Abstract: A multi-pole synchronous pulmonary artery radiofrequency ablation catheter may comprise a control handle, a catheter body and an annular ring. One end of the catheter body may be flexible, and the flexible end of the catheter body may be connected to the annular ring. The other end of the catheter body may be connected to the control handle. A shape memory wire may be arranged in the annular ring. One end of the shape memory wire may extend to an end of the annular ring and the other end of the shape memory wire may pass through a root of the annular ring and be fixed on the flexible end of the catheter body. The annular ring may be provided with an electrode group. The device possesses advantages of simple operation, short operation time and controllable precise ablation. The device can be used to treat pulmonary hypertension or heart failure with pulmonary denervation.

Abstract: Naturally-occurring cellulose-based material, such as wood, bamboo, grass, or reed, can be subjected to one or more chemical treatments to remove at least some lignin therefrom. The resulting partially-delignified material can be partially dried or fully dried and then rehydrated to yield a moldable cellulose-based material. The moldable material can be formed from a substantially flat planar configuration into a non-planar three-dimensional configuration. Once formed into a desired configuration, the moldable material can be fully dried to set its shape, thereby forming a rigid molded piece. In some embodiments, the molded piece can be used as a structural material, for example, to form a load-bearing structure or part of a composite load-bearing structure.

Abstract: A preparing process for pressing particulate aggregates into a slab and a shaping apparatus are described. The preparing process includes: mixing particulate aggregates of different sizes and a liquid material as per a preset proportion to obtain a mixture; spreading the mixture flatly in a preset thickness onto a conveyor belt or into a mould frame; vacuumizing the mixture; shaping, wherein a pressing plate is disposed covering the mixture during the shaping procedure to constantly apply a downward pressure against the mixture; a ram disposed over the pressing plate continuously smashes the pressing plate, wherein during ram smashing, the pressure created by the pressing plate forms a continuous holding force to the mixture, the smashing force created from up-down motion of the ram is transferred to the mixture via the pressing plate. The changes to the abstract are shown below: A preparing process for pressing particulate aggregates into a slab and a shaping apparatus are described.

Abstract: A desulfurization wastewater zero discharge treatment method and system suitable for multiple working conditions. A tail flue of a boiler and a bottom outlet of a wastewater drying tower are both communicated with an inlet of a dust collector; an outlet of the dust collector is communicated with flue gas inlets of a wastewater concentration tower and a desulfurization absorption tower; the wastewater concentration tower is communicated with the desulfurization absorption tower; the desulfurization absorption tower is communicated with a chimney; the desulfurization absorption tower is communicated with a gypsum cyclone; the gypsum cyclone is communicated with a filtrate water tank; the gypsum cyclone is communicated with a gypsum dewatering machine; the gypsum dewatering machine is communicated with a gas liquid separating tank; and a flue gas port of the tail flue of the boiler is communicated with the flue gas inlet of the wastewater drying tower.

Abstract: Systems and methods for predicting performance of a modulation system are provided. A neural network model is trained using performance information of a source system. The neural network model is modified with transferable knowledge about a target system to be evaluated. The neural network model is tuned using specific characteristics of the target system to create a source-based target model. The target system performance is evaluated using the source-based target model to predict system performance of the target system.

Abstract: This application discloses a resource pushing method performed by a computer device. The method includes: obtaining a target recommendation model and a preference feature and a candidate resource set corresponding to a target object, the preference feature including at least a channel preference feature and a content preference feature; obtaining at least one target resource from the candidate resource set based on the target recommendation model and the preference feature; and pushing the at least one target resource to the target object. Such a resource pushing process integrates preferences of the target object in different dimensions, so that the target resource pushed to the target object not only conforms to channel preferences of the target object, but also conforms to content references of the target object, which is beneficial to improving the resource pushing effect, and further increasing the click-through rates (CTRs) of the pushed resources.

Abstract: Reflecting light beams off of microscale three-dimensional (3D) freeform surfaces can yield highly efficient coupling into and out of optical waveguides, optical fibers, and photonic chips. The structure of the 3D freeform reflective surface determines the shape of the reflected beam. This allows freeform reflectors to control the mode profile, rotation angle, and divergence angle of light beams. Control of beam shape enables mode matching between source output mode and target input mode, which results in low-loss optical coupling. An inventive freeform reflective surface can direct light beams in plane or out of plane via specular reflection or total internal reflection. A photonic integrated circuit with this type of freeform optical coupler can operate with a bandwidth range of at least 400 nm, potentially encompassing all visible or telecommunications wavelengths, and can be volume manufactured in photonic chips.

Abstract: Aspects of the present disclosure describe a method for digital coherent transmission systems that advantageously provides low-complexity, single-step nonlinearity compensation based on artificial intelligence (AI) implemented in a deep neuron network (DNN).

Abstract: A multi-pole synchronous pulmonary artery radiofrequency ablation catheter may comprise a control handle, a catheter body and an annular ring. One end of the catheter body may be flexible, and the flexible end of the catheter body may be connected to the annular ring. The other end of the catheter body may be connected to the control handle. A shape memory wire may be arranged in the annular ring. One end of the shape memory wire may extend to an end of the annular ring and the other end of the shape memory wire may pass through a root of the annular ring and be fixed on the flexible end of the catheter body. The annular ring may be provided with an electrode group. The device possesses advantages of simple operation, short operation time and controllable precise ablation. The device can be used to treat pulmonary hypertension with pulmonary denervation.

