Information Technology Reference
In-Depth Information
here h=1,2,3,4,
x
represent sample j,
i
x
is sample j's ith index value. Therefore we
may use variable fuzzy recognition model (14) to calculate synthetic RMD of sample 7.
With Formula (7) we obtain synthetic RMD of each index for flood, after normalizing
them that we get normalized synthetic RMD of each index. Here
w
is the above index
α
weight ; m is number of recognition indexes;
is rule parameter of model optimiza-
tion,
p
is distance parameter.
When taking rule parameter of model optimization
α
=1 distance parameter
p
=1
′
ux
()
and substituting relative data into model (6) we get synthetic RMD
. After
h
j
ux
.Using Formula (7) we get disaster degree of sample 7 as
H=2.046. In the same way, we can calculate the disaster degree values of all the 32
samples as shown in Table 2, and Figure 2 shows the scatter plots of calculated values
of test samples by VFS and projection pursuit (PP) model ([6]). The disaster degree
values agree with the estimated degree values using PP ([6]).
()
normalized it is
hj
Table 2.
The disaster degree values during the 32 years in Henan province
sample degree
value
Judgement
Result
experiential
judgment
sample
degree
value
Judgement
Result
experiential
judgment
x
x
i
i
1
1.415
1
1
17
2.685
3
3
2
1.411
1
1
18
3.5
3
3.5
3
1.343
1
1
19
3.404
3
4
4
1.259
1
1
20
3.823
4
4
5
1.390
1
1
21
3.781
4
4
6
1.5
1
1.5
22
3.644
4
4
7
2.046
2
2.0
23
3.863
4
4
8
1.699
2
2.0
24
1.708
2
2
9
2.189
2
2.0
25
2.277
2
2
10
1.644
2
2.0
26
2.685
3
3
11
1.910
2
2.0
27
2.517
3
3
12
2.5
2
2.5
28
3.556
4
4
13
3.288
3
3.0
29
3.157
3
3
14
2.953
3
3.0
30
3.039
3
3
15
3.468
3
3.0
31
3.532
4
4
16
3.135
3
3.0
32
2.853
3
3
Due to the standard of four grades, so we have([4]):
(a) If 1.0 ≤ H ≤ 1.5,then desertification degree belongs to small (1 grade).
(b) If 1.5 < H ≤ 2.5,then it belongs to medium (2 grade).
(c) If 2.5 < H ≤ 3.5, then it belongs to large (3 grade).
(d) If 3.5 < H ≤ 4, then it belongs to extreme (4 grade).
Search WWH ::
Custom Search