当然，我可以为您提供一个简单的 TensorFlow 2.0 版本的 DDPG（Deep Deterministic Policy Gradient）算法的代码示例。请注意，下面的代码是一个基本的框架，您可能需要根据您的具体问题和环境进行一些调整和修改。 ```python import tensorflow as tf from tensorflow.keras.layers import Dense from tensorflow.keras.optimizers import Adam class Actor(tf.keras.Model): def __init__(self, action_dim, max_action): super(Actor, self).__init__() self.dense1 = Dense(400, activation='relu') self.dense2 = Dense(300, activation='relu') self.output_layer = Dense(action_dim, activation='tanh') self.max_action = max_action def call(self, inputs): x = self.dense1(inputs) x = self.dense2(x) x = self.output_layer(x) return self.max_action * x class Critic(tf.keras.Model): def __init__(self): super(Critic, self).__init__() self.dense1 = Dense(400, activation='relu') self.dense2 = Dense(300, activation='relu') self.output_layer = Dense(1) def call(self, state, action): x = tf.concat([state, action], axis=-1) x = self.dense1(x) x = self.dense2(x) return self.output_layer(x) class DDPG: def __init__(self, state_dim, action_dim, max_action): self.actor = Actor(action_dim, max_action) self.target_actor = Actor(action_dim, max_action) self.critic = Critic() self.target_critic = Critic() self.actor_optimizer = Adam(learning_rate=0.001) self.critic_optimizer = Adam(learning_rate=0.001) self.hard_update_target_networks() def hard_update_target_networks(self): self.target_actor.set_weights(self.actor.get_weights()) self.target_critic.set_weights(self.critic.get_weights()) def soft_update_target_networks(self, tau): actor_weights = self.actor.get_weights() target_actor_weights = self.target_actor.get_weights() critic_weights = self.critic.get_weights() target_critic_weights = self.target_critic.get_weights() for i in range(len(target_actor_weights)): target_actor_weights[i] = tau * actor_weights[i] + (1 - tau) * target_actor_weights[i] for i in range(len(target_critic_weights)): target_critic_weights[i] = tau * critic_weights[i] + (1 - tau) * target_critic_weights[i] self.target_actor.set_weights(target_actor_weights) self.target_critic.set_weights(target_critic_weights) def train(self, replay_buffer, batch_size, gamma, tau): states, actions, next_states, rewards, dones = replay_buffer.sample_batch(batch_size) # Update critic with tf.GradientTape() as tape: target_actions = self.target_actor(next_states) target_q = self.target_critic(next_states, target_actions) target_q = rewards + gamma * target_q * (1 - dones) current_q = self.critic(states, actions) critic_loss = tf.reduce_mean(tf.square(current_q - target_q)) critic_gradients = tape.gradient(critic_loss, self.critic.trainable_variables) self.critic_optimizer.apply_gradients(zip(critic_gradients, self.critic.trainable_variables)) # Update actor with tf.GradientTape() as tape: current_actions = self.actor(states) actor_loss = -tf.reduce_mean(self.critic(states, current_actions)) actor_gradients = tape.gradient(actor_loss, self.actor.trainable_variables) self.actor_optimizer.apply_gradients(zip(actor_gradients, self.actor.trainable_variables)) # Update target networks self.soft_update_target_networks(tau) ``` 这是一个简单的 DDPG 算法的代码示例，其中包括了 Actor 和 Critic 网络的定义，以及训练方法 `train()` 的实现。您需要根据您的具体问题和环境进行适当的调整和修改。 请注意，这段代码只是一个示例，并不包含完整的环境交互和经验回放等部分。在实际应用中，您还需要实现经验回放缓冲区（replay buffer）来存储和采样经验数据，并且在训练过程中进行环境交互和经验回放。 希望这个代码示例能够对您有所帮助！