Privacy-Preserving Bipartite Consensus in Aperiodic Sampled-Data Multiagent Systems: A Discrete Differential Privacy Scheme

This article investigates the bipartite consensus control problem for aperiodic sampled-data multiagent systems (MASs) withε-differential privacy. First, a discrete-time time-varying state space model is constructed for heterogeneous MASs, in which the sampling interval is time-varying within a certain known range, which can better describe the segmented continuum features caused by aperiodic sampled-data. Second, a differential privacy average output bipartite consensus algorithm for sampled MASs is given. This dynamic distributed algorithm safeguards the confidentiality of its initial configuration while minimizing the need for constant controller adjustments and real-time data exchanges. A time-varying control gain is given to ensure that the convergence point’s expectation precisely matches the average of the initial states, with the convergence accuracy bounded by an explicit expression. Furthermore, a carefully designed noise mechanism is introduced in the privacy analysis, enabling agents to flexibly adjust their privacy levels. To validate the proposed approach, a multiagent collaborative control experimental platform is applied, incorporating both uncrewed ground vehicles (UGVs) and quadcopter uncrewed aerial vehicles (UAVs). The developed methodology effectively computes the required control gain, with experimental results conclusively demonstrating the successful implementation of our algorithm.