Esmaeil Ebrahimzadeh

Improved Phasor Estimation Method for Dynamic Voltage Restorer Applications

The dynamic voltage restorer (DVR) is a series compensator for distribution system applications, which protects sensitive loads against voltage sags by fast voltage injection. The DVR must estimate the magnitude and phase of the measured voltages to achieve the desired performance. This paper proposes a phasor parameter estimation algorithm based on a recursive variable and fixed data window least error squares (LES) method for the DVR control system. The proposed algorithm, in addition to decreasing the computational burden, improves the frequency response of the control scheme based on the fixed data window LES method. The DVR control system based on the proposed algorithm provides a better compromise between the estimation speed and accuracy of the voltage and current signals and can be implemented using a simple and low-cost processor. The results of the studies indicate that the proposed algorithm is insensitive to noise, harmonics, interharmonics, and dc offset unlike the LES method, while both methods estimate the phasor parameters within 5 ms. The performance of the control scheme based on the proposed method is evaluated by multiple case studies in the PSCAD/EMTDC environment and experimentally validated based on a laboratory setup.

Efficient approach for harmonic resonance identification of large Wind Power Plants

Unlike conventional power systems where the resonance frequencies are mainly determined by the passive components parameters, large Wind Power Plants (WPPs) may introduce additional harmonic resonances because of the interactions of the wideband control systems of power converters with each other and with passive components. This paper presents an efficient approach for identification of harmonic resonances in large WPPs containing power electronic converters, cable, transformer, capacitor banks, shunt reactors, etc. The proposed approach introduces a large WPP as a Multi-Input Multi-Output (MIMO) control system by considering the linearized models of the inner control loops of grid-side converters. Therefore, the resonance frequencies of the WPP resulting from passive components and the control loop interactions are identified based on the determinant of the transfer function matrix of the introduced MIMO system. The effectiveness of the presented theoretical analysis is validated by time-domain simulations for a 400-MW WPP studied in the PSCAD/EMTDC software environment.

Harmonic Stability and Resonance Analysis in Large PMSG-Based Wind Power Plants

Compared to the conventional power systems, large wind power plants (WPPs) present a more challenging system, where the interactions between the passive elements and the wideband control systems of power converters may result in harmonic instability and new resonance frequencies. Most of researches about harmonic stability focus on small-scale systems, and it has not paid much attention yet to identify the mentioned resonance frequencies. This paper models and analyzes the harmonic instability and resonance frequencies in large permanent magnet synchronous generator (PMSG) based WPPs with full-scale converters, where linearized models of inner control loops of the power converters are considered. A large PMSG-based WPP introduces as a multi-input mulit-output (MIMO) control system, therefore, the stability of the whole power system is analyzed based on the real parts of the poles of the introduced MIMO system and the resonance frequencies are identified based on the element amplitudes of the MIMO matrix. An active damping controller is used to set the poles of the WPP in a desired location in order to mitigate the harmonic instability problems. Multiple case studies are provided to depict that wind turbine connections or disconnections in a WPP, as well as grid impedance variations can affect the harmonic stability and resonance frequencies. The effectiveness of the presented theoretical analysis is validated by time-domain simulations of a 400-MW WPP in PSCAD/EMTDC software environment.

Modeling and identification of harmonic instability problems in wind farms

In power electronics based power systems like wind farms, the interactions between the inner control systems of the power converters and the passive components may lead to high frequency oscillations, which can be called harmonic instability. In this paper, a simple methodology is presented to identify harmonic instability problems in wind farms, where many wind turbines, cables, transformers, capacitor banks, shunt reactors, etc, typically are located. This methodology introduces the wind farm as a Multi-Input Multi-Output (MIMO) control system, where the linearized models of fast inner control loops of the grid-side converters are considered. Therefore, instability problems of the whole wind farm are predicted based on the poles of the introduced MIMO system. In order to confirm the effectiveness of the proposed analytical approach, time-domain simulations are performed in the PSCAD/EMTDC software environment for a 400-MW wind farm. The proposed analytical analysis method and time-domain simulation results show that both dynamics of the power electronic converter and the parameters of the passive component can effect on the wind farm stability.

An Adaptive Least-Error Squares Filter-Based Phase-Locked Loop for Synchronization and Signal Decomposition Purposes

Without any doubt, phase-locked loops (PLLs) are the most popular and widely used technique for the synchronization purposes in the power and energy areas. They are also popular for the selective extraction of fundamental and harmonic/disturbance components of the grid voltage and current. Like most of the control algorithms, designing PLLs involves a tradeoff between the accuracy and dynamic response, and improving this tradeoff is what recent research efforts have focused on. These efforts are often based on designing advanced filters and using them as a preprocessing tool before the PLL input. A filtering technique that has received a little attention for this purpose is the least-error squares (LES)-based filter. In this paper, an adaptive LES filter-based PLL, briefly called the LES-PLL, for the synchronization and signal decomposition purposes is presented. The proposed LES filter can be understood as an adaptive complex-coefficient filter because its implementation involves cross couplings between orthogonal axes. The stability of designed LES-PLL is analyzed by the derivation of its small-signal model. Some control design guidelines are also presented. The effectiveness of proposed PLL structure is finally evaluated using experimental results.

