マルチエージェント合意制御系における多層パーセプトロンを用いた故障診断システム設計

抄録

In this paper, we designed a fault-diagnosis system using a multilayer perceptron for a multi-agent consensus control system. The proposed system diagnoses the presence or absence of a fault and the faulty agent, based on the fault features called residual signals. The residual generator and the fault-tolerant controller are designed based on the conventional method using H∞ and H- norm of the transfer function about the fault signal. A numerical simulation was performed to confirm the effectiveness of the proposed method.