Abstract: The invention provides a method for constructing a target prediction model in a multicenter small sample scenario and a prediction method. In combination with the idea of transfer learning, a training set of a new node is predicted by directly using knowledge of a trained node, and a prediction error sample is used to reflect a difference between the new node and the trained node. The difference is used as supplementary knowledge. Model knowledge of the new node is quickly acquired, to avoid training of a new node from scratch each time. Finally, parallel integration of incremental subclassifiers is implemented by using a ridge regression method, so that the deployment time and costs are greatly reduced. The generalization of models is ensured through sharing of historical knowledge and a knowledge discarding mechanism, good classification effect can also be achieved for a node with a small sample size.

Abstract: A method and a system for controlling cooling of a vehicle, and a vehicle are provided in the present disclosure. In the present disclosure, a maximum value among cooling demand values of cooling objects is used to control a servo component in a cooling system to operate according to an execution opening determined by a preset corresponding relation, so that operating states of the servo components may be correlated and may be comprehensively controlled, so as to prevent the control system from oscillating and improve control stability, thus solving a problem that periodic oscillation of the operating states of the servo components is easily caused and service life of the servo components is affected in existing control modes of the cooling system of the vehicle.

Abstract: A method of configuring a signal transducer of a foldable UE includes: providing a first panel comprising a first conductor configured to provide a first portion of the signal transducer; providing a second panel comprising a second conductor configured to provide a second portion of the signal transducer; providing a hinge connected to the first panel and the second panel and configured to allow the first panel to rotate relative to the second panel about a pivot axis of the hinge, the hinge comprising a third conductor configured to provide a third portion of the signal transducer, the third conductor electrically coupling the first conductor to the second conductor; providing a fourth conductor that is physically separate from the first conductor, the second conductor, and the third conductor; and selectively causing the fourth conductor to provide a fourth portion of the signal transducer.

Abstract: A power conversion device used in electric vehicles includes a transmission assembly and a power converter. The transmission assembly is detachably connected to the electric vehicle and receives a first power from the battery pack of the electric vehicle. The power converter is electrically connected to the transmission assembly and converts the first power into a second power or a third power. The power converter is configured outside the electric vehicle.

Abstract: A solar energy optimization device, a solar energy generation system and a power conversion system using the same are provided. The solar energy optimization device includes a plurality of conversion circuits and a plurality of control circuits. Each of conversion circuits is individually connected in series with a solar module. The conversion circuits and the solar modules are connected in series to a maximum power point tracking (MPPT) circuit. The MPPT circuit is configured to determine a photovoltaic current according to the solar modules. Each of the conversion circuits is used for converting a photovoltaic voltage of one of the solar modules into an output voltage. Each of the control circuits is used to adjust a conversion parameter of one of the conversion circuits to increase the output voltage thereof, so that an output power of each of the solar modules is optimized based on the photovoltaic current.

Abstract: Provided are a digital twinning method and system for a scene flow based on a dynamic trajectory flow, which belong to the field of traffic control. The method includes: extracting and identifying a target semantic trajectory with a detecting and tracking integrated multi-modal fusion and perception enhancement network; extracting road traffic semantics, so as to obtain a highly parameterized virtual road layout top view; obtaining a road layout traffic semantic grid encoding vector based on the virtual road layout top view; constructing a target coupling relation model; constructing a traffic force constraint model; constructing a long short term memory trajectory prediction network; predicting a motion trajectory of a target with the long short term memory trajectory prediction network, so as to obtain the predicted motion trajectory; and obtaining a digital twin of the scene flow based on trajectory extraction, semantic identification and the predicted motion trajectory.

Abstract: Techniques of providing continual augmented reality navigation include combining first and second 6DoF poses from, respectively, a localization from a map in a vicinity of a location to an image and from predicted 6DoF poses based on IMU data. The first 6DoF pose is based on a visual positioning system that uses map data from a vicinity of a location combined with an image from the world-facing camera of the AR smartglasses that provides 6DoF pose in world coordinates independent of the IMU. The second 6DoF pose is based on compensated rotational velocity and acceleration measurements derived from IMU intrinsic values.

