Abstract: A method and a system for ship stability prediction by weighted fusion of RBFNN and random forest based on GD are provided. Firstly, input characteristics when predicting failure probabilities under different failure modes are determined through prior knowledge. Secondly, a mean square error of k-fold cross-validation is used as performance evaluation criterion of the RBFNN and the RF to search for model capacities of the RBFNN and the RF. Then, network parameters of the RBFNN are updated. Multiple random sample sets are generated using a bootstrap sampling method and are parallelly trained to generate multiple regression trees. A Gini index is used as an attribute division index, and a prediction result of the random forest is obtained. Finally, weight coefficients are introduced for weighted fusion of prediction results of the RBFNN and the RF. The weight coefficient is obtained by solving through iterative optimization of the gradient descent.

Abstract: The present invention discloses a method and a system for prostate multi-modal MR image classification based on a foveated residual network, the method comprising: replacing convolution kernels of a residual network using blur kernels in a foveation operator, thereby constructing a foveated residual network; training the foveated residual network using prostate multi-modal MR images having category labels, to obtain a trained foveated residual network; and classifying, using the foveated residual network, a prostate multi-modal MR image to be classified, so as to obtain a classification result. In the present invention, a foveation operator is designed based on human visual characteristics, blur kernels of the operator are extracted and used to replace convolution kernels in a residual network, thereby constructing a foveated deep learning network which can extract features that conform to the human visual characteristics, thereby improving the classification accuracy of prostate multi-modal MR images.

Abstract: The disclosure belongs to the technical field of capsule robotics, and in particular, relates to a magnetically driven capsule robot with controllable drug administration and sampling functions and a fabrication method thereof. A magnetically driven capsule robot provided by the disclosure includes a robot body and a magnetic field-driven control module. The robot body includes a capsule shell, two ends of the capsule shell are provided with hollow air chambers, and a carrier chamber is disposed between the hollow air chambers disposed on the two ends. A substance transfer channel, a magnetic lock disposed around the substance transfer channel, and a magnetic soft switch valve matched with the magnetic lock are disposed on a side wall of the carrier chamber. The soft valve automatically closes due to the magnetic attraction force between the lock and valve, and can be flexibly opened under a magnetic torque acting on the valve.

Abstract: This disclosure relates to a dielectric layer response-based field effect transistor photodetector, comprising a gate, a photoelectric response composite dielectric layer, a carrier transport layer, a source, and a drain. The composite dielectric layer is formed by compounding a photoelectric response dielectric and a charge blocking insulating dielectric; the carrier transport layer is used for transporting electrons or holes, the photoelectric response dielectric is used for absorbing light under illumination to generate electrons, holes or excitons, and the charge blocking insulating dielectric is used for limiting passage of the electrons, holes or excitons.

Abstract: The present invention provides an anti-trapdoor-leakage on-chain data restoration system, at least comprising: a blockchain node, for broadcasting transaction data of a request-initiating person to blockchain nodes and proposer nodes in other groups, respectively; and a proposer node, for performing computation of a Chameleon-Hash function using a key set that is generated by a key-generating module provided in the proposer node, packaging the transaction data to generate a new block, and distributing the new block to all the blockchain nodes so that the blockchain nodes update their respective underlying ledgers according to the new blocks broadcasted by the proposer. The system of the present invention not only realizes such functions as restoration and editing of the transaction data, but also protects operational security and reliability of blockchains.

Abstract: The present disclosure relates to the technical field of heat dissipation for electronic devices, and discloses a manifold microchannel heat sink based on directional optimization of hotspot areas, including a power module, a heat dissipation substrate, a cold source device, and a diverter manifold; the heat dissipation substrate is arranged on the power module, a microchannel is provided on a side of the heat dissipation substrate away from the power module and on a back of a heat generating area of the power module for carrying the power module and dissipating heat from the power module. According to the manifold microchannel heat sink based on the directional optimization of hotspot areas, the local heat dissipation performance of the power module may be preliminarily changed by optimizing the microchannel on the heat dissipation substrate, which avoids poor heat dissipation temperature uniformity in a multi-heat source system.

