基于长短期记忆网络陶瓷梭式窑温度预测控制

Temperature Predictive Control of Ceramic Shuttle Kiln Based on Long-and Short-Term Memory Network

陶瓷梭式窑是一种形似抽屉陶瓷生产设备,其温度控制直接影响陶瓷产品质量。针对目前陶瓷梭式窑生产过程人工劳动强度大,控制方法落后,自动化程度低的情况,提出了基于长短期记忆网络陶瓷梭式窑温度预测控制方法。首先,通过长短期记忆网络来实现对陶瓷梭式窑温度进行预测建模;然后通过滚动优化算法和反馈校正算法实现对陶瓷梭式窑温度控制;最后,进行了仿真实验,仿真结果表明所提出的方法比PID控制、一般的预测控制以及基于普通神经网络的预测控制方法均好,并具有良好的动静态性能和强鲁棒性,因而可行有效。所提出的方法为提高陶瓷梭式窑自动化水平提供一种新方法和新途径。

Ceramic shuttle kiln is a kind of drawer-like ceramic production equipment,and its temperature control directly affects the quality of ceramic products.In view of the high labor intensity,backward control methods and low automation in the ceramic shuttle kiln production process,a temperature predictive control method based on long-short-term memory network is proposed in this paper.Firstly,the temperature of ceramic shuttle kiln is predicted and modeled by long-term and short-term memory network,and then the temperature of ceramic shuttle kiln is controlled by rolling optimization algorithm and feedback correction algorithm.Finally,the simulation results show that the proposed method is better than PID control,general predictive control and predictive control based on ordinary neural network,and is with good dynamic performance,static performance and strong robustness.Hence,this method is feasible and effective.The proposed method provides a new method and a new way to improve the automation level of ceramic shuttle kiln.