基于Q-Learning的青霉素发酵过程控制方法

Control Method for Penicillin Fermentation Process Based on Q-Learning

针对在复杂场景下,保持青霉素发酵过程中的温度稳定性以及提高青霉素产量的问题,本文建立了一种基于强化学习的青霉素发酵系统模型。选取强化学习算法中的Q-Learning算法,设计基于Q-Learning强化学习理论的控制策略,将控制任务分解成两个子任务分别进行控制,使青霉素在发酵过程中保持稳定。建立仿真环境并进行实验,仿真结果表明,该方法可以在未知环境变化的青霉素发酵过程中有效地使温度趋于稳定,具有一定的应用价值。

A penicillin fermentation system model based on reinforcement learning was established to maintain the temperature stability and improve the penicillin production in the process of penicillin fermentation in complex scenes.The Q-Learning algorithm in the reinforcement learning algorithm is selected,and the control strategy based on the Q-Learning reinforcement learning theory is designed.The control task is divided into two subtasks for control respectively,so that penicillin is stable in the fermentation process.A simulation environment was established to conduct experiments.The simulation results show that this method can effectively stabilize the temperature during penicillin fermentation in unknown environmental changes,and has certain application value.