Modeling and multi-objective optimal sizing of standalone photovoltaic system based on evolutionary algorithms
The standalone photovoltaic (SAPV) system is one of the most widely used applications of the PV system. However, the main drawbacks of this system are the high capital investment and low energy efficiency mainly due to the simple PV model and the ineffective sizing method of SAP...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2020
|
| Subjects: | |
| Online Access: | http://psasir.upm.edu.my/id/eprint/85650/1/FK%202020%2060%20-%20ir.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The standalone photovoltaic (SAPV) system is one of the most widely used
applications of the PV system. However, the main drawbacks of this system are the high capital
investment and low energy efficiency mainly due to the simple PV model and the ineffective
sizing method of SAPV system. Therefore, precisely modeling method to determine the unknown
parameters of the PV module is essential to give a realistic evaluation for the extracted energy of
the PV array. Moreover, optimization of the standalone PV system is necessary to maximum the
reliability and minimize the total cost of the system in both urban and rural areas.
The research presented in this thesis is divided into two phases, namely, modeling of the PV
module, and optimal sizing of the entire system to obtain reliable and cost- effectiveness SAPV
system. Due to the effective attraction-repulsion mechanism of electromagnetic-like (EM) algorithm
and reliable exploration and exploitation phases of differential evolution (DE), these two methods
were used to determine parameters of the single diode PV model and finding optimal sizing of the
SAPV system. Firstly, an improved EM (IEM) algorithm is presented to estimate the five parameters
of the single PV-module system. The IEM algorithm uses the attraction-repulsion
mechanism to change the positions of solutions towards the optimality. The key to improvement is
performed by adding a nonlinear equation to adjust the length of the particle. Moreover, the total
force formula is simplified to speed up the exploration for an optimal solution. Six statistical
tools are used to show the superiority of the proposed PV model as compared to other models
proposed in the literature. Secondly, the modeling method of the proposed PV module is validated by
experimental data. In the sizing of the SAPV system, the mutation adaptive DE (MADE) algorithm
based multi-objective functions minimizes three constraint objective functions. A new
mutation vector inspired by the two-opposite path (2-Opt) algorithm with adaptive mutation scalar (F ) and crossover rate (CR) control parameters were employed to
enhance the exploration and exploitation phases of the proposed algorithm. The
objective functions are loss of load probability (LLP) and life cycle cost (LCC) and
levelized cost of energy (LCE). In the current sizing optimization problem, the three
individual objectives are normalized, weighted, and then aggregated by a single
function which is minimized to select the optimal configuration of the SAPV system.
Moreover, the performance of the SAPV system is carried out based on three types of
storage batteries which are lead-acid battery, crown battery, and lithium-ion battery
using hourly meteorological data for one year.
Performance results show that the MADE algorithm based on lead-acid battery has a
high level of LLP and minimum cost among other types of storage batteries. The LLP
value of lead-acid is 0.0019 which describes the availability of the proposed SAPV
system. Moreover, the LCC and LCE values are 54895.68 USD, and 1.5803 USD,
respectively. Finally, the proposed sizing method is compared with a numerical sizing
method to show the accuracy and efficiency of the proposed method. The results of
the comparison indicated that the MADE method has an excellent level of accuracy
and outperforms the iterative method in terms of CPU-execution time. |
|---|