Generative Adversarial Networks for Multi-agent Consistency System
Jinrong Ma, Yang Yang, Xiaozhong Qi · 2019 3rd International Conference on Electronic Information Technology and Computer Engineering (EITCE) · 2019
The inconsistency of the states of agents in infinite discrete time domain is a kernel problem that must be addressed. In this study, the optimal strategy of distributed suboptimal controller is proposed under the framework of generating adversarial networks to optimize the state disparity between agents. Firstly, the Nash equilibrium solution is obtained by calculating the game between the generator and the discriminator, which is used as a qualification condition for multi-agent consistency. Then, with the minimum cost function as the performance optimization index, the control vector function is proposed that make the actual output of the system always follows the change of the expected state. Finally, the effectiveness of the proposed suboptimal controller is verified by simulation experiments. The simulation results show that the selected control vector function minimizes the cost function and satisfies the optimal solution for generating adversarial networks framework. At this time, the multi-agent system can not only achieve multi-agent consistency, but also greatly improve the consistency accuracy.