Abstract: This specification provides a training method of a hybrid graph neural network model. The hybrid graph neural network model includes an encoding function and a decoding function. The method includes the following: using instances corresponding to all targets in training samples and several nearest neighbors of the instances as nodes in a graph, a graph representation vector of each instance is generated by using the encoding function based on graph data of all the instances. t rounds of training are performed on a decoding parameter; and in each round, bs targets are extracted from training samples, a predicted quantity of each target is generated by using the decoding function based on the graph representation vector of the instance corresponding to each target and non-graph data corresponding to each target, and the decoding parameter is optimized based on a loss quantity of the current round that is determined by the predicted quantities and label quantities of the bs targets in the current round.

Abstract: A method for improving estimation of leaf area index in an early growth stage of wheat based on red-edge bands of Sentinel-2 satellite images includes the following steps: acquiring field background spectrum and wheat canopy spectrum information by means of Sentinel-2; respectively acquiring a “straw-soil” spectrum and a “wheat-straw-soil” spectrum, and calculating slopes of red-edge areas of the “straw-soil” spectrum and the “wheat-straw-soil” spectrum; calculating a tangent function SATF corrected by a background adjustment coefficient ?; constructing a wheat LAI estimation model based on a spectral variable SATFNIR-RE2; performing preliminary screening on the LAI estimation model using cross validation; and then testing the screened model with independent data.

Abstract: Provided are a charging control method for battery system, and a charging control apparatus. The charging control method includes: under a condition that a battery with a smallest SOC in N batteries has been connected in parallel to the battery system, performing charging control on the battery system based on a difference between a SOC of a first battery not yet connected in parallel to the battery system and a SOC of a second battery already connected in parallel to the battery system; or under a condition that the battery with the smallest SOC in the N batteries has not been connected in parallel to the battery system, performing charging control on the battery system based on a difference between a voltage of a third battery and a voltage of a fourth battery.

Abstract: A display substrate including a drive-circuit layer and a light-emitting structure layer, a preparation method thereof, and a display device, the light-emitting structure layer includes an anode, a pixel definition layer, an organic light-emitting layer and a cathode, and an auxiliary electrode and an organic light-emitting block, arranged sequentially, the pixel definition layer includes an anode opening exposing the anode and an electrode opening exposing the auxiliary electrode, the organic light-emitting block is separated from the organic light-emitting layer, the auxiliary electrode includes the first, second and third auxiliary electrodes arranged sequentially; the cathode includes a first horizontal lapping part lapping with the first auxiliary electrode and a second sidewall lapping part lapping with the second auxiliary electrode, the thickness of the second sidewall lapping part in the direction parallel to the substrate is greater than that of the first horizontal lapping part in the direction per

Abstract: The embodiment of the present disclosure discloses a method, an electronic device, and a storage medium for a ship route optimization. The method for the ship route optimization considers a dynamic feature of a multi-functional emergency rescue ship and an interference effect caused by an airflow, uses a movement model with kinematics non-holonomic constraints, such as a ship total cost assessment function to simulate the movement of the ship, and based on a sparrow search algorithm, learns how to apply the algorithm to a route plan of the emergency rescue ship. The method can improve the ship route optimization algorithm, improve an accuracy and practical applicability of a calculated plan. The method can effectively and accurately locate obstacles and hidden reefs, and consider effects of an airflow and a non-holonomic constraint effect, so as to make the route planning of the ship more efficient and intelligent.

Abstract: A method for locking repetition rate of an optical frequency comb is provided. A cavity length adjusting actuator is adopted to lock the repetition rate of the optical frequency comb. When a locking state of the cavity length adjusting actuator is in a critical threshold state, an optical delay line of an optical frequency comb optical system is adjusted to adjust the repetition rate of the optical frequency comb, so as to keep locking the repetition rate of the optical frequency comb. A locking device of the repetition rate is also provided, in which an optical delay line control system can judge the locking state of a cavity length adjusting actuator control system in real time, such that the repetition rate can be slowly adjusted within a wide range, and the locking state of the optical frequency comb can be maintained.

