Optimal location and size estimation of distributed generators by employing grouping particle swarm optimization and grouping genetic algorithm
Distributed Generators is being implemented in the distribution network to improve the performance of the network by reducing the real and reactive power losses, while improving voltage profile during the operation. These advantages could be confirmed and accomplished if optimal size distribution ge...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2017
|
| Subjects: | |
| Online Access: | http://psasir.upm.edu.my/id/eprint/67919/1/FK%202018%2052%20IR.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Distributed Generators is being implemented in the distribution network to improve the performance of the network by reducing the real and reactive power losses, while improving voltage profile during the operation. These advantages could be confirmed and accomplished if optimal size distribution generation units are installed at the optimal location in the distribution network. Otherwise, the problem of generation may increase when the DG is located in non-optimal location and size which can lead to increase the real and reactive power losses and high voltage deviation. Therefore, there are various algorithms that could be applied in order to integrate distribution generation units into the distribution network. These algorithms can be enhanced to increase their efficient and effective. This work is aimed to decrease the total real and reactive power losses while enhancing the voltage profile of the distribution network with less computation time by proposing two new artificial intelligence algorithms, i.e. grouping particle swarm optimization algorithm and grouping genetic algorithm. These two algorithms are compared to their original artificial intelligence algorithms, i.e. particle swarm optimization algorithm and genetic algorithm. These algorithms are used to obtain the optimal size of distributed generators units to be installed at optimal locations, which are obtained using loss sensitivity factor. Multi-objective function is the summation of three indices that are considered in these algorithms, i.e. real power loss index (PLI), reactive power loss index (QLI), and cumulative voltage index (CVD), and implemented on an IEEE 30-bus test system. This is to test the performance of the four artificial intelligence algorithms by taking into consideration the installation of 5 distributed generators units in the bus test system. It was observed that the grouping particle swarm optimization algorithm has achieved high reduction of total real and reactive power losses, by installing five distributed generators, when compared to the base case. It was observed that the voltage profile was improved with the shortest computation times demonstrated in determining the optimal size (global optimal) for the multi DG units that were installed in the IEEE 30-bus test system comparing to all the four artificial intelligence algorithms that were employed in this work. Also, it is found that case of installing four distributed generators units is not much difference from case of five distributed generators which is make it optimum to be installed in the IEEE 30-bus test system in terms of reducing the real and reactive power losses. This also relates to the stability of the voltage. |
|---|