A type of wavelet neural network, in which the scale function isadopted only, is proposed in this paper for non-linear dynamicprocess modelling. Its network size is decreased significantly andthe weight coefficients c...A type of wavelet neural network, in which the scale function isadopted only, is proposed in this paper for non-linear dynamicprocess modelling. Its network size is decreased significantly andthe weight coefficients can be estimated by a linear algorithm. Thewavelet neural network holds some advantages superior to other typesof neural networks. First, its network structure is easy to specifybased on its theoretical analysis and intuition. Secondly, networktraining does not rely on stochastic gradient type techniques andavoids the problem of poor convergence or undesirable local minima.展开更多
In marine seismic exploration,especially in deep-water and hard ocean-bottom cases,free-surface multiples are strongly developed.Compared with primary waves,the wider illumination aperture of the multiples is benefici...In marine seismic exploration,especially in deep-water and hard ocean-bottom cases,free-surface multiples are strongly developed.Compared with primary waves,the wider illumination aperture of the multiples is beneficial for high-resolution seismic imaging.In this study,by introducing a new compound source composed of primaries and free-surface multiples and by ignoring internal multiples,we derive a new linearized forward problem(free-surface-multiple prediction model)under a weak-scattering assumption(i.e.,first-order Born approximation).On the basis of the new linearized problem,we propose a joint inversion-imaging method by simultaneously using the primaries and free-surface multiples under the general framework of least square inversion.To eliminate the crosstalk artifacts introduced by the cross-correlation of multiples with different orders,we prove that the crosstalk artifacts can be gradually eliminated during the inversion if a proper step length is selected.Synthetic-andfield-data tests demonstrate the effectiveness of the proposed method.展开更多
The Gaussian vortex beam is assumed to be linearly polarized.The analytical expression of the electric field of a linearly polarized Gaussian vortex beam propagating in free space is derived by using the vectorial Ray...The Gaussian vortex beam is assumed to be linearly polarized.The analytical expression of the electric field of a linearly polarized Gaussian vortex beam propagating in free space is derived by using the vectorial Rayleigh-Sommerfeld integral formulae.The propagating magnetic field of the linearly polarized Gaussian vortex beam is presented by taking the curl of the electric field.By employing the electromagnetic field of the linearly polarized Gaussian vortex beam beyond the paraxial approximation,the analytical expression of the angular momentum density of the linearly polarized Gaussian vortex beam is derived.The three components of the angular momentum density of a linearly polarized Gaussian vortex beam are demonstrated in the reference plane.The effects of the linearly polarized angle and the topological charge on the three components of the angular momentum density are investigated.To acquire the more longitudinal angular momentum density requires such an optimal choice that the linearly polarized angle is set to be zero and the topological charge increases.This research is useful to the optical trapping,the optical guiding,and the optical manipulation.展开更多
基金Supported by the Eu Information Technologies Programme Project(No. 22416) and National High Tech R&D Project(863/Computer Integrated Manufacture System AA413130) of China.
文摘A type of wavelet neural network, in which the scale function isadopted only, is proposed in this paper for non-linear dynamicprocess modelling. Its network size is decreased significantly andthe weight coefficients can be estimated by a linear algorithm. Thewavelet neural network holds some advantages superior to other typesof neural networks. First, its network structure is easy to specifybased on its theoretical analysis and intuition. Secondly, networktraining does not rely on stochastic gradient type techniques andavoids the problem of poor convergence or undesirable local minima.
基金the sponsors of the WPI group for their financial supportfinancially supported by the National Key R&D Program of China (Grant Number: 2018YFA0702503, 2019YFC0312004)+2 种基金National Natural Science Foundation of China (Grant Number: 41774126)Southern Marine Science and Engineering Guangdong Laboratory (Zhanjiang) (ZJW-2019-04)National Science and Technology Major Project of China (Grant Number: 2016ZX05024-001, 2016ZX05006-002)。
文摘In marine seismic exploration,especially in deep-water and hard ocean-bottom cases,free-surface multiples are strongly developed.Compared with primary waves,the wider illumination aperture of the multiples is beneficial for high-resolution seismic imaging.In this study,by introducing a new compound source composed of primaries and free-surface multiples and by ignoring internal multiples,we derive a new linearized forward problem(free-surface-multiple prediction model)under a weak-scattering assumption(i.e.,first-order Born approximation).On the basis of the new linearized problem,we propose a joint inversion-imaging method by simultaneously using the primaries and free-surface multiples under the general framework of least square inversion.To eliminate the crosstalk artifacts introduced by the cross-correlation of multiples with different orders,we prove that the crosstalk artifacts can be gradually eliminated during the inversion if a proper step length is selected.Synthetic-andfield-data tests demonstrate the effectiveness of the proposed method.
基金supported by the National Natural Science Foundation of China(Grant Nos.61178016 and 10974179)Zhejiang Provincial Natural Science Foundation of China(Grant No.Y1090073)
文摘The Gaussian vortex beam is assumed to be linearly polarized.The analytical expression of the electric field of a linearly polarized Gaussian vortex beam propagating in free space is derived by using the vectorial Rayleigh-Sommerfeld integral formulae.The propagating magnetic field of the linearly polarized Gaussian vortex beam is presented by taking the curl of the electric field.By employing the electromagnetic field of the linearly polarized Gaussian vortex beam beyond the paraxial approximation,the analytical expression of the angular momentum density of the linearly polarized Gaussian vortex beam is derived.The three components of the angular momentum density of a linearly polarized Gaussian vortex beam are demonstrated in the reference plane.The effects of the linearly polarized angle and the topological charge on the three components of the angular momentum density are investigated.To acquire the more longitudinal angular momentum density requires such an optimal choice that the linearly polarized angle is set to be zero and the topological charge increases.This research is useful to the optical trapping,the optical guiding,and the optical manipulation.
