Agriculture Reference
In-Depth Information
The objective weight of
j
th
index was:
θ
j
=(1−
e
j
)
/
i
=
n
(1−
e
j
),
(
j
=1, 2, 3, …,
n
)
It was clear that:
n
0≤
θ
j
≤1;
j
=1
θ
j
=1
This study took the subjective information into the calculations, the comprehensive weight
could be obtained by combining the subjective weight
w
1
, w
2
, w
3
…w
n
of the decision makers
with the objective weight
θ
j
(
j=1,2,3,…,n)
:
α
j
=
θ
j
ω
¯
/
j
=1
n
θ
j
ω
¯
,
(
j
=1, 2, 3, …,
n
)
Recording the optimum value of each row as
r
j
*
, normalize the elements in the matrix, and
r
j
*
value was varied with the index characters. The indexes could be divided into two classes: The
profitable indexes and the damnous indexes, which were “the larger the better” and “the
smaller the better”, listing as follows:
d
ij
=
{
r
ij
r
j
∗
,
r
j
∗
=max
{
r
ij
}
r
j
∗
r
ij
,
r
j
∗
=min
{
r
ij
}
The entropy coefficient value (A better management scheme would obtain a higher entropy
coefficient value) of each treatment could be calculated by:
n
λ
i
=
j
=1
αd
ij
,
i
=1, 2, 3, …,
m
.
The calculated objective weight of the tobacco yield,
IWUE,
Ca content in tobacco leaves and
the cost were 0.2142, 0.2135, 0.2003, and 0.3720. However, since the strict requirement on water
saving, the subjective weight of which was assigned as 0.4, 0.4, 0.15, 0.05 respectively. Fig. 8
showed the entropy weight coefficient of different treatments, based on the principle of
entropy weight coefficient “the higher the better”, P4 was supposed to be the best scheme, in
other words, 30 g/plant MP3005 water-retaining agent combined with EM-calcium spray (‰1
Ca
2+
) during the root-extending stage of flue-cured tobaccos was the optimal management
scheme, and the interval of spraying time was 3 days with 2 mm each time on the back side of
tobacco leaves. Additionally, entropy weight coefficient value of P1 and P4 were similar, but
the mechanism was different, P1 tended to obtain a lower cost, and P4 tended to obtain a higher
yield and
IWUE
.
