2

我试图在 PyBrain 中实现类似于迷宫问题的东西。但是，它更类似于带有紧急出口的房间，您可以将代理人留在其中一个房间中以找到出口。要将其转换为计算机方法，可以使用双向图，权重显示房间之间的路径。

我试图实现一个新环境，但我有点迷失了应该是什么。例如，基于抽象环境类我想到了这个：

#!/usr/bin/python2.7

class RoomEnv(Environment):
    # number of action values acceptable by the environment
    # Two events: go forward and go back through the door (but, how we know what room is connect to another?)
    indim = 2
    # Maybe a matrix where 0 is no connection and 1 is a connection(?)
    #            A,B,C,D,E,F
    #indim = array([[0,0,0,0,0,0],  # A
                    [0,0,0,0,0,1],  # B
                    [0,0,0,0,0,0],  # C
                    [0,0,0,0,0,0],  # D
                    [0,0,0,0,0,1],  # E
                    [0,0,0,0,0,1],  # F
                  ])

    # the number of sensors is the number of the rooms
    outdim = 6

    def getSensors(self):
        # Initial state:
        # Could be any room, maybe something random(?)

    def performAction(self, action):
        # We should look at all the states possible to learn what are the best option to go to the outside state.
        # Maybe a for loop that goes through all the paths and use some weight to know where is the best option?

        print "Action performed: ", action

    def reset(self):
        #Most environments will implement this optional method that allows for reinitialization.

真挚地，

4

1 回答 1

1

在pybrain中，您可以将房间定义为一个数组，然后将结构传递给 Maze 以创建一个新环境。例如：

structure = array([[1, 1, 1, 1, 1, 1, 1, 1, 1],
                   [1, 0, 0, 1, 0, 0, 0, 0, 1],
                   [1, 0, 0, 1, 0, 0, 1, 0, 1],
                   [1, 0, 0, 1, 0, 0, 1, 0, 1],
                   [1, 0, 0, 1, 0, 1, 1, 0, 1],
                   [1, 0, 0, 0, 0, 0, 1, 0, 1],
                   [1, 1, 1, 1, 1, 1, 1, 0, 1],
                   [1, 0, 0, 0, 0, 0, 0, 0, 1],
                   [1, 1, 1, 1, 1, 1, 1, 1, 1]])

# defining the environment
environment = Maze(structure, (7, 7))

在上面的示例中，1 表示墙壁，0 表示代理可以在上面行走的网格。所以你可以修改结构来制作你自己的。

于 2014-02-24T15:18:43.403 回答