基于双权M-估计子的鲁棒数据校正新方法New Method for Robust Data Reconciliation Based on Biweight M-Estimator
金思毅,刘猛,李宁,陶少辉
摘要(Abstract):
准确可靠的测量数据是实现装置过程控制、模拟、优化和生产管理的前提条件,而通过仪表测量获取的过程数据中可能存在过失误差,直接影响数据校正的准确性,现有的数据校正方法不能完全有效地避免过失误差的影响。根据双权M-估计的原理,今以相对残差为变量构造了一种新型的具有强鲁棒性的目标函数,使含有过失误差的变量对函数的贡献为一常数,从而避免了过失误差对数据校正过程的影响。选取了具有代表性的一个线性问题和一个非线性问题进行实例研究,并与现有的Huber法和Cauchy法进行了对比分析。计算结果表明,对线性系统和非线性系统,新方法的过失误差侦破性能均优于Huber法和Cauchy法,且其稳定性更高。因此,在进行数据校正时应首选新方法。
关键词(KeyWords): 数据校正;过失误差侦破;双权M-估计子;鲁棒性
基金项目(Foundation): 重质油国家重点实验室开放课题基金资助项目(2008-08)
作者(Author): 金思毅,刘猛,李宁,陶少辉
参考文献(References):
- [1]YUAN Yong-gen(袁永根),LI Hua-sheng(李华生).Data Reconciliation Technology of Process System Measurements(过程系统测量数据校正技术)[M].Beijing(北京):China Petroleum and Chemistry Press(中国石化出版社),1996.
- [2]Bhat S A,Saraf D N.Steady-state identification,gross error detection,and data reconciliation for industrial process units[J].IECR J,2004,43(15):4323-4336.
- [3]Huber P J.Robust Statistics[M].New York:Willey,1981.
- [4]Johnston L P M,Kramer M A.Maximum likelihood data reconciliation:steady-state systems[J].AIChE J,1995,41(11):2415426.
- [5]zyurt D B,Pike R W.Theory and practice of simultaneous data reconciliation and gross error detection for chemical processes[J].Computers Chem Eng,2004,28(3):381402.
- [6]ZHOU Ling-ke,SU Hong-ye,CHU Jian.A new method to solve robust data reconciliation in nonlinear process[J].Chinese J Chem Eng,2006,14(3):357-363.
- [7]ZHANG Zheng-jiang(张正江),ZHU Ling-yu(祝钰铃).Data reconciliation based on large scale rigorous model with equation oriented method(基于大规模严格机理模型的数据校正)[J].J Chem Eng of Chinese Univ(高校化学工程学报)[J].2008,22(5):877-882.
- [8]Tukey J W.Explortary Data Analysis[M].Reading,MA:Addison-Wesley,1977.
- [9]Rosenberg J,Mah R S H,Iordache C.Evaluation of schemes for detecting and identifying gross errors in process data[J].Ind Engng Chem Res,1987,26(3):555564.
- [10]Tjoa I B,Biegler L T.Simultaneous strategies for data reconciliation and gross error detection of nonlinear systems[J].Computers Chem Eng,1991,15(10):679690.