Detailed electrostatic potential (ESP) analyses were performed to compare the directionality of halogen bonds with those of hydrogen bonds and lithium bonds. To do this, the interactions of HOOOH with the molecules XF (X = Cl, Br, H, Li) were investigated. For each molecule, the percentage of the van der Waals (vdW) molecular surface that intersected with the ESP surface was used to roughly quantify the directionality of the halogen/hydrogen/lithium bond associated with the molecule. The size of the region of intersection was found to increase in the following order: ClF < BrF < HF < LiF. The maximum ESP in the region of intersection, V S, max, was observed to become more positive according to the sequence ClF < BrF < HF < LiF. For ClF and BrF, the positive electrostatic potential was concentrated in a very small region of the vdW molecular surface. On the other hand, for HF and LiF, the positive electrostatic potential was more diffusely scattered across the vdW surface than for ClF and BrF. Also, the optimized geometries of the dipolymers HOOOH··· XF (X = Cl, Br, H, Li) indicated that halogen bonds are more directional than hydrogen bonds and lithium bonds, consistent with the results of ESP analyses.

An accurate electrocardiogram (ECG) beat classification can benefit the diagnosis of the cardiovascular disease. Deep convolutional neural networks (CNN) can automatically extract valid features from data, which is an effective way for the classification of the ECG beats. However, the fully-connected layer in CNNs requires a fixed input dimension, which limits the CNNs to receive fixed-scale inputs. Signals of different scales are generally processed into the same size by segmentation and downsampling. If information loss occurs during a uniformly-sized process, the classification accuracy will ultimately be affected. To solve this problem, this paper constructs a new CNN framework spatial pyramid pooling (SPP) method, which solves the deficiency caused by the size of input data. The Massachusetts Institute of Technology-Biotechnology (MIT-BIH) arrhythmia database is employed as the training and testing data for the classification of heartbeat signals into six categories. Compared with the traditional method, which may lose a large amount of important information and easy to be over-fitted, the robustness of the proposed method can be guaranteed by extracting data features from different sizes. Experimental results show that the proposed architecture network can extract more high-quality features and exhibits higher classification accuracy (94%) than the traditional deep CNNs (90.4%).

For nuclear reactor physics, uncertainties in the multigroup cross sections inevitably exist, and these uncertainties are considered as the most significant uncertainty source. Based on the home-developed 3D high-fidelity neutron transport code HNET, the perturbation theory was used to directly calculate the sensitivity coefficient of keff to the multigroup cross sections, and a reasonable relative covariance matrix with a specific energy group structure was generated directly from the evaluated covariance data by using the transforming method. Then, the “Sandwich Rule” was applied to quantify the uncertainty of keff. Based on these methods, a new SU module in HNET was developed to directly quantify the keff uncertainty with one-step deterministic transport methods. To verify the accuracy of the sensitivity and uncertainty analysis of HNET, an infinite-medium problem and the 2D pin-cell problem were used to perform SU analysis, and the numerical results demonstrate that acceptable accuracy of sensitivity and uncertainty analysis of the HNET are achievable. Finally, keff SU analysis of a 3D minicore was analyzed by using the HNET, and some important conclusions were also drawn from the numerical results.

No abstract