actor-critic代码逐行解析（tensorflow版）

深度强化学习算法actor-critic代码逐行解析（tensorflow版）

Actor是基于Policy-Gradients。可以选择连续动作，但是必须循环一个回合才可以更新策略。学习效率低。
Critic网络继承了Q-learning 的传统，依然可以逐步更新。
对于首先导入所需的包，没有什么可说的。

import numpy as np
import tensorflow as tf
import gym
import matplotlib.pyplot as plt

np.random.seed(2)
tf.set_random_seed(2)

OUTPUT_GRAPH = False
MAX_EPISODE = 5
DISPLAY_REWARD_THRESHOLD = 200
MAX_EP_STEPS = 500
RENDER = False
GAMMA = 0.9
LR_A = 0.001
LR_C = 0.01

env = gym.make('CartPole-v0')
env.seed(1)
env = env.unwrapped

N_F = env.observation_space.shape[0]
N_A = env.action_space.n

Actor网络

class Actor(object):
    def __init__(self, sess, n_features, n_actions, lr=0.001):
        self.sess = sess

        self.s = tf.placeholder(tf.float32, [1, n_features], "state")
        self.a = tf.placeholder(tf.int32, None, "act")
        self.td_error = tf.placeholder(tf.float32, None, "td_error")

        with tf.variable_scope('Actor'):
            l1 = tf.layers.dense(
                inputs=self.s,
                units=20,
                activation=tf.nn.relu,
                kernel_initializer=tf.random_normal_initializer(0., .1),
                bias_initializer=tf.constant_initializer(0.1),
                name='l1'
            )

            self.acts_prob = tf.layers.dense(
                inputs=l1,
                units=n_actions,
                activation=tf.nn.softmax,
                kernel_initializer=tf.random_normal_initializer(0., .1),
                bias_initializer=tf.constant_initializer(0.1),
                name='acts_prob'
            )

        with tf.variable_scope('exp_v'):
            log_prob = tf.log(self.acts_prob[0, self.a])
            self.exp_v = tf.reduce_mean(log_prob * self.td_error)

        with tf.variable_scope('train'):
            self.train_op = tf.train.AdamOptimizer(lr).minimize(-self.exp_v)

    def learn(self, s, a, td):
        s = s[np.newaxis, :]
        feed_dict = {self.s: s, self.a: a, self.td_error: td}
        _, exp_v = self.sess.run([self.train_op, self.exp_v], feed_dict)
        return exp_v

    def choose_action(self, s):
        s = s[np.newaxis, :]
        probs = self.sess.run(self.acts_prob, {self.s: s})
        return np.random.choice(np.arange(probs.shape[1]), p=probs.ravel())

具体流程如下：
1、两个全连接层网络，一层神经元的个数为20个，第二层的输入是第一层l1的输出。
2、loss依然是Policy-Gradients的-log(probs)*vt。probs可以看出是第二层神经网络的输出，是动作空间下所有动作的概率。vt是Critic计算出的时间差分误差td_error。
3、训练步骤。喂入动作，状态。和td误差即可。
4、选择动作.

Critic网络
伟大的Critic网络负责给Actor网络最后输出的动作打分，并通过td_error返回给Actor用于更新Actor的参数。同样的也是两层网络。
但是输入有三个。但是和Actor的三个输入不同。因为a是固定的值。
输入当前状态、当前奖励和下一状态的折扣奖励：

class Critic(object):
    def __init__(self, sess, n_features, lr=0.01):
        self.sess = sess

        self.s = tf.placeholder(tf.float32, [1, n_features], "state")
        self.v_ = tf.placeholder(tf.float32, [1, 1], "v_next")
        self.r = tf.placeholder(tf.float32, None, 'r')

        with tf.variable_scope('Critic'):
            l1 = tf.layers.dense(
                inputs=self.s,
                units=20,
                activation=tf.nn.relu,

                kernel_initializer=tf.random_normal_initializer(0., .1),
                bias_initializer=tf.constant_initializer(0.1),
                name='l1'
            )

            self.v = tf.layers.dense(
                inputs=l1,
                units=1,
                activation=None,
                kernel_initializer=tf.random_normal_initializer(0., .1),
                bias_initializer=tf.constant_initializer(0.1),
                name='V'
            )

        with tf.variable_scope('squared_TD_error'):
            self.td_error = self.r + GAMMA * self.v_ - self.v
            self.loss = tf.square(self.td_error)
        with tf.variable_scope('train'):
            self.train_op = tf.train.AdamOptimizer(lr).minimize(self.loss)

    def learn(self, s, r, s_):
        s, s_ = s[np.newaxis, :], s_[np.newaxis, :]

        v_ = self.sess.run(self.v, {self.s: s_})
        td_error, _ = self.sess.run([self.td_error, self.train_op],
                                    {self.s: s, self.v_: v_, self.r: r})
        return td_error

具体流程如下：
1、两个全连接层网络，一层神经元的个数为20个，第二层的输入是第一层l1的输出。可以发现这个输出只有一维度。就是a做这个动作的奖励。
2、loss：时间差分值的平方（取出下一时刻的动作的奖励）
3、学习步骤：Critic神经网络前向传播一波，将下一个输入，得到评价动作的值，和第一步一样，输入当前状态，当前奖励，下一个状态的折扣奖励v_，用Adam优化器反向传播一波。
.
大功告成网络

sess = tf.Session()
actor = Actor(sess, n_features=N_F, n_actions=N_A, lr=LR_A)
critic = Critic(sess, n_features=N_F, lr=LR_C)
sess.run(tf.global_variables_initializer())

if OUTPUT_GRAPH:
    tf.summary.FileWriter("logs/", sess.graph)
track_r = []

for i_episode in range(MAX_EPISODE):
    s = env.reset()
    t = 0

    while True:
        if RENDER: env.render()
        a = actor.choose_action(s)
        s_, r, done, info = env.step(a)
        if done: r = -20

        track_r.append(r)
        td_error = critic.learn(s, r, s_)
        actor.learn(s, a, td_error)
        s = s_
        t += 1

        if done or t >= MAX_EP_STEPS:

            ep_rs_sum = sum(track_r)
            if 'running_reward' not in globals():
                running_reward = ep_rs_sum
            else:
                running_reward = running_reward * 0.95 + ep_rs_sum * 0.05
            if running_reward > DISPLAY_REWARD_THRESHOLD: RENDER = True
            print("episode:", i_episode, "  reward:", int(running_reward))
            break

Original: https://blog.csdn.net/u014136435/article/details/123585523
Author: Blackquiten
Title: actor-critic代码逐行解析（tensorflow版）