摘要
针对纺纱车间采用传统的PID控制车间温湿度不稳定问题,采用机器学习的BP神经网络对后台历史数据进行监督学习,并建立空调设备自控参数的专家库。当PID调控产生温湿度较大波动时采用信息库建立好的空调控制参数直接介入辅助控制。基于热量平衡、湿量平衡以及风量平衡开发了纺纱空调仿真系统,并分析了风机、水泵、二回风窗、新风窗对于车间温湿度控制的影响。为验证辅助调控的效果,将BP神经网络预测的空调控制参数介入车间温湿度控制,并与传统PID控制车间的温湿度进行了对比。结果表明:采用BP神经网络辅助调控空调系统的参数,车间温湿度稳定性优于传统PID控制。认为:机器学习辅助调控纺纱空调系统可达到稳定控制温湿度的目的。
Aiming at the instability of temperature&humidity in spinning workshop controlled by traditional PID,BP neural network based on machine learning was used to supervise and learn the background historical data,an expert database of automatic control parameters of air conditioner equipment was established.When PID control results in large temperature&humidity fluctuations,air conditioner control parameters established by the information database were directly involved in auxiliary control.Based on heat balance,moisture balance and air balance,spinning air conditioner simulation system was developed.The influence of fan,water pump,secondary return air window and fresh air window on temperature&humidity control in workshop was analyzed.In order to verify the effect of auxiliary control,the temperature&humidity in the workshop were compared with that of the traditional PID control when the air conditioning control parameters predicted by BP neural network were involved in the control.The results showed that the stability of temperature&humidity in workshop was better than that of traditional PID control when BP neural network was used to control the parameters of air conditioner system.It is considered that machine learning assisted control of spinning air conditioner system can achieve stable control of temperature&humidity.
作者
吴清贵
韩云龙
陆彪
高杰
汪虎明
WU Qinggui;HAN Yunlong;LU Biao;GAO Jie;WANG Huming(Anhui University of Technology,Ma'anshan,243032,China;Jiangsu Jingya Environmental Technology Co.,Ltd.,Wuxi,214426,China)
出处
《棉纺织技术》
CAS
2024年第6期69-74,共6页
Cotton Textile Technology
关键词
纺纱空调
PID控制
机器学习
BP神经网络
车间温湿度
spinning air conditioner
PID control
machine learning
BP neural network
temperature&humidity in workshop