Abstract: A method includes calculating an initial phase ?gv of each air gap working magnetic field Bgv required to generate positive back EMF based on certain parameters of each armature harmonic vth; generating Bgv and maximizing a sum of equivalent air gap flux density amplitude Beqv of each of the air gap working magnetic fields Bgv while calculating a phase ?mv and a pole arc coefficient ?v corresponding to each permeance harmonic; designing a number, a position(s) and a length(s) along a circumference of modulation teeth corresponding to each of the permeance harmonics based on the phase ?mv and the pole arc coefficient ?v, such that a generated permeance model is consistent with the phase ?mv of each of the permeance harmonics; and optimizing radial dimensions of modulation teeth corresponding to each of the permeance harmonics to maximize the sum of equivalent air gap flux density amplitude Beqv.

Abstract: The present invention provides a Ga-based van der Waals room-temperature ferromagnetic crystal material, preparation and use thereof, which belong to the technical field of nano magnetic material preparation. The materials include Fe3-aGabTe2 (a=?0.3 to 0.1, b=0.8 to 1.2) and Fe5-c GeGadTe2 (c=?0.2 to 0.2, d=0.01 to 0.5). The growth method of Fe3-aGabTe2 (a=?0.3 to 0.1, b=0.8 to 1.2) is a self-flux method, using excess Ga and Te as flux to grow crystals. The growth method of Fe5-c GeGadTe2 (c=?0.2 to 0.2, d=0.01 to 0.5) uses iodine as a transport agent to grow crystals. The Ga-based van der Waals room-temperature ferromagnetic crystal Fe3-a GabTe2 (a=?0.3 to 0.1, b=0.8 to 1.2) and Fe5-cGeGadTe2 (c=?0.2 to 0.2, d=0.01 to 0.5) materials have Curie temperature of 330 K to 367 K and 320 K to 345 K, and the saturation magnetic moments are 50 emu/g to 57.2 emu/g and 80 emu/g to 88.5 emu/g, respectively.

Abstract: A method for operating a dynamic memory is provided, and the method includes the following steps. A refresh operation is performed on the dynamic memory according to predetermined interval time T, an operation command is received in real time at the same time, a read operation is performed on a selected memory cell according to position information of the selected memory cell in the operation command when the operation command is received, and state data read in the read operation is temporarily stored in a read buffer. The interval time T is less than time t required for a voltage value of a capacitor in the memory cell to drop to a critical capacitor voltage value for the read operation to correctly read the state data of the memory cell during a write operation. According to operation command type information in the operation command, corresponding operations are performed on the selected memory cell.

Abstract: The present invention discloses an electromagnetic wave spatial analysis method based on multi-level dipole group modeling.

Abstract: A method for operating a dynamic memory is provided, and the method includes the following steps. A refresh operation is performed on the dynamic memory according to predetermined interval time T, an operation command is received in real time at the same time, a read operation is performed on a selected memory cell according to position information of the selected memory cell in the operation command when the operation command is received, and state data read in the read operation is temporarily stored in a read buffer. The interval time T is less than time t required for a voltage value of a capacitor in the memory cell to drop to a critical capacitor voltage value for the read operation to correctly read the state data of the memory cell during a write operation. According to operation command type information in the operation command, corresponding operations are performed on the selected memory cell.

Abstract: A partition design and molding method for an acetabular cup prosthesis with a porous surface includes the following: modeling of a spherical shell of an acetabular cup; cutting of the spherical shell of the acetabular cup into equal parts; integrated design on a porous layer of a surface of the partitioned acetabular cup prosthesis; reorganization of the partitioned acetabular cup prosthesis; construction of a hemispherical acetabular cup prosthesis; and 3D printing and manufacturing of the acetabular cup prosthesis. Through the disclosure, the formed porous layer is perpendicular to the surface of the acetabular cup, so favorable mechanical properties are provided. A fully interconnected branch-rod structure is provided, and there is no isolated rod, so that inflammation or even re-revision led by the detachment of isolated rods during long-term postoperative activities is prevented.

Abstract: The disclosure belongs to the field of image segmentation in medical image processing and discloses a three-dimensional medical image segmentation method and system based on short-term and long-term memory self-attention models, in which the method can segment a target area image in the medical image, which includes the following. (1) A training set sample is established. (2) Processing is performed on the original three-dimensional medical image to be segmented to obtain a sample to be segmented. (3) A three-dimensional medical image segmentation network based on short-term and long-term memory self-attention is established and trained. (4) The sample to be segmented is input to the network, and then a segmentation result of the target area in the sample to be segmented is output. By combining CNN and Transformer, a new model for accurate real-time segmentation of the target area (such as a tumor) in the three-dimensional medical image is obtained.

