Robust Distributed Control of DC Microgrids Considering Heterogeneous Communication Delays

Data exchange among nodes in a distributed dc microgrid (DCMG) causes time delays that can negatively impact reliability and performance. Previous literature often assumes that local data are delayed by the same value as the transmission delay between neighboring nodes, aiming to achieve convergence of average bus voltages to the desired level. However, this assumption relies on having precise knowledge of the communication delay, which is not always feasible in real-world applications. This article presents a more realistic scenario by considering unknown and varying communication delays among links. The assumption of no self-delay for local data results in a steady-state deviation of the global average bus voltages. To address this challenge, a novel observer for average bus voltage is proposed, which significantly expands the acceptable delay range and reduces the communication burden by transmitting fewer variables through the communication links. Furthermore, a modified distributed secondary control scheme is developed, and the system model is transformed into a standard linear form represented by retarded differential equations. Stability analysis and the design of control gains for the system, considering uncertain local and public loads, are presented using the Lyapunov–Krasovskii approach. The proposed method is verified through theoretical analysis, simulation, and experimental results.