Using an example of a wall erection that is included in the execution
problem, the rating procedure of several construction alternatives
is demonstrated. Due to the huge number of construction
processes for the entire second story, the procedure is shown on
a single building component. Nevertheless, it is valid in the same
manner for the entire construction. Two possible construction
alternatives were considered for the construction of a reinforced
concrete wall. In the first alternative, A, the wall is build of field
concrete that is mixed on-site. The second wall erection alternative,
B, consists of self-compacting concrete that is transported to
the site from a concrete factory. The super-ordinate planning
objective is to keep the fixed completion date.
Each of the construction alternatives consist of a different number
of tasks. The decision to realize exactly one alternative will be
made after the comparison of these two construction alternatives
with the criteria execution time, material costs, and error sensitivity
that are defined with regard to the planning objective. The first
two criteria are definite and their total rating values are calculated
by summation and maximal path algebra respectively. The error
sensitivity is modeled as an indefinite criterion by applying fuzzy
values.
Five different fuzzy ratings values that are stored in the internal
knowledge base are specified for this evaluation criterion: very
low, low, middle, high, and extremely. The fuzzy values and associated
membership functions are specified manually and are
shown in Fig. 16.The rating values of the construction tasks and
the total ratings of the alternatives are shown in Table 2. The concrete
transportation time is not considered in the rating, since the
material is ordered to deliver it on-site just-in-time. This work
does not cause time effort. Material costs emerged for the considered
tasks by material consumption, hire charges, or using/building-
in of supporting material. The total rating of error sensitivity is calculated by the averaged aggregation of all task ratings after
defuzzification using the centre-of-gravity method for each fuzzy
value.