

Why do we need to do this?

This story is needed to handle more complex strategies of opponents and not exploit a simple never changing opponent.

Tasks for this user story

Acceptance criteria

This user story is done when policies get trained in a multi agent environment.