Abstract The reconstruction problem of missing data cannot be solved by the existing linear data methods in nonlinear industrial process. In order to realize reconstruction of nonlinear data, a novel neural network named partly input-adjusting neural network is proposed, whose missing variables are selected as the inputs. Different from the conventional network, the weights and threshold of the novel network have already been obtained by other network training. By back-propagation algorithm, the reconstruction is achieved. The inputs of the network are adjusted by backpropagation algorithm, then the reconstruction is achieved and the training is completed. The simulation result proves the validity of the proposed method.
[1] Walczak B, Massart D L. Dealing with Missing Data: PartⅠ. Chemometrics and Intelligent Laboratory Systems, 2001, 58(1): 15-27 [2] Walczak B, Massart D L. Dealing with Missing Data: Part Ⅱ. Chemometrics and Intelligent Laboratory Systems, 2001, 58(1): 29-42 [3] Nelson P R C, Taylar P A, MacGregor J F. Missing Data Methods in PCA and PLS: Score Calculations with Incomplete Observation. Chemometrics and Intelligent Laboratory Systems, 2001, 35(1): 45-65 [4] Nelson P R C, Taylar P A, MacGregor J F. The Impact of Missing Measurements on PCA and PLS Prediction and Monitoring Application. Chemometrics and Intelligent Laboratory Systems, 2006, 80(1): 1-12 [5] Goulding P R, Lennox B, Sandoz D J, et al. Fault Detection in Continuous Processes Using Multivariate Statistical Methods. International Journal of Systems Science, 2000, 31(11): 1459-1471 [6] Lieftucht D, Kruger U, Irwin G W. Improved Diagnosis of Sensor Faults Using Multivariate Statistics // Proc of the American Control Conference. Boston, USA, 2004, Ⅴ: 4403-4407 [7] Lee J M, Yoo C, Choi S W, et al. Nonlinear Process Monitoring Using Kernel Principal Component Analysis. Chemical Engineering Science, 2004, 59(1): 223-234 [8] Tan S, Mavrovouniotis M L. Reducing Data Dimensionality through Optimizing Neural Network Inputs. AIChE Journal, 1995, 41(6): 1471-1480 [9] Kramer M A. Nonlinear Principal Component Analysis Using Autoassociative Neural Networks. AIChE Journal, 1991, 37(2): 233-243 [10] Saegusa R, Sakana H, Hashimoto S. Nonlinear Principal Analysis to Preserve the Order of Principal Components. Neurocomputing, 2004, 61(1): 57-70 [11] Zhao Zhonggai, Liu Fei, Xu Baoguo. Nonlinear Principal Component Analysis Based on Hierarchical Input-Training Neural Network. Information and Control, 2005, 34(6): 656-659 (in Chinese) (赵忠盖,刘 飞,徐保国.一种基于分级输入训练神经网络的非线性主元分析.信息与控制, 2005, 34(6): 656-659) [12] Dong D, McAvoy T J. Nonlinear Principal Component Analysis Based on Principal Curves and Neural Networks. Computers and Chemical Engineering, 1996, 20(1): 65-78
2007, Vol. 20 
