Dynamic Economic Dispatch For Power System
The research work in this dissertation deals with dynamic economic dispatch problem for large power systems. The work mathematically proves the dynamicity of the economic dispatch. Many physical and operational constraints were considered in the model of the dynamic economic dispatch problem. The pr...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English |
| Published: |
2016
|
| Subjects: | |
| Online Access: | http://eprints.utem.edu.my/id/eprint/18388/1/Dynamic%20Economic%20Dispatch%20For%20Power%20System.pdf http://eprints.utem.edu.my/id/eprint/18388/2/Dynamic%20Economic%20Dispatch%20For%20Power%20System.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The research work in this dissertation deals with dynamic economic dispatch problem for large power systems. The work mathematically proves the dynamicity of the economic dispatch. Many physical and operational constraints were considered in the model of the dynamic economic dispatch problem. The problem is to optimize the total generation costs while satisfying the operational constraints. Through an appropriate utilization of the structural features of the model, a solution algorithm based on Particle Swarm Optimization is developed. The performance of the PSO-based developed algorithm was tested on simple case studies with a small number of generation units and limited constraints, and then on more complex case studies with a large number of variables and complicated constraints. The solution algorithm based on a constraint relaxation and period-by-period is developed and tested. The last part of the dissertation is dedicated to the comparison of solution results obtained by using PSO method and the Dantzig-Wolfe decomposition method for different cases of size and complexity. This research finds large variable size DED problems can be easily implemented, PSO method is reliable and is suitable for real-time analysis. Also, time-segmentation of the solution, or as known as a period by period solution, always results in sub-optimality, while, only by solving the optimization problem in totality can lead to an optimal solution. By modifying constraints, the method can provide alternate solutions to the dispatcher. Trade-offs between the level of convergence to the global solution and the required execution time necessitate finding a mean to enhance the social component and determine an appropriate value that leads to limiting the search space of the swarm. |
|---|