Abstract: A NAT method, apparatus and device are provided. According to the method, a target IP address and its reference port are obtained from a NAT resource pool, the reference port being a port in a corresponding consecutive port range. A first five-tuple is generated based on the target IP address, the reference port and an original five-tuple of the packet, and a second five-tuple is obtained by masking first-class bits of the reference port of the first five-tuple. Based on a hash result of the second five-tuple, a target bit indicating a non-conflicting state is determined from a pre-constructed bitmap. The state indicated by the target bit is set to be a conflicting state, and a target five-tuple is generated based on the target bit. The target five-tuple and the original five-tuple are recorded in a session table, and the packet is NAT-processed based on the target five-tuple.

Abstract: A multi-pole synchronous pulmonary artery radiofrequency ablation catheter may comprise a control handle, a catheter body and an annular ring. One end of the catheter body may be flexible, and the flexible end of the catheter body may be connected to the annular ring. The other end of the catheter body may be connected to the control handle. A shape memory wire may be arranged in the annular ring. One end of the shape memory wire may extend to an end of the annular ring and the other end of the shape memory wire may pass through a root of the annular ring and be fixed on the flexible end of the catheter body. The annular ring may be provided with an electrode group. The device possesses advantages of simple operation, short operation time and controllable precise ablation. The device can be used to treat pulmonary hypertension with pulmonary denervation.

Abstract: A NAT method, apparatus and device are provided. According to the method, a target IP address and its reference port are obtained from a NAT resource pool, the reference port being a port in a corresponding consecutive port range. A first five-tuple is generated based on the target IP address, the reference port and an original five-tuple of the packet, and a second five-tuple is obtained by masking first-class bits of the reference port of the first five-tuple. Based on a hash result of the second five-tuple, a target bit indicating a non-conflicting state is determined from a pre-constructed bitmap. The state indicated by the target bit is set to be a conflicting state, and a target five-tuple is generated based on the target bit. The target five-tuple and the original five-tuple are recorded in a session table, and the packet is NAT-processed based on the target five-tuple.

Abstract: Aspects of the present disclosure describe systems, methods. and structures in which a hybrid neural network combining a CNN and several ANNs are shown useful for predicting G-ONSR for Ps-256QAM raw data in deployed SSMF metro networks with 0.27 dB RMSE. As demonstrated, the CNN classifier is trained with 80.96% testing accuracy to identify channel shaping factor. Several ANN regression models are trained to estimate G-OSNR with 0.2 dB for channels with various constellation shaping. Further aspects include the tuning of existing optical networks and the characterization of retrofit/upgraded optical networks to estimate capacity—both aspects employing our inventive hybrid neural network methodology.

Abstract: Aspects of the present disclosure describe systems, methods, and structures that advantageously amplify optical signals through the effect of optical pump signals generated by a multicore laser diode and multicore rare-earth doped optical fiber in optical communication with a 3D waveguide structure and a multicore input signal fiber providing a plurality of optical signals for amplification.

Abstract: Aspects of the present disclosure describe systems, methods. and structures in which a hybrid neural network combining a CNN and several ANNs are shown useful for predicting G-ONSR for Ps-256QAM raw data in deployed SSMF metro networks with 0.27 dB RMSE. As demonstrated, the CNN classifier is trained with 80.96% testing accuracy to identify channel shaping factor. Several ANN regression models are trained to estimate G-OSNR with 0.2 dB for channels with various constellation shaping. Further aspects include the tuning of existing optical networks and the characterization of retrofit/upgraded optical networks to estimate capacity—both aspects employing our inventive hybrid neural network methodology.

Abstract: A Network Address Translation (NAT) method, apparatus and device are provided. Based on the method, a target IP address and its reference port are obtained from a NAT resource pool, wherein the reference port is a port in a consecutive port range of the target IP address; a first five-tuple corresponding to a packet is generated based on the target IP address, the reference port and an original five-tuple of the packet, and a second five-tuple is obtained by masking first-class bits of two classes of ports of the first five-tuple respectively; a target five-tuple is determined in a plurality of consecutive hash buckets of a hash table based on a hash result of the second five-tuple; and the target five-tuple and the original five-tuple are recorded in the hash table and a corresponding result table, and the packet is NAT-processed based on the target five-tuple.

Abstract: A NAT method, apparatus and device are provided. According to the method, a target IP address and its reference port are obtained from a NAT resource pool, the reference port being a port in a corresponding consecutive port range. A first five-tuple is generated based on the target IP address, the reference port and an original five-tuple of the packet, and a second five-tuple is obtained by masking first-class bits of the reference port of the first five-tuple. Based on a hash result of the second five-tuple, a target bit indicating a non-conflicting state is determined from a pre-constructed bitmap. The state indicated by the target bit is set to be a conflicting state, and a target five-tuple is generated based on the target bit. The target five-tuple and the original five-tuple are recorded in a session table, and the packet is NAT-processed based on the target five-tuple.

Abstract: A Network Address Translation (NAT) method, apparatus and device are provided. Based on the method, a target IP address and its reference port are obtained from a NAT resource pool, wherein the reference port is a port in a consecutive port range of the target IP address; a first five-tuple corresponding to a packet is generated based on the target IP address, the reference port and an original five-tuple of the packet, and a second five-tuple is obtained by masking first-class bits of two classes of ports of the first five-tuple respectively; a target five-tuple is determined in a plurality of consecutive hash buckets of a hash table based on a hash result of the second five-tuple; and the target five-tuple and the original five-tuple are recorded in the hash table and a corresponding result table, and the packet is NAT-processed based on the target five-tuple.