PhD defence - Maher Abdelkhalek Azzouz AbdelkhalekExport this event to calendar

Friday, September 25, 2015 — 9:00 AM EDT


Maher Abdelkhalek Azzouz Abdelkhalek


New Control Algorithms for the Robust Operation and Stabilization of Active Distribution Networks


Ehab El-Saadany


The integration of renewable distributed generation units (DGs) alters distribution systems so that rather than having passive structures, with unidirectional power flow, they become active distribution networks (ADNs), with multi-directional power flow. While numerous technical, economic, and environmental benefits are associated with the shift toward ADNs, this transition also represents important control challenges from the perspective of both the supervisory and primary control of DGs. Voltage regulation is considered one of the main operational control challenges that accompany a high penetration of renewable DG. The intermittent nature of renewable energy sources, such as wind and solar energy, can significantly change the voltage profile of the system and can interact negatively with conventional schemes for controlling on-load tap changers (OLTCs). Another factor is the growing penetration of plug-in electric vehicles (PEVs), which creates additional stress on voltage control devices due to their stochastic and concentrated power profiles. These combined generation and load power profiles can lead to overvoltages, undervoltages, increases in system losses, excessive tap operation, infeasible solutions (hunting) with respect to OLTCs, and/or limits on the penetration of either PEVs or DGs. With regard to the dynamic control level, DG interfaces are typically applied using power electronic converters, which lack physical inertia and are thus sensitive to variations and uncertainties in the system parameters. Grid impedance (or admittance), which has a substantial effect on the performance and stability of primary DG controllers, is nonlinear, time-varying, and not passive in nature. In addition, constant power loads, such as those interfaced through power electronic converters, are also characterized by inherited negative impedance that results in destabilizing effects, creating instability and damping issues.

Motivated by these challenges, the research presented in this thesis was conducted with the primary goal of proposing new control algorithms for both the supervisory and primary control of DGs, and ultimately of developing robust and stable ADNs. Achieve this objective entailed the completion of four studies:

Study#1: Development of a coordinated fuzzy-based voltage regulation scheme with reduced communication requirements.
Study#2: Integration of PEVs into the voltage regulation scheme through the implementation of a vehicle-to-grid reactive power support strategy.
Study#3: Creation of an estimation tool for multivariable grid admittance that can be used to develop adaptive controllers for DGs.
Study#4: Development of self-tuning primary DG controllers based on the estimated grid admittance so that stable performance is guaranteed under time-varying DG operating points (dispatched by the schemes developed in Study#1 and Study#2) and under changing grid impedance (created by network reconfiguration and load variations).

As the first research component, a coordinated fuzzy-based voltage regulation scheme for OLTCs and DGs has been proposed. The primary reason for applying fuzzy logic is that it provides the ability to address the challenges associated with imperfect information environments, and can thus reduce communication requirements. The proposed regulation scheme consists of three fuzzy-based control algorithms. The first control algorithm was designed to enable the OLTC to mitigate the effects of DGs on the voltage profile. The second algorithm was created to provide reactive power sharing among DGs, which will relax OLTC tap operation. The third algorithm is aimed at partially curtailing active power levels in DGs so as to restore a feasible solution that will satisfy OLTC requirements. The proposed fuzzy algorithms offer the advantage of effective voltage regulation with relaxed tap operation and with utilization of only the estimated minimum and maximum system voltages. Because no optimization algorithm is required, it also avoids the numerical instability and convergence problems associated with centralized approaches. An OPAL real-time simulator (RTS) was employed to run test simulations in order to demonstrate the success of the proposed fuzzy algorithms in a typical distribution network.

The second element, a vehicle-to-grid reactive power support (V2GQ) strategy, has been developed as a means of offering optimal coordinated voltage regulation in distribution networks with high DG and PEV penetration. The proposed algorithm employs PEVs, DGs, and OLTCs in order to satisfy the PEV charging demand and grid voltage requirements while maintaining relaxed tap operation and minimum curtailment of DG active power. The voltage regulation problem is formulated as nonlinear programming and consists of three consecutive stages, with each successive stage applying the output from the preceding stage as constraints. The task of the first stage is to maximize the energy delivered to PEVs in order to ensure PEV owner satisfaction. The second stage maximizes the active power extracted from the DGs, and the third stage minimizes any deviation of the voltage from its nominal value through the use of available PEV and DG reactive power. The primary implicit objective of the third stage problem is the relaxation of OLTC tap operation. This objective is addressed by replacing conventional OLTC control with a proposed centralized controller that utilizes the output of the third stage to set its tap position. The effectiveness of the proposed algorithm in a typical distribution network has been validated in real time using an OPAL-RTS in a hardware-in-the loop (HiL) application.

The third part of the research has resulted in the proposal of a new multivariable grid admittance identification algorithm with adaptive model order selection as an ancillary function to be applied in inverter-based DG controllers. Cross-coupling between the and grid admittance necessitates multivariable estimation. To ensure persistence of excitation for the grid admittance, sensitivity analysis is first employed as a means of determining the injection of controlled voltage pulses by the DG. Grid admittance is then estimated based on the processing of the extracted grid dynamics by the refined instrumental variable for continuous-time identification (RIVC) algorithm. Unlike nonparametric identification algorithms, the proposed RIVC algorithm provides a parametric multivariable model of grid admittance, which is essential for designing adaptive controllers for DGs. HiL applications using OPAL-RTS have been utilized for validating the proposed algorithm for both grid-connected and isolated ADNs.

The final section of the research is a proposed adaptive control algorithm for optimally reshaping DG output impedance so that system damping and bandwidth are maximized. Such adaptation is essential for managing variations in grid impedance and changes in DG operating conditions. The proposed algorithm is generic so that it can be applied for both grid-connected and islanded DGs. It involves three design stages. First, the multivariable DG output impedance is derived mathematically and verified using a frequency sweep identification method. The grid impedance is also estimated so that the impedance stability criteria can be formulated. In the second stage, multi-objective programming is formulated using the -constraint method in order to maximize system damping and bandwidth. As a final stage, the solutions provided by the optimization stage are employed for training an adaptation scheme based on a neural network (NN) that tunes the DG control parameters online. The proposed algorithm has been validated in both grid-connected and isolated distribution networks, with the use of OPAL RTS and HiL applications.

EIT building
Room 3145


  1. 2021 (104)
    1. December (3)
    2. November (4)
    3. October (5)
    4. September (4)
    5. August (12)
    6. July (4)
    7. June (5)
    8. May (10)
    9. April (25)
    10. March (14)
    11. February (10)
    12. January (9)
  2. 2020 (248)
    1. December (20)
    2. November (20)
    3. October (16)
    4. September (15)
    5. August (16)
    6. July (32)
    7. June (29)
    8. May (32)
    9. April (27)
    10. March (13)
    11. February (20)
    12. January (16)
  3. 2019 (282)
  4. 2018 (150)
  5. 2017 (212)
  6. 2016 (242)
  7. 2015 (242)
  8. 2014 (268)
  9. 2013 (190)
  10. 2012 (31)