摘要
针对预脱硅过程的RP值在线实时检测,影响生产质量和生产控制实时性问题,提出了RP值软测量的方法,开发了一套基于神经网络软测量的软件系统,利用此系统对预脱硅过程RP值进行预测。现场运行表明,系统误差在4%以内,对生产质量和控制实时性问题可以起到指导作用,证明了该软测量方法的有效性。
According to RP value on-line real-time detection of pre-desilication process, influencing the production quality and real-time production control, this article proposed RP value soft measurement method, and developed a software system based on neural network soft sensor, by using this system to forecast RP value of pre-desilication process. Field operation shows that, the system error is less than 4%, and can play a guiding role for the quality of production and real time control, and also prove the validity of the soft measuring method.
出处
《中国仪器仪表》
2013年第1期47-50,共4页
China Instrumentation
关键词
RP值
软测量
神经网络
RP value
Soft measurement
Neural network
