Page 363 - Decision Making Applications in Modern Power Systems

Heuristic methods for the evaluation of environmental impacts Chapter | 12 325

Step-4: Initialize the population, P t .
Step-5: Create a young population or descendants Q t of the current popu-
lation P t
Step-6: Combine the two populations Q t and P t to form R t where
R t 5 P t , Q t .
Step-7: Find the nondominated Pareto fronts F i andR t .
Step-8: Start the new population P t11 5 0 and the count for inclusion
i 5 1.
Step-9: While P t11 1 F i # N pop do: P t11 ’P t11 , F i , where i’i 1 1:
Step-10: Order the last front F i using the distance agglomeration in des-

cending order and choose the first elements N pop 2 P t11 ofF i .
Step-11: Use the selection of operators, crossover, and mutation to cre-
ate the young population or the descendants of the new population
Q t11 .

12.3.3 Analysis and discussion of results

The solution report presents the input parameters to run the program, such as
the energy demand, the minimum and maximum power of the engines and
the results of the total cost of fuel, total power loss, and optimal power for
each machine in the plant to meet the load demand.
Table 12.12 shows the results of the case study of the plant located in the
city of Manaus (first case study). These results were obtained after the exe-
cution of the program for a power demand of 20 MW.
As can be seen from Table 12.12, there is a certain difference between
the levels of emission of generators, and the power demand is distributed
among all generators with lower values assigned to the generators 2 and 10.
It can also be seen that the power is not always the maximum power that is
related to the maximum emission.
Table 12.13 shows the results for the case study of the IEEE test system
[75]. These results were obtained after the execution of the program for a
1036 MW power demand, which is the power between the maximum and
minimum power of this system. This system has 10 units.
As is shown in Table 12.13, there are some differences between the emis-
sion index of the engine test system.
This is mainly due to the power difference between the engines of this
system. This also leads to different emission index of generators.
Fig. 12.7 shows the trade-off between emission index and the fuel cost of
the first case study after the application of NSGA-II, generated by
MATLAB.
Fig. 12.8 shows the trade-off between emission index and the fuel cost of
the test system IEEE 118-bars after applying NSGA-II, generated by
MATLAB.