Three feature extraction methods of sucker-rod pump indicator card data have been studied, simulated, and compared in this\r\npaper, which are based on Fourier Descriptors (FD), GeometricMoment Vector (GMV), and Gray LevelMatrix Statistics (GLMX),\r\nrespectively.Numerical experiments show that the FourierDescriptors algorithm requires less running time and less memory space\r\nwith possible loss of information due to nonoptimal numbers of Fourier Descriptors, the Geometric Moment Vector algorithm is\r\nmore time-consuming and requires more memory space, while the Gray LevelMatrix Statistics algorithm provides low-dimension\r\nfeature vectors withmore time consumption andmore memory space. Furthermore, the characteristic of rotational invariance, both\r\nin the Fourier Descriptors algorithm and the Geometric Moment Vector algorithm, may result in improper pattern recognition of\r\nindicator card data when used for sucker-rod pump working condition diagnosis.
Loading....