Frequency Adaptive Parameter Estimation of Unbalanced and Distorted Power Grid

Abstract

Grid synchronization plays an important role in the grid integration of renewable energy sources. To achieve grid synchronization, accurate information of the grid voltage signal parameters are needed. Motivated by this important practical application, this paper proposes a state observer-based approach for the parameter estimation of unbalanced three-phase grid voltage signal. The proposed technique can extract the frequency of the distorted grid voltage signal and is able to quantify the grid unbalances. First, a dynamical model of the grid voltage signal is developed considering the disturbances. In the model, frequency of the grid is considered as a constant and/or slowly-varying but unknown quantity. Based on the developed dynamical model, a state observer is proposed. Then using Lyapunov function-based approach, a frequency adaptation law is proposed. The chosen frequency adaptation law guarantees the global convergence of the estimation error dynamics and as a consequence, ensures the global asymptotic convergence of the estimated parameters in the fundamental frequency case. Gain tuning of the proposed state observer is very simple and can be done using Matlab commands. Some guidelines are also provided in this regard. Matlab/Simulink based numerical simulation results and dSPACE 1104 board-based experimental results are provided. Test results demonstrate the superiority and effectiveness of the proposed approach over another state-of-the art technique.

Coordinate Transformation-Free Observer-Based Adaptive Estimation of Distorted Single-Phase Grid Voltage Signal

“…non-smooth jump of phase, frequency, DC offset and amplitude. In the traditional state space observer design [20]- [22], by considering y = x 1 , andẏ = x 2 , the dynamics of the grid voltage signal in the state-space form is given by:…”

“…various variants of least-square estimation [12], [13], adaptive notch filter (ANF) [14], various variants of delayed signal cancellation (DSC) techniques [15]- [19], linear observer [20]- [23], harmonic oscillators (linear and nonlinear) [24], [25], various variants of phase-locked loop (PLL) [26]- [30], demodulation-based techniques [31], [32], open-loop techniques [33], [34], etc.…”

“…Moreover, online computation of matrix inverse is also required. Linear observers proposed in [20]- [23] use similar dynamic model as used by Kalman filter. However, linear observers also uses coordinate transformation.…”

“…Moreover, the model explicitly considers DC offset. Secondly, thanks to the novel state-space model, no need of coordinate transformation unlike [20]- [23]. The inverse of a n × n matrix requires n 3 multiplications/divisions and n 3 − 2n 2 + n additions/subtractions operations [36].…”

Enhanced Frequency Adaptive Demodulation Technique for Grid-Connected Converters

“…State-space observer [13], [15], [16] use a state-space model of the grid voltage as opposed to parameterized model used by regression-based techniques. State-space observer can either use fixed gain (e.g.…”