当前位置：首页 > news >正文

强化学习PPO/DDPG算法学习记录

news 2025/9/4 15:06:16

强化学习PPO算法详解

核心思想

直接学习一个策略函数pi(a|s), 在状态s下要输出的动作a的概率分布（离散情况下是每个action的概率，连续情况下是mean和std指定的高斯分布）
策略梯度算法如果更新的步长太大，一次更新就能毁掉整个策略，所以通过一个裁剪函数，防止新策略和旧策略差距太大，保证“小幅且安全”。
多步更新：在收集一批数据后，用小批量数据对策略进行多次epochs更新，提高样本效率。

参考代码：

这个PPO实现代码有什么问题？里面好像没有用到critic网络的结果：import torch
import torch.nn as nn
import torch.optim as optimclass PPO(nn.Module):def __init__(self, state_dim, action_dim):super(PPO, self).__init__()self.actor = nn.Sequential(nn.Linear(state_dim, 64),nn.Tanh(),nn.Linear(64, 64),nn.Tanh(),nn.Linear(64, action_dim),nn.Softmax(dim=-1))self.critic = nn.Sequential(nn.Linear(state_dim, 64),nn.Tanh(),nn.Linear(64, 64),nn.Tanh(),nn.Linear(64, 1))def forward(self, state):return self.actor(state), self.critic(state)class PPOAgent:def __init__(self, state_dim, action_dim, lr=3e-4, gamma=0.99, epsilon=0.2):self.ppo = PPO(state_dim, action_dim)self.optimizer = optim.Adam(self.ppo.parameters(), lr=lr)self.gamma = gammaself.epsilon = epsilondef update(self, states, actions, rewards, next_states, dones):states = torch.FloatTensor(states)actions = torch.LongTensor(actions)rewards = torch.FloatTensor(rewards)next_states = torch.FloatTensor(next_states)dones = torch.FloatTensor(dones)# 1. 首先，用当前的策略网络（不计算梯度）计算旧概率 old_probswith torch.no_grad():old_probs, old_state_values = self.ppo(states)old_probs = old_probs.gather(1, actions.unsqueeze(1)).squeeze(1)# 计算下一个状态的价值_, next_state_values = self.ppo(next_states)# 计算价值目标：如果done了，next_state_value就是0value_targets = rewards + self.gamma * next_state_values.squeeze(1) * (1 - dones)# 计算优势函数 A(s,a) = value_target - old_state_valueadvantages = value_targets - old_state_values.squeeze(1)# 通常会对advantages进行标准化，以减少方差advantages = (advantages - advantages.mean()) / (advantages.std() + 1e-8)for _ in range(10):  # 多次更新# 2. 用当前策略网络计算新概率和状态价值new_probs, state_values = self.ppo(states)new_probs = new_probs.gather(1, actions.unsqueeze(1)).squeeze(1)# 3. 计算重要性采样比率ratio = new_probs / old_probs# 4. 计算Clipped Surrogate Losssurr1 = ratio * advantagessurr2 = torch.clamp(ratio, 1 - self.epsilon, 1 + self.epsilon) * advantagesactor_loss = -torch.min(surr1, surr2).mean()# 5. 计算Critic Loss (MSE between value_targets and current value estimates)critic_loss = nn.MSELoss()(state_values.squeeze(1), value_targets)# 6. 总损失loss = actor_loss + 0.5 * critic_lossself.optimizer.zero_grad()loss.backward()self.optimizer.step()def get_action(self, state):state = torch.FloatTensor(state)probs, _ = self.ppo(state)return torch.multinomial(probs, 1).item()

PPO是在线策略，输出的是概率，更新稳健，是策略网络的集大成者。

DDPG算法详解

核心改进

相比于DQN，DDPG的核心改进在于：

DQN的动作空间是离散的，例如上下左右开火等，而DDPG的动作空间是连续的
DQN输出的是Q值，然后选择最大的对应的动作，DDPG直接输出动作
DQN是value-based，而DDPG是Policy- based
DQN通常只有一个Q网络，DDPG要有Actor和critic两个网络

一句话总结：DQN是为了解决离散控制问题，DDPG主要是针对连续控制领域的,是DQN和策略网络的结合。
参考代码(原文章：Deep Reinforcement Learning (DRL) 算法在 PyTorch 中的实现与应用）：

import torch
import torch.nn as nn
import torch.optim as optimclass Actor(nn.Module):def __init__(self, state_dim, action_dim, max_action):super(Actor, self).__init__()self.fc1 = nn.Linear(state_dim, 400)self.fc2 = nn.Linear(400, 300)self.fc3 = nn.Linear(300, action_dim)self.max_action = max_actiondef forward(self, state):a = torch.relu(self.fc1(state))a = torch.relu(self.fc2(a))return self.max_action * torch.tanh(self.fc3(a))class Critic(nn.Module):def __init__(self, state_dim, action_dim):super(Critic, self).__init__()self.fc1 = nn.Linear(state_dim + action_dim, 400)self.fc2 = nn.Linear(400, 300)self.fc3 = nn.Linear(300, 1)def forward(self, state, action):q = torch.cat([state, action], 1)q = torch.relu(self.fc1(q))q = torch.relu(self.fc2(q))return self.fc3(q)class DDPGAgent:def __init__(self, state_dim, action_dim, max_action, lr=1e-4, gamma=0.99, tau=0.001):self.actor = Actor(state_dim, action_dim, max_action)self.actor_target = Actor(state_dim, action_dim, max_action)self.actor_target.load_state_dict(self.actor.state_dict())self.actor_optimizer = optim.Adam(self.actor.parameters(), lr=lr)self.critic = Critic(state_dim, action_dim)self.critic_target = Critic(state_dim, action_dim)self.critic_target.load_state_dict(self.critic.state_dict())self.critic_optimizer = optim.Adam(self.critic.parameters(), lr=lr)self.gamma = gammaself.tau = taudef select_action(self, state):state = torch.FloatTensor(state.reshape(1, -1))return self.actor(state).cpu().data.numpy().flatten()def update(self, replay_buffer, batch_size=100):# 从经验回放中采样state, action, next_state, reward, done = replay_buffer.sample(batch_size)# 计算目标Q值target_Q = self.critic_target(next_state, self.actor_target(next_state))target_Q = reward + (1 - done) * self.gamma * target_Q.detach()# 更新Criticcurrent_Q = self.critic(state, action)critic_loss = nn.MSELoss()(current_Q, target_Q)self.critic_optimizer.zero_grad()critic_loss.backward()self.critic_optimizer.step()# 更新Actoractor_loss = -self.critic(state, self.actor(state)).mean()self.actor_optimizer.zero_grad()actor_loss.backward()self.actor_optimizer.step()# 软更新目标网络for param, target_param in zip(self.critic.parameters(), self.critic_target.parameters()):target_param.data.copy_(self.tau * param.data + (1 - self.tau) * target_param.data)for param, target_param in zip(self.actor.parameters(), self.actor_target.parameters()):target_param.data.copy_(self.tau * param.data + (1 - self.tau) * target_param.data)