Closely Cooperative Multi-Agent Reinforcement Learning Based on Intention Sharing and Credit Assignment | IEEE Journals & Magazine | IEEE Xplore