Distributed cooperative learning control for multiagent systems with data protection and disturbance observation

This paper presents a distributed cooperative learning control strategy that incorporates data protection and a disturbance observer. Specifically, an encryption-decryption mechanism is proposed through the construction of mask functions and command filtering. This mechanism not only protects agent information from being exposed but also reduces the impact on system performance caused by the mask function.Additionally, a self-correction mechanism with an amendment function is proposed for the disturbance observer, addressing the issue of error over-compensation. Furthermore, neighbor information is incorporated into the adaptive law to enhance the robustness of system. The designed trigger condition accounts for the effects of trigger errors and system tracking performance, effectively conserving communication resources while maintaining system performance. Finally, the simulation results validate the theoretical findings.