人工智能搜索问题.pptxVIP

第 1 章 搜索问题;状态空间;例 智力游戏问题：传教士与野人渡河问题;初始状态： （3，3，1） 目标状态： （0，0，0）;2;1 4 3 7 6 5 8 2;例 旅行推销员问题;旅行推销员问题的搜索空间;1.1 回溯策略 回溯策略的主要思想: 只保留从初始状态到当前状态的一条解路径, 给变换状态的规则给出一个排序方法, 对当前状态使用规则产生新的状态, 不断地向前延伸解路径. 当没有规则可用, 或向前延伸的状态都是无解状态(称为死点,deadend)时, 沿解路径后退到前一个状态(回溯), 重新开始搜索, 直至???到解或宣布失败. 回溯策略是一种穷尽的搜索方法.; 2.1 回溯算法Backtracking Strategies 递归过程A simple recursive procedure 输入: 问题的初始状态. . The input: the initial state. 输出:一个规则表. 应用这个规则表可以把初始状态变为目标状态. 否则回答FAIL.  The output of the procedure, a list of rules, using it we can get the goal from the initial state. If the procedure can not find the solution, it return FAIL.;4 LOOP: if NULL(RULES), return FAIL;5 R ← FIRST(RULES);6 RULES ← TAIL(RULES);7 RDATA ← R(DATA);8 PATH ←BACKTRACK( RDATA);9 if PATH = FAIL , go LOOP;10 return CONS( R, PATH);In step 3, after get the list of rules using the function APPRULES, we need to order the rules in the lists. The ordering is very important. If the “better”rule is put in the front of the list, it can be used firstly, the efficiency of the system is high. If the “worse” rule is put in the front, the system needs to trymore rules, the efficiency of the system is poor. The Example of Backtracking procedure. The 4 queen problem ;The problem representation the global database: 4*4 array the rule: Rij If i= 1 : there are no queen in the array 1 i= 4: There is a queen in the row i-1 then put a queen in the row i, column jWe order the rules according to the column.我们用Rij表示在棋盘的第 i 行第 j 列放一个王后. 按行的增加顺序依次放皇后, 在规则的排序上依列的上升次序排序.  Termination condition: there are 4 queens in the chess board, they can not kill each other.终止条件: 4 皇后都放到棋盘上, 且不能互相杀死. ;;We can use more informed rule ordering using function diag(i,j)我们可以采用含有较多信息的函数diag(i,j) . Diag(i,j) = the length of the longest diagonal pa