Expert-Based Method of Integrated Waste Management Systems for Developing Fuzzy Cognitive Map - Complex System Modelling and Control Through Intelligent Soft Computations - page 133

Table 6 The time series
predictions

t 1 ′ t 2 ′ t 3 ′ t 4 ′ Normalized values

C1 0.53 0.68 0.69 0.72 0.74 1.00

-

C2 0.51 0.56 0.62 0.65 0.71 - 0.90

C3 0.48 0.45 0.50 0.52 0.67 0.72

-

C4 0.44 0.33 0.34 0.34 0.61 0.47

-

C5 0.56 0.62 0.70 0.72 0.78 - 1.00

C6 0.46 0.39 0.41 0.42 0.67 0.72

-

Despite the fact that a waste management system consists of only six main
factors, it is obvious now that overviewing the whole procedure properly needs an
approach based on the systems of systems concept (Perusich 2010 ). This approach
is namely suitable to handle problems with essentially different type system com-
ponents where interoperability and seamless interfacing is necessary. The appli-

cation of this approach then easily leads to unexpected emerging phenomena

'

such
as e.g. the surprising values in the resulting connection matrix. The results obtained
by the FCM model are unambiguously such emerging features that will necessarily
lead to re-evaluation of the knowledge and views of experts dealing with waste
management.
While in this approach we tried to

—

find to optimize parameters with the help of
the BEA and thus obtained a single set of results for the connection matrix in an
alternative research (Buruzs et al. 2013b ) we found that results obtained with
various, non-optimal steepness values , the results differed essentially only in the

scaling. All estimated time series predictions converged to essentially the same limit
values (Table 6 ).
Findings from these simulations are not surprising. From the above table
(Table 7 ), it can be stated, that the ranking of the factors in

λ

uencing the sustain-
ability of the waste management systems is similar in both simulations. This is the
way how the roles and weights of the factors should be considered within an IWMS
in order to ensure overall ef

fl

ciency. In the next part, we summarize the results and
give a short overview about the future research intentions.

Table 7 The normalized
values of the two simulations

Normalized values
(FCM simulation)

Normalized values (BEA
optimisation)

C1 0.87 1.00 0.74 1.00

- -

C2 0.78 0.89 0.71 0.90

- -

C3 0.68 - 0.78 0.67 - 0.72

C4 0.64 0.69 0.61 0.47

- -

C5 0.78 0.89 0.78 1.00

- -

C6 0.65 - 0.71 0.67 - 0.72