Abstract: A method for manufacturing a core-shell coaxial gallium nitride (GaN) piezoelectric nanogenerator is provided. A mask covering a center part of a gallium nitride wafer is removed. An electrodeless photoelectrochemical etching is performed on the gallium nitride wafer to form a primary GaN nanowire array on a surface of the gallium nitride wafer. A precious metal layer provided on the surface of the gallium nitride wafer is removed and an alumina layer is deposited on the surface of the gallium nitride wafer to cover the primary GaN nanowire array to obtain a core-shell coaxial GaN nanowire array. A first conductive layer is provided on a flexible substrate to which the core-shell coaxial GaN nanowire array is transferred. A second conductive layer is provided at a top end of the core-shell coaxial GaN nanowire array, and is connected to an external circuit to obtain the core-shell coaxial GaN piezoelectric nanogenerator.

Abstract: Provided are a search request processing method, an apparatus, a computer device, a storage medium, a computer product, and a computer program. The method includes: in response to a search instruction, sending a search request comprising a target question; acquiring a plurality of sets of answer information corresponding to the target question, wherein each set of answer information includes answer summary information and a multimedia content supporting the answer summary information; the genre of the multimedia content is determined on the basis of a category of the target question, and the answer summary information is extracted on the basis of the multimedia content; and displaying the answer summary information in each set of answer information on a search result page, and displaying, when any piece of answer summary information is selected, at least part of the multimedia content supporting the selected answer summary information.

Abstract: The sense amplifier includes: an amplification module configured to amplify a voltage transmitted by a bit line or a reference bit line, when the sense amplifier is at an amplification stage; a first switch module configured to control the amplification module to be disconnected from the reference bit line, when the sense amplifier performs a read operation for the bit line and is at the amplification stage. In the disclosure, the power consumption of the sense amplifier may be reduced.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for jointly training an image embedding model and a text embedding model. In one aspect, a method comprises: processing data from a historical query log of a search system to generate a candidate set of training examples, wherein each training example comprises: (i) a search query comprising a sequence of one or more words, (ii) an image, and (iii) selection data characterizing how often users selected the image in response to the image being identified by a search result for the search query; selecting a plurality of training examples from the candidate set of training examples; and using the training data to jointly train the image embedding model and the text embedding model.

Abstract: A method of forming a 1T?-phase transition metal dichalcogenide monolayer. A transition metal precursor and a first solvent are mixed to form a first mixture. A non-oxygen chalcogen or non-oxygen chalcogen precursor and a second solvent are mixed to form a second mixture. The first mixture is rapidly injected into the second mixture at a temperature between approximately 250 and 350° C. to form a third mixture. 1T?-transition metal dichalcogenide monolayers are recovered from the third mixture. The transition metals may be molybdenum or tungsten, while the non-oxygen chalcogens may be sulfur, selenium, or tellurium. The 1T?-transition metal dichalcogenide monolayers can be grown on a variety of metal substrates to form metal@ 1T?-transition metal dichalcogenide monolayer heterostructures. A flexible 4H-Au@1T?-WS2/SiO2/PDMS SERS tape was fabricated to detect the SARS-CoV-2 spike protein. The tape is capable of attomole-level detection of SARS-CoV-2 spike protein, for real-time monitoring of COVID-19.

Abstract: A core-shell coaxial gallium nitride piezoelectric nanogenerator includes a core-shell coaxial gallium nitride nanowire array and a flexible substrate. A first conductive layer is provided on a surface of the flexible substrate. The core-shell coaxial gallium nitride nanowire array is fixed to the flexible substrate. A top end of the core-shell coaxial gallium nitride nanowire array is provided with a second conductive layer. The first conductive layer and the second conductive layer are both connected to an external circuit via a wire. A nanowire of the core-shell coaxial gallium nitride nanowire array is covered with an alumina layer. A method for preparing the core-shell coaxial gallium nitride piezoelectric nanogenerator is further provided. The gallium nitride nanowire array is formed by electrodeless photoelectrochemical etching.