Abstract: The present disclosure relates to clock circuits, computing chips, hash boards and data processing devices. A clock circuit comprises M stages of clock drive circuits that are connected in series, M being an integer that is no less than 2, wherein N inverters connected in series are arranged between an input port and an output port of each stage of the M stages of clock drive circuits, N being an odd number that is no less than 3. The clock circuit may provide clock signals with excellent performance.

Abstract: A power transmission and transformation project acceptance method based on building information modeling (BIM) and point cloud measurement includes: S1: constructing a BIM model based on a design drawing, where the BIM model includes standard parameters of key components of a power transmission and transformation project; S2: obtaining point cloud data, and constructing a point cloud data model based on the point cloud data, where the point cloud data model includes actual parameters of the key components of the power transmission and transformation project; S3: comparing the BIM model with the point cloud data model to obtain construction acceptance data; and S4: visualizing the construction acceptance data and outputting an acceptance report to complete acceptance of the power transmission and transformation project. The BIM model and the point cloud data model are compared to implement rapid and efficient acceptance of the power transmission and transformation project.

Abstract: An air ionization display device includes: a pulse-light-source module configured to generate a plurality of synchronous pulse light beams; and a light field control module, the plurality of pulse light beams being projected to the light field control module and the light field control module being configured to adjust and merge the plurality of pulse light beams and ionize air in a display area to form a holographic real image.

Abstract: A display substrate, including: a base substrate; a plurality of gate lines and a plurality of data lines that intersect to surround a plurality of pixels; wherein each pixel includes a first thin film transistor, a second thin film transistor and a pixel electrode. A first electrode of the first thin film transistor is electrically connected to a pixel conductive layer of the pixel electrode, a second electrode of the first thin film transistor is electrically connected to a first electrode of the second thin film transistor, and a second electrode of the second thin film transistor is electrically connected to the data line. An orthographic projection of a combination of the second electrode of the first thin film transistor and the first electrode of the second thin film transistor on the base substrate at least partially overlaps with an orthographic projection of the pixel conductive layer on the base substrate.

Abstract: The disclosure provides a display panel, a compensation of brightness method and a displaying device. The display panel comprises an array substrate, a luminescent layer, a plurality of optical film layers and a photoelectric sensor, wherein the luminescent layer is located on one side of the array substrate, the array substrate is used for driving the luminescent layer to emit light, the optical film layers are located on a side, away from the array substrate, of the luminescent layer, light emitted by the luminescent layer forms optical waveguide on the array substrate and/or the optical film layers, and the photoelectric sensor is used for acquiring the optical waveguide to perform compensation of brightness on the light emitted by the luminescent layer according to data of the optical waveguide.

Abstract: Abnormally-shaped tobacco granules are provided. The tobacco granules comprise a tobacco granule body and a coating attached to the surface of the tobacco granule body. The coating contains an adhesive. The surface of the abnormally-shaped tobacco granule has one or more grooves and/or ridges, and/or, the abnormally-shaped tobacco granule has through holes therein. The surface grooves of the abnormally-shaped tobacco granules can serve as smoke channels after the tobacco granules are bonded, cured and shaped, so that the air permeability inside a total-granule tobacco product is greatly improved. The smoking resistance of cigarettes is reduced. The flammability of cigarettes is improved and the consumption of tobacco raw materials is reduced. A preparation method of the abnormally-shaped tobacco granules, a tobacco product comprising the abnormally-shaped tobacco granules, and a preparation method thereof are also provided.

Abstract: A low-temperature smoking body and a preparation method thereof are disclosed. The low-temperature smoking body includes tobacco particles. The tobacco particle includes a particle body and a shell wrapped on the particle body. A carrier is distributed in the particle body and/or the shell. The carrier includes at least one of a raw tobacco material, a non-tobacco material, and a porous material. The carrier carries a smoking agent, and the smoking agent includes a tobacco extract and/or an atomizer. In addition, a preparation method of the low-temperature smoking body is further disclosed. The low-temperature smoking body has better smoking stability and uniformity, and the tobacco particles having core-shell structures can effectively solve the problem of liquid leakage.

