Multi-empirical Discriminant Multi-Agent Reinforcement Learning Algorithm Based on Intra-group Evolution
Zhonglei Zhang, Jiaoling Zheng, Chang-jie Zou · Journal of Physics Conference Series · 2020
Abstract In order to find the optimal target in the unknown environment more quickly, a multi-empirical discriminant multi-agent reinforcement learning algorithm based on intra-group evolution is proposed based on the deep deterministic policy gradient algorithm (DDPG). In the unknown environment, the agent can find the target faster by performing Information exchange, group genetics and other mechanisms. The comparison with the traditional reinforcement learning algorithm shows that the proposed algorithm is superior to the traditional reinforcement learning algorithm in solving time and solving accuracy, and can quickly and effectively find the optimal target in the environment. CCS Concepts •Computing methodologies → Multi-agent systems