Location of voltage SAG source by using artificial neural network
Power quality (PQ) is a major concern for number of electrical equipment such as of sophisticated electronics equipment, high efficiency variable speed drive (VSD) and power electronic controller. The most common power quality event is the voltage sag. The objectives are to analyse the voltage sag a...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2016
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/id/eprint/81059/1/KhairulRijalWagimanMFKE2016.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Power quality (PQ) is a major concern for number of electrical equipment such as of sophisticated electronics equipment, high efficiency variable speed drive (VSD) and power electronic controller. The most common power quality event is the voltage sag. The objectives are to analyse the voltage sag and to estimate the voltage sag source location using artificial neural network (ANN). In this project, the multi-monitor based method was used. Based on the simulation results, the voltage deviation (VD) index of voltage sag was calculated and assigned as a training data for ANN. The Radial Basis Function Network (RBFN) was used due to its superior performances (lower training time and errors). The three types of performance analysis considered are coefficient of determination (R2), root mean square error (RMSE) and sum of square error (SSE). The RBFN was developed by using MATLAB software. The proposed method was tested on the CIVANLAR distribution test system and the Permas Jaya distribution network. The voltage sags were simulated using Power World software which is a common simulation tool for power system analysis. The asymmetrical fault namely line to ground (LG) fault, double line to ground (LLG) fault and line to line (LL) fault were applied in the simulation. Based on the simulation results of VD for CIVANLAR distribution test system and the Permas Jaya distribution network, the highest VD was contributed by LLG which were 0.491 and 0.751, respectively. Based on the proposed RBFN results, the best performance analysis were R2, RMSE and SSE of 0.9999, 5.24E-04 and 3.90E-05, respectively. Based on the results, the highest VD showed the location of voltage sag source in the system. The proposed RBFN accurately identified the location of voltage sag source for both test systems. |
|---|