alpha beta 滤波_alpha-beta-search算法

func max(alpha,beta,state):

v=max(action,min(alpha,-infinite,infinite)) for action in action(s)

return action which in actions(state) with value=v

func max-value(alpha,beta,state):

if end(state):return u(state)

v=-infinite

for action in action(s):

v=max(v,min-value(alpha,beta,result(state,action))

if v<=beta:return v

alpha=max(v,alpha)

return v

func min-value(alpha,beta,state):

if end(state):return u(state)

v=infinite

for action in action(s):

v=min(v,max-value(alpha,beta,result(state,action)))

if v>=alpha:return v

beta=min(v,beta)

return v

提供一种框架，上下剪枝，更新上下界，从而减少不必要的搜索量。