【摘要】 针对传统柔性演员-评论家算法在探索能力和复杂环境中状态表征不足的问题，提出一种改进的柔性演员-评论家算法.首先，该算法通过引入优先经验回放机制，利用时序差分误差对经验样本进行动态优先级评估，从而提高关键经验的利用率，进而提升学习效率；其次，该算法将生成式Transformer架构集成到演员网络中以增强对状态特征的动态捕捉能力，从而显著提升其在复杂优化任务中的性能；最后，在高校后勤人员动态调度优化问题上进行应用实验.实验结果表明，与原始柔性演员-评论家算法及经典深度Q网络算法相比，改进的柔性演员-评论家算法在人力需求动态拟合方面误差更小，从而有效验证了其在实际应用中的优势和实用性.更多还原

【Abstract】 Aiming at the problem that the conventional soft actor-critic(SAC) algorithm lacked exploration capability and state representation in complex environments, we proposed an improved soft actor-critic(ISAC) algorithm. Firstly, the ISAC algorithm introduced a prioritized experience replay(PER) mechanism, which dynamically evaluated the priority of experience samples by using the temporal differential errors, thereby enhancing the utilization of crucial experiences and improving learning efficiency of the algorithm. Secondly, the algorithm integrated generative Transformer architecture into the actor network to strengthen its ability to dynamically capture state features, thereby significantly improving its performance in complex optimization tasks. Finally, we conducted an application experiment on the dynamic scheduling optimization problem of university logistics staff. The experimental results show that, compared with the original SAC algorithm and the classic deep Q-network(DQN) algorithm, the proposed ISAC algorithm has smaller errors in dynamically fitting human resource demand, which effectively demonstrates its advantages and practicality in practical applications.更多还原