6. Simualation
Results
The current system has variety of units, including wind unit,
photovoltaic unit, fuel cell unit, diesel generator, used in order to
storage electrical power in batteries. A diesel generator has been used
to control the load oscillations and microgrid frequency that its cost
function is of a quadratic equation:
(16)
To make the micro-grid load and power generation units balanced,
upstream grid hourly sell tariff is presented at the point which it
possess the delivery of power from micro-grid to upstream grid. Both
Weibull and normal curves can be applied so as to model the generated
power of wind and solar units and hence normal curve has been used in
this section for power description of solar unit. Uncertainty of wind
power unit generation can be the consequence of disregarding effective
parameters such as moisture and temperature in prediction of wind
velocity. At first, to model this uncertainty, standard deviation of
main scenario is calculated and for obtaining 2 other scenarios, this
amount of standard deviation is added to the main scenario hourly
(scenario1) and by minimizing (scenario3) using the presented method,
three individual scenarios for each solar and wind power generation
units are obtained which defines occurrence probability for the main
scenarios in each unit with upper probability and two generation
scenarios with lower probability where sum of each scenario probability
is equal to one. Nine final states are generated by the combination of
these scenarios. Optimal scheduling for generation costs of units from
the least to the most is precisely performed and in all scenarios,
charge of batteries has been done with accuracy for the least cost and
low load grid at night hours. Sale of micro-grid extra energy to
upstream grid at hours which the tariff is in the highest level shows
that optimal scheduling has been done accurately. In order to guarantee
the micro-grid load at peak hours, the maximum battery discharge is used
and all the constraints of the micro-grid have been also taken into
account. Best scheduling of micro-grid can be done for one day or one
month period, and in this regard the best scheduling for one day and
nine scenarios have been done. Figures 5 to 13 show optimal scheduling
of the nine scenarios where planning and determining the optimal
strategy of micro-grid energy resources by predicting uncertainties
indicates the generation schedule and possible generation intervals of
the system. It can be observed that the battery can be charged in all
scenarios while the discharge can be occurred at specific hours. As the
wind and solar power usually experience fluctuations in their power
outputs, their scheduling has been determined according to the
uncertainties. The intervals for photovoltaic power are predicted to be
between hours 6th to 19th.
Furthermore, utility grid can sell power to the main grid in order to
meet the required load demand at hours 8th to
12th and 19th to
22th. By considering scenarios which are created
through standard deviation, the system operational cost is less compared
to GA, and by applying the presented method, 169 euro is saved in
comparison with GA which emphasizes that the methodology is more
profitable and efficient for cost optimizing.