基于约束调度的失败导向搜索
1. 搜索算法
以下是搜索算法的伪代码:
1 pick :
2 if heap is empty then begin
3 solution found ;
4 add improving objective cut ;
5 goto backtrack ;
6 end ;
7 remove choice c with the best rating from heap ;
8 if c is waiting then begin
9 let c monitor the underlying interval variable ;
10 goto pick ;
11 end ;
12 add c to stack ;
13 if c is resolved then
14 goto pick ;
15 b := branch of c with the better rating
16 propagate :
17 apply branch b ;
18 propagate until a fixed point ;
19 update rating of branch b ;
20 if not infeasible then goto pick ;
21 backtrack :
22 let c is the last choice with open branch on the stack ;
23 if there is no such c then
24 terminate ;
// whole search space was explored
25 put back into the hea
超级会员免费看
订阅专栏 解锁全文
17

被折叠的 条评论
为什么被折叠?