Abstract: Disclosed is a selective field-assisted machining system. The system includes a micron-level high-speed identification module, an in-situ laser assisted module, an ultrasonic vibration module, an energy field loading high-speed control module, and a diamond tool. The micron-level high-speed identification module is used to quickly identify the type of a material substrate of a workpiece to be processed, process the identification information into a corresponding control signal, and send same to the energy field loading high-speed control module to implement selective processing of the workpiece to be processed, i.e. to process brittle particles using in-situ laser assisted machining and to process a soft metal substrate using ultrasonic vibration processing. In the present invention, ultra-precision cutting of brittle particles and a soft metal substrate can be completed at the same time in a single processing process.

Abstract: The disclosure discloses a low-loss transmission line structure, which belongs to the field of radio frequency transmission lines and includes at least two metal layers stacked in a vertical manner. A dielectric layer is filled between the metal layers. The metal layers include a signal transmission strip in a middle portion. Ground strips are provided on both sides of the signal transmission strip. Through holes are evenly distributed on the dielectric layer, and the signal transmission strips on each of the metal layers are connected through the through holes to form a signal transmission line. The ground strips on each metal layer are connected through the through holes.

Abstract: An optical structure includes an input waveguide, a beam splitter, a first arm, a second arm, a beam combiner, and an output waveguide. The first arm and the second arm each include a conventional waveguide region and a modulation waveguide region. The modulation waveguide region of the first arm includes a first modulation waveguide region and a second modulation waveguide region. The modulation waveguide region of the second arm includes a third modulation waveguide region and a fourth modulation waveguide region. The electrical structure includes a traveling wave electrode including a ground-signal-signal-ground electrode structure. The traveling wave electrode includes a signal input region, a modulation electrode region, and a matching resistor region. The modulation electrode region includes a first modulation electrode region and a second modulation electrode region. The first modulation electrode region is connected to the second modulation electrode region.

Abstract: The present disclosure provides a superconducting power generation device and a power generation method. The device includes a superconductor, a conductive coil, a permanent magnet and a cooling medium. When ambient temperature is lower than its superconducting critical temperature, the superconductor, made of the second-type superconducting material, is capable of generating a magnetic levitation force to an outer permanent magnet and levitate it. When an external force is applied to the permanent magnet, its position changes compared to the conductive coil, which affects the magnetic flux passing through the coil and induces the generation of electromotive force in the coil, thereby converting mechanical energy to electric energy.

Abstract: The present invention relates a device for improving robustness of deep-learning based detection of source-code vulnerability, the device at least comprises a code-converting module, a mapping module, and a neural-network module, wherein the mapping module is in data connection with the code-converting module, the mapping module is in data connection with the neural-network module, respectively, and the neural-network module includes at least two first classifiers, based on a received first training program source code, the mapping module maps a plurality of code snippets, and the neural-network module trains the at least two first classifiers according to a first sample vector. The present invention improves the robustness of detection of source-code vulnerability by performing classification training on the feature generators and the classifiers.

Abstract: A partition design and molding method for an acetabular cup prosthesis with a porous surface includes the following: modeling of a spherical shell of an acetabular cup; cutting of the spherical shell of the acetabular cup into equal parts; integrated design on a porous layer of a surface of the partitioned acetabular cup prosthesis; reorganization of the partitioned acetabular cup prosthesis; construction of a hemispherical acetabular cup prosthesis; and 3D printing and manufacturing of the acetabular cup prosthesis. Through the disclosure, the formed porous layer is perpendicular to the surface of the acetabular cup, so favorable mechanical properties are provided. A fully interconnected branch-rod structure is provided, and there is no isolated rod, so that inflammation or even re-revision led by the detachment of isolated rods during long-term postoperative activities is prevented.

Abstract: An intra-frame predictive coding method, medium and system for a 360-degree video belong to video coding field are provided. The method includes: inputting features of a coding unit (CU) block to be partitioned into a neural network when intra-frame block partition (IFBP) is performed on a current CU block (CCUB), skipping a partition mode that the output of the neural network is smaller than threshold, and obtaining a plurality of IFBP schemes; determining a length n of a rough mode decision candidate list according to a CU feature and a texture feature of the sub-CU block when an intra-frame angle mode is decided to decide the intra-frame angle mode and to fill pixels in each IFBP scheme; calculating rate distortion optimization (RDO) loss between each IFBP scheme and the CCUB after pixel filling and performing intra-frame coding predictive coding on the CCUB based on the scheme.