A GP-AHP method for solving group decision-making fuzzy AHP ...
A GP-AHP method for solving group decision-making fuzzy AHP ...
A GP-AHP method for solving group decision-making fuzzy AHP ...
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
C.-S. Yu / Computers & Operations Research 29 (2002) 1969–2001 1997<br />
= −0:083683478 (log a 2 24) − 0:023296643y 2 24;1 − 0:004102164v 2 24;1 − 1:19009d 2 24;2<br />
− 0:904953978d 2 24;3 +0:243037743; (6.25)<br />
where y2 24;1 ¿ log a224 +(v2 24;1 − 1)M, log a224 + d224;2 + d224;3 6 0:17609, 0 6 d224;2 6 0:17609,<br />
0 6 d2 24;3 ¿ 0:17609, a224 ;y2 24;1 ¿ 0, v2 24;1 is a 0–1 variable, M is a big value.<br />
Step 2: In the <strong>for</strong>m of the proposed model 4, Example 5 is <strong>for</strong>mulated as follows:<br />
Minimize<br />
+<br />
⎧<br />
⎨<br />
⎩<br />
⎧<br />
⎨<br />
−<br />
⎩<br />
4�<br />
4�<br />
q=1 q ′ ¿q e=1<br />
4�<br />
3�<br />
3�<br />
q=1 i=1 j¿i e=1<br />
4�<br />
4�<br />
q=1 q ′ ¿q e=1<br />
3�<br />
(log vq − log vq ′ − log aeqq ′ +2 e qq ′)<br />
3�<br />
(log vqi − log vqj − log ae qij +2 e ⎫<br />
⎬<br />
qij)<br />
⎭<br />
3�<br />
(log a e qq ′)+<br />
Subject to (6:1)–(6:7); (6:10)–(6:25);<br />
4�<br />
3�<br />
3�<br />
3�<br />
q=1 i=1 j¿i e=1<br />
(log ae ⎫<br />
⎬<br />
qij)<br />
⎭<br />
log vq − log vq ′ − log ae qq ′ + e qq ′ ¿ 0; log vqi − log vqj − log a e qij + e qij ¿ 0;<br />
( 3<br />
2 ) 6 log a3 5<br />
24 6 ( 2 );vq;vqi;ae qq ′;ae e e<br />
qij; qq ′; qij;de ij;de qij ¿ 0;<br />
(q; q ′ ) ∈{(q; q ′ ) | 1 6 q¡q ′ 6 4}; (i; j) ∈{(i; j) | 1 6 i¡j6 3}:<br />
Step 3: Based on Theorem 1, the obtained M 1 is 2.05645, M 0 is 0.5751544, and value is<br />
31.72898104.<br />
Step 4: By executing the <strong>for</strong>mulated model on LINDO or EXCEL, the acquired solution set<br />
is (log v1 =0:23571, log v2 =0:41353; log v3 =0, log v4 =0:21752, log v11 =0:17592, log v12 =0,<br />
log v13 =0:27616, log v21 =0:47664, log v22 =0, log v23 =0, log v31 =0:47664, log v32 =0,<br />
log v33 =0, log v41 =0:10024, log v42 =0:20049, log v43 =0, v1 =1:72072, v2 =2:59162, v3 =1,<br />
v4 =1:65014, v11 =1:49939, v12 =1, v13 =1:88869, v21 =2:99671, v22 =1, v23 =1, v31 =2:99671,<br />
v32 =1, v33 =1, v41 =1:25963, v42 =1:58667 and v43 = 1), where the average grade of total<br />
membership functions is computed as 0.74416.<br />
Step 5: After calculating normalized the vector V ,<br />
⎡<br />
⎢<br />
[v1;v2;v3;v4] ⎢<br />
⎣<br />
v11 v12 v13<br />
v21 v22 v23<br />
v31 v32 v33<br />
v41 v42 v43<br />
⎤<br />
⎥<br />
⎦