Multi-objective optimization of large wind farm parameters for harmonic instability and resonance conditions

In large wind farms, the mutual interactions between the power converter control systems and passive components may result in harmonic instability and resonance frequencies at a various frequency range. This paper presents an optimized parameter design of the power converter controllers in large wind farms in order to reduce the resonance probability and guarantee harmonic stability. In fact, a general multiobjective optimization procedure based on the genetic algorithm is proposed to set the poles of the wind farm in a desired location in order to minimize the number of the resonance frequencies and to improve the harmonic stability. Time-domain simulations of a 400-MW wind farm in the PSCAD/EMTDC environment demonstrate the effectiveness of the proposed design technique.

Performance improvement of DFIG-based wind farm using multilevel cascaded H-bridge converter under unbalanced grid voltage conditions

This paper presents a new configuration of doubly fed induction generator (DFIG)-based wind farm with multilevel cascaded H-bridge (CHB) converter and its control system, which is suitable not only for unbalanced grid voltage conditions, but also for small voltage sags. The proposed configuration of DFIG can balance the point of common coupling (PCC) voltage without needing additional DC link capacitor or energy storage unlike other configurations and methods. Consequently, the stator voltage and the other loads voltage connected to the PCC are kept constant at the nominal value. The new configuration improves all the DFIG signals under unbalanced grid voltage and small voltage sag conditions unlike other methods. Some of the DFIG units in wind farm require to connect the CHB converter, therefore, it is cost-effective. Matlab/Simulink is used for simulation of a 13.5-MW wind farm consisting of nine 1.5-MW DFIG units and the results demonstrate the good efficiency of the proposed scheme.

A modulation technique for four-leg voltage source inverter without needing the look-up table

In this paper, a modulation technique is proposed for four-leg voltage source inverter which is simpler than other techniques. Unlike previous techniques, the proposed technique needs no switching vectors determination and duty ratio calculation. In this technique, turn-on time interval of upper switches, in every switching period, is calculated without needing the look-up table and complex equations. As a result, it is very simple to implement and produce the dead time. Several Simulations in PSCAD/EMTDC environment confirm the validity and advantages of the proposed technique. Experimental results by using DSP-TMS320F2812 are also provided to confirm the validity of the proposed technique.

A novel model predictive control for single-phase grid-connected photovoltaic inverters

Single-phase grid-connected inverters with LCL filter are widely used to connect photovoltaic systems to the utility grid. Among the existing control schemes, predictive control methods are faster and more accurate but also more complicated to implement. Recently, the Model Predictive Control (MPC) algorithm for single-phase inverter has been presented, where the algorithm implementation is straightforward. In the MPC approach, all switching states are considered in each switching period to achieve the control objectives. However, since the number of switching states in single-phase inverters is small, the inverter output current has a high Total Harmonic Distortions (THD). In order to reduce this, this paper presents an improved MPC for single-phase grid-connected inverters. In the proposed approach, the switching algorithm is changed and the number of the switching states is increased by means of virtual vectors. Simulation results show that the proposed approach lead to a lower THD in the injected current combined with fast dynamics. The proposed predictive control has been simulated and implemented on a 1 kW single-phase HERIC (highly efficient and reliable inverter concept) inverter with an LCL filter at the output.

Robust MPC-based Current Controller against Grid Impedance Variations for Single-Phase Grid-Connected Inverters

Recently, LCL filters have been widely used in the output of single phase inverters. Since, the grid side inductor in these filters is in series with the grid impedance at the Point of Common Coupling (PCC), it may create new resonances. This phenomena may take the control loop toward instability. In this case, in order to have a reliable operation, the current controller should be insensitive to the grid impedance variation. In order to damp these resonances, researchers have presented some methods using active or passive damping. These methods added an extra loop to the control loop, an extra passive component in the filter or extra sensor in the control process. But in most of them, the complexity and the cost of controller have been increased. Therefore, presenting a simple control method without extra sensor, passive component or extra arrangement can be a promising approach. This paper presents an MPC-based current controller, which is simple and robust against the grid impedance variation and even the variation of the LCL filter parameters. In contrast to classical multi-loop controller like Proportional-Resonant (PR) controllers, the proposed control method does not need any parameter tuning. In the proposed controller, the switching plan and duty cycles are determined by a cost function and a switching table. Therefore, at the same time with any variation in grid impedance, the proposed controller changes the next switching state and duty cycle. Operating performance like look-up table, searching in all possible switching states to find the best state for the next switching period, makes the controller adaptive and robust against the variation of LCL filter parameters. In order to confirm the effectiveness of the proposed controller, simulations and experimental results of the proposed controller are compared with a classical PR controller.