Note that all the input parameters were divided into 2 classes, deterministic
and non-deterministic. The former ones were identified with
their average values and the latter ones were represented as fuzzy
sets. Not all the parameters were represented as non-deterministic
forms due to one or several of the following reasons: 1) they have
been found not sensitive enough to remarkably affect the model solutions;
2) their variation ranges are rather small, potentially suggesting
their low uncertainty; and 3) an increase in the number of nondeterministic parameters would lead to enhancement in computational cost.