Functions f_i_
_W_i_ and f_i−1__W_i−1__ are optimal return functions per state variables
W_i_ and W_i−1_ in time stages i and i−1 respectfully; Pel_i_
__Qd_i__=return function from stage i per decision variable Qd_i_;
and Eq. _4_ expresses soil moisture constraint; Eq. _5_ expresses
the constraint that the model will disregard cases when precipitation
exceeds evapotranspiration; Eq. _6_ expresses the cost efficiency
constraint _separately elaborated hereinafter_; Eq. _7_
expresses constraint of water abundance from borehole; and Eq.
_8_ defines the initial value of soil moisture, present in Croatian
climate conditions in winter and early spring.
State variable values in observed time stages W_i_ and in previous
time stages W_i−1_, and values of decision variable Qd_i_, as
well as returns per various stages Pel_i_, are calculated in the
course of the process.
Optimal electric power of the PV generator, which is obtained
as an output result, should meet the demands of the consumers
throughout the whole observed period. However, the defined objective
function should not be confused with defining the maximum
area that can be economically irrigated, Amax, because the
optimizing model was developed for the purpose of defining the
optimal size of the PV generator. Further research showed that the
same model could also be
used for defining Amax in the manner
described in the work.