Abstract: A system can arrange a plurality of transport services in a geographic region by matching received transport requests from requesters with standard vehicles to facilitate transport of the requesters to destinations indicated in the transport requests. The transport requests can comprise both requests for standard vehicles and high-capacity vehicle (HCV) requests from HCV requesters. Based on a set of transport services for HCV requesters, from the plurality of transport services, the system can determine over a specified duration of time, a set of demand data for HCV transport services in the geographic region. The system can then execute a route design model using the set of demand data to generate an HCV route network for the geographic region.

Abstract: The invention discloses a method and a device for constructing a SQL statement based on reinforcement learning, wherein the method includes: initializing an actor-critic network parameter; acquiring a sequence pair of natural language and real SQL statement from a data set; inputting a natural language sequence into an actor network encoder, and inputting a real SQL sequence into a critic network encoder; using an encoded hidden state as an initialized hidden state of a corresponding decoder; gradually predicting, by an actor network decoder, a SQL statement action, and inputting the SQL statement action to a critic network decoder and an environment to obtain a corresponding reward; and using a gradient descent algorithm to update the network parameters, and obtaining a constructing model of the natural language to the SQL statement after repeated iteration training.

Abstract: The present invention discloses a method for detecting abnormity in an unsupervised industrial system based on deep transfer learning. Labeled machine sensor sequence data from a source domain and unlabeled sensor sequence data from a target domain are used in the present invention to train an industrial system abnormal detection model with good generalization ability, and the industrial system abnormal detection model is trained and tested to finally generate a trained industrial system abnormity discrimination model. Using the model, received machine sensor sequence data can be analyzed and whether a machine is abnormal is discriminated.

Abstract: Disclosed is a device for measuring speed of a tennis ball, which includes: a housing, a power supply, a main control device, a sensor device, a transmission device, a clamping device and a limiting device. The clamping device is clamped at the serve opening of a tennis ball machine, which can detect the serve speed of a tennis ball accurately with simple operation. The transmission device is engaged with the second sensor device, and the housing is slidably connected with the second sensor device. The emitting ends of the second sensor device and the first sensor device are positioned in the same vertical plane. The limiting device is connected with the second sensor device to limit excessive operation of the transmission device.

Abstract: Disclosed is a device for measuring speed of a tennis ball, which includes: a housing, a power supply, a main control device, a sensor device, a transmission device, a clamping device and a limiting device. The clamping device is clamped at the serve opening of a tennis ball machine, which can detect the serve speed of a tennis ball accurately with simple operation. The transmission device is engaged with the second sensor device, and the housing is slidably connected with the second sensor device. The emitting ends of the second sensor device and the first sensor device are positioned in the same vertical plane. The limiting device is connected with the second sensor device to limit excessive operation of the transmission device.

Abstract: The invention discloses a method and a device for constructing a SQL statement based on reinforcement learning, wherein the method includes: initializing an actor-critic network parameter; acquiring a sequence pair of natural language and real SQL statement from a data set; inputting a natural language sequence into an actor network encoder, and inputting a real SQL sequence into a critic network encoder; using an encoded hidden state as an initialized hidden state of a corresponding decoder; gradually predicting, by an actor network decoder, a SQL statement action, and inputting the SQL statement action to a critic network decoder and an environment to obtain a corresponding reward; and using a gradient descent algorithm to update the network parameters, and obtaining a constructing model of the natural language to the SQL statement after repeated iteration training.

Abstract: The present invention discloses a method for detecting abnormity in an unsupervised industrial system based on deep transfer learning. Labeled machine sensor sequence data from a source domain and unlabeled sensor sequence data from a target domain are used in the present invention to train an industrial system abnormal detection model with good generalization ability, and the industrial system abnormal detection model is trained and tested to finally generate a trained industrial system abnormity discrimination model. Using the model, received machine sensor sequence data can be analyzed and whether a machine is abnormal is discriminated.