Abstract: A data transmission cable (100) includes: a signal bundle (110), where the signal bundle (110) includes at least three signal cables, the at least three signal cables are disposed at intervals, pairwise signal cables form a differential pair signal cable, and the differential pair signal cable is used to transmit a differential data signal; a ground cable (120), where the ground cable (120) encircles and covers the signal bundle (110), and the ground cable (120) is used to transmit a ground signal and isolate the signal bundle (110) from a signal bundle (110) of another data transmission cable (100); and a filling medium (130), where the filling medium (130) is disposed in space on an inner side of the ground cable (120) except the signal cable, so that a problem that a MIPI bus has poor transmission quality and a short transmission distance can be resolved.

Abstract: A 3D target detection method based on multimodal fusion and a depth attention mechanism includes the following steps: obtaining and preprocessing original point cloud data and original image data; inputting the preprocessed point cloud data and image data into a 3D target detection network based on multimodal fusion and the depth attention mechanism, where the 3D target detection network based on multimodal fusion and the depth attention mechanism includes a generation phase for 3D bounding box proposals and a refinement phase for 3D bounding boxes, and the network outputs parameters and classification confidence of a target bounding box; training the 3D target detection network based on multimodal fusion and the depth attention mechanism; and processing the collected lidar point cloud data and image data by using the trained detection network, and outputting 3D target information, to realize 3D target detection.

Abstract: The present disclosure discloses a method for reducing the fructan content in food made of rice and flour using microwave heating and belongs to the technical field of food processing. Fermented products made of rice and flour are prepared using microwave heating and fructan in the products can be effectively degraded. Compared with traditional steaming, the method can remarkably reduce the fructan content in the fermented food, effectively alleviate the bowel stress sensitivity to the fermented rice and flour food by patients with irritable bowel syndrome (IBS), meet the daily dietary requirements of special crowds, has the advantages of simple process, green processing and environmental protection, and is suitable for large-scale industrial production.

Abstract: A driving mechanism for moving an optical element is provided, including a fixed part, a movable part, a driving assembly, and a first guiding member connected between the fixed part and the movable part. The optical element is disposed on the movable part, and the driving assembly drives the movable part to move relative to the fixed part. The first guiding member is configured for guiding the movable part to move relative to the fixed part.

Abstract: The invention provides a method for constructing a target prediction model in a multicenter small sample scenario and a prediction method. In combination with the idea of transfer learning, a training set of a new node is predicted by directly using knowledge of a trained node, and a prediction error sample is used to reflect a difference between the new node and the trained node. The difference is used as supplementary knowledge. Model knowledge of the new node is quickly acquired, to avoid training of a new node from scratch each time. Finally, parallel integration of incremental subclassifiers is implemented by using a ridge regression method, so that the deployment time and costs are greatly reduced. The generalization of models is ensured through sharing of historical knowledge and a knowledge discarding mechanism, good classification effect can also be achieved for a node with a small sample size.

Abstract: The present disclosure provides a perovskite nanocrystal fluorescent material comprising a micro-porous template and/or a mesoporous template, a substrate, and a perovskite nanocrystal. A substrate is attached to an inner surface of the micro-porous template and/or the mesoporous template and the perovskite nanocrystal is grown on a surface of the substrate. A melting point of the substrate is higher than a melting point of the perovskite nanocrystal. The perovskite nanocrystal fluorescent material has a high luminescence rate and high luminescence stability. The present disclosure also discloses a method for preparing the perovskite nanocrystal fluorescent material. The preparation method does not require organic solvents, and the preparation process is more green and environmentally friendly. The melt crystallization method adopted has a simple and controllable process, which is suitable for large-scale preparation of the perovskite nanocrystal.