(1)解:
min??15y1?7y2s.t.2y15y1y1
(2)解:
?4y2?3y2?10?3 ?5 y1,y2?0max??6y1?8y2s.t.3y1?5y1y12y1?y2?y24y2?2y3?3y3?7y3?y3?5y3?3?2?4 ?0?2 y1?0,y2?0,y3无限制
解:例3原问题
minz?x1?x2?x3?x4?x5?x6s.t.x1?x2?70x2?x3?60x3?x4?50x4?x5?20x5?x6?30x6?x1?60xj?0,j?1,?,6对偶问题:
max??70y1?60y2?50y3?20y4?30y5?60y6s.t.y1?y6?1y1?y2?1y2?y3?1y3?y4?1y4?x5?1y5?y6?1yj?0,j?1,?,6
解:
(1)由最优单纯形表可以知道原问题求max,其初始基变量为x4,x5,最优基的逆阵为
B?1?1???21????6?0??。 1??3??1由P32式()()()可知b??Bb,Pj??B?1Pj,?j?cj?CBPj?,j?1,?,5,其中b和Pj?aj1??b1??都是初始数据。设b???b??,Pj???a?,j?1,?,5,C??c1,c2,c3?,则
?2??j2??1??1b??Bb??21????6?1??1Pj??BPj??21????65??1?5?0??b???b??212?b1?512?????,解得? ??5,即?11??15bb?10?2??b1?b2??????2??3?32?2??61???0??a01??a12a13?112???,即 ????1??1aaa?2223?0??1????213?2???1?2a11?0?11?a11?0??a11?a21?1?a?33?6?21?1a?1??2122?a12?1,解得? ?111a??1??a12?a22???2232?6?a13?2?1?a?1??213?a23?1?11??a13?a23?03?6?1??j?cj?CBPj????4,?4,?2???c2,0,0???c3,c1??21????2121?6?0??,即 1??3?11?c?c??2232c1??4?c2??2?1??1??c3?c1??4,解得?c3?10
6?c?6?21?1??c??21??3所以原问题为:
maxz?6x1?2x2?10x3s.t.3x1x2?x2?2x3?x3?5?10
x1,x2,x3?0对偶问题为:
min??5y1?10y2s.t.y12y1
*(2)由于对偶问题的最优解为Y?CIB??IB??c4,c5????4,?5???4,2?
3y2?y2?y2?6??2 ?10 y1,y2?0 解:
cj? -5 5 13 0 0 CB 5 0 XB b? 20 10 100 x1 -1 16 0 x2 1 0 0 x3 3 -2 -2 x4 1 -4 -5 x5 0 1 0 x2 x5 (1)因为x3的检验数c3?5?3?0,所以c3的可变范围是c3?15。c3由13变为6在可变范围内,所以最优解不变,目标函数值减少?c3?x3??13?6??0?0,即目标函数最优值不变。
(2)因为x2是基变量,所以c2的变化会引起所有变量检验数的变化。根据最优性准则,有:
??1??5???c2??013???3?13??c2?3??0,解得c2的可变范围是?c2?5,c2由5变为,在可变范围内,
3???0??c?1??02?4所以最优解不变,目标函数值减少?c2?x2??5?4.5??20?10,即最优值为90。 (3)从最优单纯形表可以知道最优基的逆阵为
B?1?10?????41??,要保证基变量的值非负,即要求: ???20??10??20?b2?80,B?1?即?80?b2?0,所以b2的可变范围为b2?80。??41????b?????b???0,
??2??2??第二个约束条件的右端项由90变为80,在可变范围内,所以最优基不变。
?10??20??20?T*?1?20????X?0,20,0,0,0x?20,x?0???????,即,最优解为,b?B?????25???????80???41??80??0?最优值为100。
(4)从最优单纯形表可以知道最优基的逆阵为
B?1?10?????41??,要保证基变量的值非负,即要求: ???b1?0?b1??10??b1?45?????,即解得,所以b1的可变范围0?b?B?1???0?1?90???41??90?2??4b1?90?0??????为0?b1?45。第一个约束条件的右端项由20变为30,超过了可变范围,所以最优基变2化。需用单纯形表重新计算。
cj? -5 5 13 0 0 CB 0 0 XB b? 30 90 0 x1 -1 12 -5 x2 1 4 5 x3 3 [10] 13 x4 1 0 0 x5 0 1 0 ?i 10 9 x4 x5 cj? -5 5 13 0 0 CB 0 13 XB b? 3 9 117 x1 x2 x3 0 1 0 x4 1 0 0 x5 ?i x4 x3
(5)解:最优单纯形表如下:
从最优单纯形表可以知道最优基的逆阵为
?10???1??0??x???,的系数列向量由变为B?1??1??41??12??5??,根据公式可以计算出最优单纯形表
???????1?中x1的系数列向量为P1?BP1????4??10??0??0????5?????5??,重新计算检验数,结果如下: 1??????0 0 cj? -5 5 13 CB 5 0 XB b? 20 10 0 x1 0 5 -5 x2 1 0 0 x3 3 -2 -2 x4 1 -4 -5 x5 0 1 0 x2 x5 因为x1的检验数为-5<0,所以原最优解不变,最优值不变。