0

在下面的工作示例中,我试图检索匹配的模型,在这种情况下有两个令人满意的模型:

 t1= cl7          t2= cl4      t3= cl5

 t1= cl4          t2= cl3      t3= cl9 

问题是重复匹配的模型,直到运行求解器超时。如何在不重复的情况下检索满意的模型。

非常感谢,

        S, (cl1, cl2, cl3, cl4, cl5, cl6, cl7,cl8,cl9) = EnumSort('S', ['cl1', 'cl2', 'cl3', 'cl4', 'cl5','cl6' ,'cl7', 'cl8', 'cl9'])



       s = Solver()

       x = Const('x', S)
      def fun1(x):
        return Or(x == cl1, x == cl2, x == cl3, x == cl4, x == cl5, x == cl6, x == cl7, x == cl8, x == cl9)


       y1, y2 = Consts('y1 y2', S)
      def fun2(y1, y2):
          return Or(And(y1 == cl7, y2 == cl4), And(y1 == cl4, y2 == cl3), And(y1 == cl3, y2 == cl2),And(y1 == cl8, y2 == cl9))


      q1,q2 = Consts('q1 q2', S)
      def fun3(q1,q2):
        return Or(And(q1 == cl7, q2 == cl5), And(q1 == cl4, q2 == cl9))



 t1, t2 ,t3 = Consts('t1 t2 t3', S)

 s.add(fun1(t1))
 s.add(fun2(t1,t2))
 s.add(fun3(t1,t3))


while s.check() == sat:
s.add(Or(t1 != s.model()[t1], t2 != s.model()[t2],t3 != s.model()[t2]))
print s.model()
4

1 回答 1

2

您似乎在阻止相同模型在新的可满足性检查中再次使用的部分中有一个小错字,特别是倒数第二行应该有t3 != s.model()[t3]而不是t3 != s.model()[t2].

这是生成您在开头放置的两个模型的更新示例(z3py 链接:http ://rise4fun.com/Z3Py/AxJv ):

S, (cl1, cl2, cl3, cl4, cl5, cl6, cl7,cl8,cl9) = EnumSort('S', ['cl1', 'cl2', 'cl3', 'cl4', 'cl5','cl6' ,'cl7', 'cl8', 'cl9'])

s = Solver()

x = Const('x', S)
def fun1(x):
  return Or(x == cl1, x == cl2, x == cl3, x == cl4, x == cl5, x == cl6, x == cl7, x == cl8, x == cl9)

y1, y2 = Consts('y1 y2', S)
def fun2(y1, y2):
  return Or(And(y1 == cl7, y2 == cl4), And(y1 == cl4, y2 == cl3), And(y1 == cl3, y2 == cl2),And(y1 == cl8, y2 == cl9))

q1,q2 = Consts('q1 q2', S)
def fun3(q1,q2):
  return Or(And(q1 == cl7, q2 == cl5), And(q1 == cl4, q2 == cl9))

t1, t2 ,t3 = Consts('t1 t2 t3', S)

s.add(fun1(t1))
s.add(fun2(t1,t2))
s.add(fun3(t1,t3))

while s.check() == sat:
  s.add(Or(t1 != s.model()[t1], t2 != s.model()[t2], t3 != s.model()[t3]))
  print s.model()
于 2013-03-10T15:20:06.857 回答