Distributed privacy protection control of multi-agent systems subject to fuzzy dead-zone nonlinearities and sensor faults

Abstract

This article proposes a distributed adaptive privacy-preserving consensus tracking control scheme for leader-follower multi-agent systems with fuzzy dead-zone actuator nonlinearities and sensor bias faults. The core innovations lie in integrating a masking mechanism for privacy protection with a tolerant adaptive method that compensates for fuzzy dead zones without requiring precise knowledge of breakpoints or membership functions. First, unlike traditional dead-zone models that assume known slope boundaries, the adopted fuzzy dead-zone model with center-of-gravity defuzzification effectively captures the physical actuator outputs. Second, distinct from differential privacy methods that sacrifice control accuracy by injecting persistent noise, a vanishing output-masking mechanism is developed. This mask encrypts initial states during the critical transient phase and decays asymptotically, thereby guaranteeing exact consensus tracking with zero steady-state error while preventing information leakage. Third, a coordinate transformation is introduced to incorporate sensor faults and masking dynamics into a prescribed performance function. This innovation ensures that the transient fluctuations induced by the privacy mask are strictly confined within user-defined performance envelopes. A rigorous Lyapunov analysis establishes semi-global boundedness of the consensus tracking error and boundedness of all closed-loop signals. The effectiveness and superiority of the proposed scheme are validated through simulations.