Abstract: A method for manufacturing a substrate is disclosed. The method comprises the following steps: step one, depositing an amorphous silicon layer on a base material; step two, depositing a silicon dioxide layer with a first thickness on the amorphous silicon layer; and step three, etching the silicon dioxide layer until a thickness thereof is reduced to a second thickness. According to the method of the present disclosure, the silicon dioxide layer with a needed thickness can be manufactured on the amorphous silicon layer. When the ELA procedure is performed, the silicon dioxide layer has an enough thickness to prevent the formation of protrusions at grain boundary of polysilicon, so that the semi-conductive layer manufactured therein can have a relatively low roughness.

Abstract: Provided is a cleaning device, comprising a plurality of cleaning units, and a conveyor which is used for connecting the plurality of cleaning units. Each of the plurality of cleaning units includes a body having a working chamber, and a liquid spray unit. At least one of the plurality of cleaning units comprises a plurality of baffles arranged between the liquid spray unit and the conveyor, and a driving mechanism. The driving mechanism is configured to actuate each of the plurality of baffles into rotation along a first direction to overlap with adjacent baffles, and further actuate each of the plurality of baffles into rotation along an opposite direction of the first direction.

Abstract: The present disclosure relates to an array substrate and the manufacturing method thereof. The manufacturing method includes depositing a conductive layer on a substrate, forming three poles of at least one thin film transistor (TFT), a first signal line, and a second signal line by etching the conductive layer via a first mask, The method also includes depositing an intermediate layer in sequence, forming a first connecting bridge connecting a first portion and a second portion by etching the intermediate layer via a second mask, depositing a conductive electrode, and forming at least one pixel electrode and a connecting line between the first portion and the second portion by etching the conductive electrode via a third mask. In this way, the time of the manufacturing process of the array substrates may be reduced such that the manufacturing cost may be decreased.

Abstract: A method for manufacturing a substrate is disclosed. The method comprises the following steps: step one, depositing an amorphous silicon layer on a base material; step two, depositing a silicon dioxide layer with a first thickness on the amorphous silicon layer; and step three, etching the silicon dioxide layer until a thickness thereof is reduced to a second thickness. According to the method of the present disclosure, the silicon dioxide layer with a needed thickness can be manufactured on the amorphous silicon layer. When the ELA procedure is performed, the silicon dioxide layer has an enough thickness to prevent the formation of protrusions at grain boundary of polysilicon, so that the semi-conductive layer manufactured therein can have a relatively low roughness.

Abstract: The present disclosure relates to an array substrate and the manufacturing method thereof. The manufacturing method includes depositing a conductive layer on a substrate, forming three poles of at least one thin film transistor (TFT), a first signal line, and a second signal line by etching the conductive layer via a first mask, The method also includes depositing an intermediate layer in sequence, forming a first connecting bridge connecting a first portion and a second portion by etching the intermediate layer via a second mask, depositing a conductive electrode, and forming at least one pixel electrode and a connecting line between the first portion and the second portion by etching the conductive electrode via a third mask. In this way, the time of the manufacturing process of the array substrates may be reduced such that the manufacturing cost may be decreased.

Abstract: A diode includes: a semiconductor substrate; a cathode metal layer contacting a bottom of the substrate; a semiconductor drift layer on the substrate; a graded aluminum gallium nitride (AlGaN) semiconductor barrier layer on the drift layer and having a larger bandgap than the drift layer, the barrier layer having a top surface and a bottom surface between the drift layer and the top surface, the barrier layer having an increasing aluminum composition from the bottom surface to the top surface; and an anode metal layer directly contacting the top surface of the barrier layer.

Abstract: The present disclosure relates to a LTPS TFT unit for liquid crystal modules and the manufacturing method thereof. The manufacturing method includes: forming a SiNx layer on a glass substrate; forming a SiOx layer and an a-Si layer on the SiNx layer in sequence; scanning the a-Si layer by laser beams to remove hydrogen within the a-Si layer; adopting excimer laser to re-crystallization anneal the a-Si layer to form the polysilicon layer; forming a gate insulation layer on the polysilicon layer; forming a gate on the gate insulation layer; and forming a drain insulation layer on the gate.