Author

Embaiya Salih

Author Identifier

Embaiya Salih

https://orcid.org/0000-0001-5586-6897

Date of Award

2018

Document Type

Thesis - ECU Access Only

Publisher

Edith Cowan University

Degree Name

Doctor of Philosophy

School

School of Engineering

First Supervisor

Stephan Lachowicz

Second Supervisor

Octavian Bass

Abstract

The stability of power systems has become the most critical issue affecting their quality and performance. Concerns over climate change are now moving the energy sector into a new era of modern power grids. To sustain the reliable operation of power systems and improve the quality of power generation, several instability issues that affect the quality of the operation of power systems are addressed in this thesis. The first is the fluctuations in generated power due to variations in wind velocity in wind power systems. The fluctuation in the wind, which is the main energy source in a wind power system, leads directly to voltage and frequency fluctuations in both generation and load and affects the stability of power systems. In microgrids, a long period of transient, which occurs after the switching operation of microgrid (MG) and load demand changes, is the second issue addressed. The third instability issue is the impact of instability on the power load because of voltage and frequency variations.

Therefore, as a contribution to overcoming the impacts of these instability issues on power systems, this thesis proposes to apply a superconducting magnetic energy storage (SMES)-based neural network (NN) control strategy to enhance the quality of the wind power supply, increase the stability of the MG, and protect critical power loads. In this research, as a method, NNs and the adaptive control method are proposed and applied to control the power flow via its power conversion system. NNs are applied to forecast renewable energy as a short-term prediction of wind power fluctuations. The backpropagation function is used for training the NNs on wind speed variations and then NN is used as a predictive controller. To increase the stability of wind power systems, the proposed SMES-based NN is connected to a wind power system for stabilising wind power fluctuations. SMES-based NN is also applied to an MG to reduce the transient period that occurs after switching of the MG and because of the load demand changes. Furthermore, SMES is operated as an uninterruptible power supply (UPS) to reduce the fluctuations in generated power to protect the critical loads.

In this research, an NN controller is built and trained to predict and track the fluctuations of wind power. This NN controller as a reference model, along with the adaptive control strategy, is implemented and applied as a control system for SMES. The behaviour of the proposed system is verified to decrease the voltage and frequency fluctuations in wind power supply with variations of wind speed. The results show that with a high dynamic response, the proposed NN controller based SMES maintains the voltage and frequency within acceptable limits and stabilises its generating power significantly. Also, the reliability of the SMES-NN for stabilising an MG is investigated. The results verify that the MG was stabilised under the proposed power controller. Moreover, the system’s ability to protect and support the loads during power interruptions and instability events affecting the quality of the supply is tested. The results show that as a short-term storage system, the UPS-SMES efficiently protects the loads by injecting power into the system when it is needed.

Consequently, this research with the developed techniques of reducing instability impacts on power systems could underpin the reliability of renewable power sources as well as supporting and protecting equipment and power loads from such critical issues.

Share

 
COinS