In modern electrical grids, planning and operation processes require efficient optimization tools. Optimal placement and sizing of Flexible AC transmission system (FACTS) devices, renewable energy resources, and energy storage units, to name a few, are optimization tasks in the planning process. Minimizing the cost of generated power from committed generators in the operation process is an important part of power system operations. For power system optimization problems, several optimization algorithms have been proposed and used in the past two decades. However, the need for efficient optimization algorithms customized to power system problems still exists. The research reported in this thesis develops novel evolutionary optimization approaches for two applications: optimal power flow (OPF) and optimal placement and sizing of FACTS to enhance electrical network loadability. For optimal power flow, two new feedback-based self-adaptive differential evolution algorithms are proposed. Prior to applying the proposed methods to the power system test cases, they are tested on standard mathematical benchmark problems. The self-adaptive differential evolution algorithms showed significant improvement in the benchmark problems compared to other algorithms. More importantly, in this work, the feedback-based self-adaptive differential evolution algorithms demonstrated good improvement in results and in convergence rate in several power system test cases. To enhance the loadability of an electrical network, a new multiobjective-based frame work is proposed for optimal placement and sizing of FACTS devices. The proposed method has been applied to commonly used FACTS devices, thyristor-controlled series controllers (TCSCs), and demonstrated excellent results in the electrical loading margins as well as the investment costs compared to other available methods.
Read
