Distributed Weight-Average-Prediction Control and Stability Analysis for an Islanded Microgrid With Communication Time Delay

Abstract

This paper studies the communication time-delay issues in islanded microgrids (MGs) with the distributed secondary control architecture. Firstly, a time-delayed MG smallsignal model is developed. Then, a new weight-average-prediction (WAP) controller is proposed to compensate the delayed system state. By introducing a time-delayed differential term in the proposed control law, the traditional time-delayed small-signal model is transformed into a neutral time-delayed mathematic model. Based on the developed model, the stability analysis is conducted considering both fixed time delay and time-varying delay. For the fixed time delay, a novel graphic analytical method is proposed to evaluate the time delay margin, which eliminates the conservatism compared with existing time-domain methods. For the time-varying delay, stability condition is established by a Lyapunov-Krasovskii function and linear matrix inequalities. In addition, some non-linear WAP control methods are discussed to guide the parameter tuning with a higher resolution. Lastly, the proposed method and analytical result are verified in the OPAL-RT real-time test platform. The results demonstrate the effectiveness and high performance of the proposed controller.

Robust fuzzy model predictive control for voltage regulation in islanded microgrids

“…Theorem 2. For the considered μG representation in (16), the set (B1) is an RPI set with respect to the corresponding feedback control law 𝜋 t , if there exists a positive definite matrix X t , general matrices Y a , Y b , Y l , Z, G b and two positive scalars 𝜓, 𝜆 ∈ ℝ (0,1) , such that the following three matrix inequalities are feasible [33]:…”

“…Toward this end, Theorem 2 in Appendix 2 is developed to obtain a sequence of robust constraint sets. (16) with respect to the corresponding feedback control law 𝜋 e , if there exists a positive definite matrix X e , general matrices Y a , Y b , Y l , Z, and two positive scalars 𝜉 e ,𝜆 ∈ ℝ (0,1) , such that the following matrix inequalities are feasible:…”

“…Lemma 1. The set Ω t in (B1) is an RPI set for the considered μG in (16), if there exists a positive scalar 𝜆 ∈ ℝ (0,1) such that:…”

“…Then, the global T-S fuzzy system using Equations ( 15) and ( 16) can be written as follows: (16) where…”

“…However, this method was not able to handle any communication delays. Authors in [16] proposed a weighted average predictive control in order to improve consensus convergence and strong anti-interference ability; however, this method suffers from uncertainties, and the random delay was not considered. Authors in [17] proposed a Smith predictor controller that can handle both fixed delay and ran-dom communication delays.…”