Stochastic Consensus-Based Control of $\mu$Gs With Communication Delays and Noises

Abstract

We propose a novel stochastic consensus-based secondary voltage and frequency restoration scheme with communication delays and noises for islanded microgrids (µGs). Existing distributed methods commonly design restoration layer with the assumption of ideal communication among distributed generations (DGs). Albeit, the communication channels are exposed to delay and noise, whereas any DG receives delayed and noisy measurements from its neighbors due to peripheral noises and communication delays. Delay and noise have a great impact on the control of µG, which terribly reduce the stability and quality of it. To eliminate the adverse effects of delays and noises, we propose novel consensus protocols that consider effects of the communication noises and delays simultaneously for complete plant dynamics, and study mean square consensus for frequency and voltage restoration of µGs whereas providing stringent real power sharing. To this end, we derive the mean square consensus restoration proof using rigorous Lyapunov analysis. As a result the suggested method decreases the sensitivity of the system to failures and increases its reliability. Finally, we have done several simulation scenarios in MATLAB/SimPowerSystems Toolbox to verify the proposed strategy performance.

Robust distributed stochastic secondary control of microgrids with system and communication noises

“…in which I and 0 are identity and zero matrices, K is the gain to be designed in (21), and ⊗ represents kronecker product. In the following, we claim that Ξ = lim t → ∞ E[π(t)π(t) T ] exists and is obtained from the following Lyapunov equation:…”

“…Proof: See equations (3)- (21). □ The main feature of our proposed distributed protocol is that the designed protocols are based on the relative estimated state, under which the consensus error can be characterised using the Lyapunov equation's solution.…”

“…So far, many types of distributed controllers have been proposed to solve the MGs problems such as disturbances, uncertainties, unmodeled dynamics, noises, faults, and time delay in the secondary layer by researchers. For example, in [16,20,21], authors present distributed stochastic and/or distributed time-delay methods for restoration of frequency/voltage of MG. Unmodelled dynamics, uncertainties, and disturbances' effects on the secondary restoration layer of an islanded MG have been investigated in [15,16,22,23]. The adverse effects of time delays in the network communication for islanded MGs in the secondary layer are discussed by a small-signal model in [24][25][26][27].…”

An accurate current sharing method for an Islandeddroop‐controlled dcmicrogrid without communication between remote buses

“…14 Several research studies on "noise-resilient" control strategies of microgrids have been reported. 14,[50][51][52][53][54] Additive noise is statistically modeled with "Gaussian distribution". 14,[50][51][52][53][54] Without any loss of generality, we have modeled noise with a "zero-mean white Gaussian" model.…”

“…14 White noise with a variance of 0.1 is considered for the simulation. 14,50 The time response with a communication delay (of 30 ms) and white noise ("zero-mean white Gaussian" with a variance of 0.1) is shown in Figure 15. We observe that the system performs satisfactorily even with additive noise.…”