The optimal design method for an open Magnetic Resonance Imaging （MRI） superconducting magnet with an active shielding configuration is proposed. Firstly, three pairs of current rings are employed as seed coils. By ...The optimal design method for an open Magnetic Resonance Imaging （MRI） superconducting magnet with an active shielding configuration is proposed. Firstly, three pairs of current rings are employed as seed coils. By optimizing the homogeneity of Diameter Sphere Voltnne （DSV）, the positions and currents of the seed coils will be obtained. Secondly, according to the positions and currents of the seed coils, the current density of superconducting wires is determined, and then the original sections for the coils can be achieved. An optimization for the homogeneity based on the constrained nonlincar optimization method is employed to determine the coils with good homogeneity. Thirdly, the magnetic field generated by previous coils is set as the background field, then add two coils with reverse current, and optimize the stray field line of 5 Gauss in a certain scope. Finally, a further optimization for the homogeneity is used to get Final coils. This method can also be used in the design of other axisynmaetfic superconducting MRI magnets.展开更多
Multi-label learning is more complicated than single-label learning since the semantics of the instances are usually overlapped and not identical.The effectiveness of many algorithms often fails when the correlations ...Multi-label learning is more complicated than single-label learning since the semantics of the instances are usually overlapped and not identical.The effectiveness of many algorithms often fails when the correlations in the feature and label space are not fully exploited.To this end,we propose a novel non-negative matrix factorization(NMF)based modeling and training algorithm that learns from both the adjacencies of the instances and the labels of the training set.In the modeling process,a set of generators are constructed,and the associations among generators,instances,and labels are set up,with which the label prediction is conducted.In the training process,the parameters involved in the process of modeling are determined.Specifically,an NMF based algorithm is proposed to determine the associations between generators and instances,and a non-negative least square optimization algorithm is applied to determine the associations between generators and labels.The proposed algorithm fully takes the advantage of smoothness assumption,so that the labels are properly propagated.The experiments were carried out on six set of benchmarks.The results demonstrate the effectiveness of the proposed algorithms.展开更多
基金supported by the National Natural Science Foundation of China(No.50577063)
文摘The optimal design method for an open Magnetic Resonance Imaging （MRI） superconducting magnet with an active shielding configuration is proposed. Firstly, three pairs of current rings are employed as seed coils. By optimizing the homogeneity of Diameter Sphere Voltnne （DSV）, the positions and currents of the seed coils will be obtained. Secondly, according to the positions and currents of the seed coils, the current density of superconducting wires is determined, and then the original sections for the coils can be achieved. An optimization for the homogeneity based on the constrained nonlincar optimization method is employed to determine the coils with good homogeneity. Thirdly, the magnetic field generated by previous coils is set as the background field, then add two coils with reverse current, and optimize the stray field line of 5 Gauss in a certain scope. Finally, a further optimization for the homogeneity is used to get Final coils. This method can also be used in the design of other axisynmaetfic superconducting MRI magnets.
基金support of the National Natural Science Foundation of China(Grant Nos.61402076,61572104,61103146)the Fundamental Research Funds for the Central Universities(DUT17JC04)the Project of the Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education,Jilin University(93K172017K03).
文摘Multi-label learning is more complicated than single-label learning since the semantics of the instances are usually overlapped and not identical.The effectiveness of many algorithms often fails when the correlations in the feature and label space are not fully exploited.To this end,we propose a novel non-negative matrix factorization(NMF)based modeling and training algorithm that learns from both the adjacencies of the instances and the labels of the training set.In the modeling process,a set of generators are constructed,and the associations among generators,instances,and labels are set up,with which the label prediction is conducted.In the training process,the parameters involved in the process of modeling are determined.Specifically,an NMF based algorithm is proposed to determine the associations between generators and instances,and a non-negative least square optimization algorithm is applied to determine the associations between generators and labels.The proposed algorithm fully takes the advantage of smoothness assumption,so that the labels are properly propagated.The experiments were carried out on six set of benchmarks.The results demonstrate the effectiveness of the proposed algorithms.
