Parameter-Adaptive Distributed Model-Predictive Control for Islanded AC Microgrids: Privacy-Preserving Perspective
Ziqiang Wang, Lidan Zhou, Linyun Xiong, Haosen Yang, Jie Wang, Sunhua Huang
Abstract
Information disclosure is a potential challenge in the practical implement of distributed algorithm and wireless communications on islanded AC microgrid (MG) for better flexibility and low cost. It can be solved by the initial-value-around and real-time-values privacy preservations against internal curious distributed generations (DGs) and external cyber attackers respectively. However, three problems should be solved before the existing information privacy-preserving mechanisms implemented on control level: (i) Encryption algorithm for real-time-values privacy will exacerbate the delay problem; (ii) When the delay problem is solved by the existing distributed model predictive controls, the decaying disturbance for initial-value-around privacy will impact the stability of islanded AC MG; (iii) The compromise problem between the privacy-preserving performance and transient performance of distributed model-predictive control (DMPC) is unacceptable. For the above problems, this paper proposes a parameter-adaptive DMPC (PA-DMPC), where two decaying functions are designed to adjust parameters in terminal constraints and optimal cost function for the third problem. The optimal control results are obtained by solving time-varying quadratic programming (QP) problem. The real-time simulation platform based on NI-PXI (National Instruments-PCI eXtensions for Instrumentation) and Wireless Local Area Network (WLAN) is built to validate its effectiveness and advantages.