基于软测量的自动配煤控制系统设计  

Methodology of Coal Auto-blending System Based on Soft-sensing

作　　者：盛金辉[1,2] 黄振利[1] 庄诚[1] 林红权[1] 

机构地区：[1]中国科学院自动化研究所 [2]中国科学院研究生院,北京100080

出　　处：《控制工程》2005年第S1期42-44,共3页Control Engineering of China

摘　　要：针对配煤这一炼焦生产过程中的重要环节,其配比的准确性和配煤系统的可靠性将直接影响焦炭的质量和经济效益以及能否顺利推焦,某焦化厂由于所使用的皮带秤精度差,难以投入闭环控制,提出了一种基于多元线性回归的软测量技术,实现了配煤的准确称重;同时配煤过程被控对象具有非线性、时变、时滞,并且很难建立精确的数学模型,提出了基于专家知识的参数自整定模糊PID控制方法。这种基于软测量的模糊自整定方法在该厂配煤过程中成功应用,投运以来配煤合格率由原来的70%提高到95%,实现了稳定优化配煤,稳定焦炉生产,提高焦碳质量的目的。该方法具有广阔的应用前景和推广价值。Coal blending is an important tache in coking process. Coke cjuality is affected by the precision of coal blending ratio and reliability of blending system. The precision of the belt balance is bad, and closed loop control can not work well in the coal blending process. A soft-sensing technique based on multi-element linear regression is presented, and good precision of weight is attained. Furthermore, the coal blending process is nonlinear, time-varying with time-delays and the accurate mathematical model can not be obtained. Therefore a self-tuning fuzzy PID control algorithm based on expert knowledge is proposed. The practical result indicates that the methodology of self-tuning fuzzy PID based on soft-sensing technique is very effective.

关 键 词：配煤 软测量 模糊PID 

分 类 号：TP273.5[自动化与计算机技术—检测技术与自动化装置]