Fault diagnosis for automatic shell magazine using model verification, FDA and ELM

A fault diagnosis method for an automatic shell magazine based on the model verification, Functional Data Analysis (FDA) and Extreme Learning Machine (ELM) is presented in this paper. A virtual prototype model of the automatic shell magazine includes a mechanical model, and a control model is built in RecurDyn and Simulink. The failure mechanism of the automatic shell magazine is analyzed, and the corresponding fault factors are selected. Due to an insufficient number of fault samples, the magazine displacement and the rotating angle of the driving wheel are tested. The virtual prototype model is verified by comparing the test data with the output of a virtual prototype model. A large number of fault samples is generated by the verified model, and the fault samples are analyzed by FDA. Then, the eigenvalues from the FDA and FPCA are used to train the ELM to obtain a diagnostic machine. The diagnostic machine is used for the fault diagnosis of the automatic shell magazine and is proved to be very effective.

[1]  C. Ordóñez Galán,et al.  Functional data analysis as a tool to correlate textural and geochemical data , 2013, Appl. Math. Comput..

[2]  Harold A. Rothbart,et al.  Mechanical design handbook : measurement, analysis, and control of dynamic systems , 2006 .

[3]  Gustavo Arechavaleta,et al.  Practical guide to solve the minimum-effort problem with geometric algorithms and B-Splines , 2019, 2019 International Conference on Robotics and Automation (ICRA).

[4]  Ren Haifeng.  Analysis of Artillery Autoloader Faults by Current Recognition Method , 2013 .

[5]  J. Ramsay,et al.  Some Tools for Functional Data Analysis , 1991 .

[6]  Yan Fen.  Fault Diagnosis of Tiltrotor Aircraft via Improved Discrete Wavelet-OMELM , 2014 .

[7]  侯保林,et al.  Reliability analysis of rotation positioning precision for howitzer shell transfer arm based on virtual experiment , 2014 .

[8]  S. Konishi,et al.  Functional principal component analysis via regularized Gaussian basis expansions and its application to unbalanced data , 2007 .

[9]  Guang-Bin Huang,et al.  Upper bounds on the number of hidden neurons in feedforward networks with arbitrary bounded nonlinear activation functions , 1998, IEEE Trans. Neural Networks.

[10]  Alan S. Willsky,et al.  A survey of design methods for failure detection in dynamic systems , 1976, Autom..

[11]  R. K. Mehra,et al.  Correspondence item: An innovations approach to fault detection and diagnosis in dynamic systems , 1971 .

[12]  Christos Davatzikos,et al.  Functional principal component model for high-dimensional brain imaging , 2011, NeuroImage.

[13]  J. Ramsay.  When the data are functions , 1982 .

[14]  Zhixin Yang,et al.  Real-time fault diagnosis for gas turbine generator systems using extreme learning machine , 2014, Neurocomputing.

[15]  D. Serre.  Matrices: Theory and Applications , 2002 .

[16]  Spencer Graves,et al.  Functional Data Analysis with R and MATLAB , 2009 .

[17]  Jian Ma,et al.  Rolling bearing fault diagnosis under variable conditions using LMD-SVD and extreme learning machine , 2015 .

[18]  Irene Epifanio,et al.  Functional Data Analysis in Shape Analysis , 2011 .

[19]  C. R. Rao,et al.  Generalized Inverse of Matrices and its Applications , 1972 .

[20]  Guang-Bin Huang,et al.  Learning capability and storage capacity of two-hidden-layer feedforward networks , 2003, IEEE Trans. Neural Networks.

[21]  M. M. Segovia-Gonzalez,et al.  Explaining functional principal component analysis to actuarial science with an example on vehicle insurance , 2009 .

[22]  Carl de Boor,et al.  A Practical Guide to Splines , 1978, Applied Mathematical Sciences.

[23]  Chee Kheong Siew,et al.  Extreme learning machine: Theory and applications , 2006, Neurocomputing.