Aiming at the characteristics of large delay,nonlinear,slow time varying and strong coupling,an intelligent temperature control method based on improved DQN algorithm was proposed.Firstly,the model of ceramic shuttle kiln based on BP neural network was established.Then an intelligent control method based on improved DQN algorithm is proposed.Finally,the pro-posed method is simulated.The simulation results show that the relative error of the improved PRDQN algorithm is between 0 ℃ and 5 ℃,and the temperature control effect is relatively good.Therefore,the proposed method is effective and feasible.
关键词
陶瓷梭式窑/深度强化学习/BP神经网络/PRDQN算法
Key words
ceramic shuttle kiln/multi-agent deep reinforcement learning/BP neural network/improved DQN algorithm
维普
万方数据