写一个在gym环境下运行的移动障碍物网格环境

下面是一个在gym环境下运行的移动障碍物网格环境的示例代码：

import gym
from gym import spaces
import numpy as np

class GridWorldEnv(gym.Env):
    def __init__(self, grid_size=5):
        super(GridWorldEnv, self).__init__()

        self.grid_size = grid_size
        self.start_position = (0, 0)
        self.target_position = (grid_size-1, grid_size-1)
        self.current_position = self.start_position

        self.action_space = spaces.Discrete(4)  # 上、下、左、右四个动作
        self.observation_space = spaces.Tuple((
            spaces.Discrete(grid_size),
            spaces.Discrete(grid_size)
        ))

    def reset(self):
        self.current_position = self.start_position
        return self.current_position

    def step(self, action):
        x, y = self.current_position

        if action == 0:  # 上
            y += 1
        elif action == 1:  # 下
            y -= 1
        elif action == 2:  # 左
            x -= 1
        elif action == 3:  # 右
            x += 1

        # 检查新位置的合法性
        if 0 <= x < self.grid_size and 0 <= y < self.grid_size:
            self.current_position = (x, y)

        done = self.current_position == self.target_position
        reward = 1.0 if done else 0.0

        return self.current_position, reward, done, {}

    def render(self):
        grid = np.zeros((self.grid_size, self.grid_size))
        grid[self.start_position] = 0.5
        grid[self.target_position] = 0.5
        grid[self.current_position] = 1.0

        for row in grid:
            print(' '.join(['{:1.1f}'.format(cell) for cell in row]))

# 使用示例
env = GridWorldEnv(grid_size=5)
observation = env.reset()

for _ in range(10):
    env.render()
    action = env.action_space.sample()
    observation, reward, done, _ = env.step(action)
    if done:
        break

运行上述代码，您将看到一个5x5的网格环境，起始位置标记为0.5，目标位置标记为0.5，当前位置标记为1.0。每次循环中，随机选择一个动作，更新环境状态，直到到达目标位置为止。