In this study, the population size and the maximum generation
number of 40 and 25 are used, respectively, and the tolerance is
0.001. In addition, C1 and C2 are 2, and the inertia weight of 0.5
with an inertia weight damping rate of 0.99 is chosen. These control
parameters of the PSO algorithm have been selected according
to the expertise of the previous research works [44–51] and also
based on some pre-tests conducted by the authors to get the best
trade-off between the computational time and the reliability of
the Pareto optimal front. In addition, the evolution of the population
continues as long as at least one of the stop criteria is satisfied,
i.e. the maximum iteration number is reached or the average
change in the spread of the Pareto optimal front becomes lower
than the tolerance.
It should be noted that according to the range of variations of
decision variables, the number of possible configurations of the
building envelope is numerous, while the maximum number of scenarios searched by the PSO algorithm equals to the number of
population size multiplied by the number of maximum iteration,
which is equal to 40 25 = 1000. Therefore, the proposed optimization
approach leads to decrease the required computational
time significantly in comparison to a countless search and therefore,
ensures a saving of computational time.