Page 258 - Hybrid-Renewable Energy Systems in Microgrids

Solar–wind hybrid renewable energy system 235

A pilot hybrid solar–wind power generation project has been designed by Yang
et al. based on GA. This project was built to supply power for a telecommunication
relay station from renewable energy sources on a remote island of China [11].
Artificial Neural Network, ANN is a computational model based on biologi-
cal neural networks. It consists of an interrelated group of artificial neurons and
processes information using a connectionist approach to computation [66]. In their
study, Kalogirou [85] proposed an optimization model of solar systems using ANN
and GA. The system is modelled using a computer program called TRNSYS and
the climatic conditions of Cyprus. This study included in a Typical Meteorological
Year (TMY) file. The ANN has been trained using the results of a limited number of
TRNSYS simulations. Afterward, GA has been used to estimate the optimum con-
figurations, for maximizing life-cycle resulting a significantly reduced design time.
Another interactive optimization method has been reported in the literature for
sizing HRES systems is called particle swarm optimization (PSO) [86]. PSO is a
population-based stochastic optimization practice enthused by social behaviour of
bird flocking or fish schooling where bird flocking or fish schooling is the collective
motion of many self-propelled entities [6].
Dehgan et al. [87] suggested a methodology for sizing of a hydrogen based wind/
PV plant for reliability indices by applying a PSO. Wang et al. [88] suggested a modi-
fication for PSO algorithm to develop the multicultural design of the integrated power
generation system. They also conducted sensitivity study to examine the impacts of
different system parameter on the overall design performance.

4.2.4 Graphic construction technique
The graphical construction method is based on satisfaction of average value of demand
by average values of photovoltaic and wind power generation for PV generators and
wind turbines. In this method, only two decision variables are considered in the opti-
mization process. Generally, PV-wind or PV-battery combination is considered in this
method [90,91].
This technique designs the optimum configuration of PV array and battery for a
standalone hybrid PV–Wind system using long-term meteorological data such as solar
radiation and wind speed recorded for every hour of the day for very long years [91].
In some other studies, researchers used the monthly-average of the weather data [92].
For a given load demand and the desired LPSP, the optimum sizing of the HRES PV–
Wind can be achieved if the total cost of the system is linearly related to both the num-
ber of PV modules and the number of batteries [93]. The minimum cost of the system
will be at the point of tangency of the curve representing the relationship between the
number of PV modules and the number of batteries.
A seasonal analysis is made for the demand variation and availability of resources
for the generators during the winter and the summer months. Based on this, a siz-
ing curve is developed between the available various sizes of wind turbines and PV
generators. for a larger number of data, more refined curve can be obtained. In their
work, some authors used a long-term data of solar radiation and wind speed recorded
for every hour of the day for 30 years [41]. Load consumption data of a typical house
in Massachusetts has been fed as the load demand for the hybrid system design. For