Abstract: A tight-integrated navigation method assisted by Elman neural network when GNSS signals are blocked based on the tight-integrated navigation system model of the INS and GNSS, where the dynamic Elman neural network prediction model is used to train the inertial navigation error model and the GNSS compensation model, so as to solve the problem of tight-integrated navigation when the GNSS signals are blocked. When the GNSS signals are blocked, the trained neural network is used to predict the output error of GNSS and compensate the output of inertial navigation, so that the error will not diverge sharply, and the system can continue to work in the integrated navigation mode. The low-cost tight-integrated navigation module is used, and the collected information is preprocessed to form the sample data for training the neural network to train the Elman neural network model.

Abstract: A method for separating biomass components by a ternary system, which relates to a technical field of biomass separation, and includes the following steps of: cooking and separating a biomass raw material by using a cooking liquor consisting of organic acid, small aromatic nucleophilic organic molecule and hydrogen peroxide to obtain solid residue and extracting solution, washing and screening the solid residue to obtain paper pulp, and separating and extracting lignin and/or hemicellulose from the extracting solution. This cooking system could effectively minimize the content of residual lignin and other compounds with chromophore groups in the obtained pulp, directly producing the high-whiteness pulp with excellent performance without additional bleaching process. In addition, the hemicellulose saccharides and high-activity lignin can be also obtained, so that the method has an important significance for realizing high value and industrialization of biomass resource utilization.

Abstract: An electrochemical biosensor includes a working electrode modified with a redox polymer and amine-terminated capture aptamer specific for a particular detection target. The binding sequence of the capture aptamer is also complementary to part of a second ssDNA which is labeled with HRP (horseradish peroxidase). The capture aptamer will form dsDNA with the HRP-labeled ssDNA and bring HRP into electrical contact with the redox polymer and the electrode. Prior to capturing the detection target, addition of H2O2 will lead to the highest reduction current due to the redox polymer-mediated, enzyme-amplified electroreduction of H2O2.

Abstract: An electrochemical biosensor includes a working electrode modified with a redox polymer and amine-terminated capture aptamer specific for a particular detection target. The binding sequence of the capture aptamer is also complementary to part of a second ssDNA which is labeled with HRP (horseradish peroxidase). The capture aptamer will form dsDNA with the HRP-labeled ssDNA and bring HRP into electrical contact with the redox polymer and the electrode. Prior to capturing the detection target, addition of H2O2 will lead to the highest reduction current due to the redox polymer-mediated, enzyme-amplified electroreduction of H2O2.

Abstract: Provided are an optimization method, apparatus, device for a wake-up model and a storage medium, which allow for: acquiring a training set and a verification set; performing an iterative training on the wake-up model according to the training set and the verification set; during the iterative training, periodically updating the training set and the verification set according to the wake-up model and a preset corpus database, and continuing performing the iterative training on the wake-up model according to the updated training set and verification set; and outputting the wake-up model when a preset termination condition is reached. The embodiments of the present disclosure, by periodically updating the training set and the verification set according to the wake-up model and the preset corpus database during an iteration, may improve optimization efficiency and effects of the wake-up model, thereby improving stability and adaptability of the wake-up model and avoiding overfitting.

Abstract: The disclosure relates to a tight-integrated navigation method assisted by Elman neural network when GNSS signals are blocked based on the tight-integrated navigation system model of the INS and GNSS. The dynamic Elman neural network prediction model is used to train the inertial navigation error model and the GNSS compensation model, so as to solve the problem of tight-integrated navigation when the GNSS signals are blocked. When the GNSS signals are blocked, the trained neural network is used to predict the output error of GNSS and compensate the output of inertial navigation, so that the error will not diverge sharply, and the system can continue to work in the integrated navigation mode. The low-cost tight-integrated navigation module is used, and the collected information is preprocessed to form the sample data for training the neural network to train the Elman neural network model.

Abstract: Provided are an optimization method, apparatus, device for a wake-up model and a storage medium, which allow for: acquiring a training set and a verification set; performing an iterative training on the wake-up model according to the training set and the verification set; during the iterative training, periodically updating the training set and the verification set according to the wake-up model and a preset corpus database, and continuing performing the iterative training on the wake-up model according to the updated training set and verification set; and outputting the wake-up model when a preset termination condition is reached. The embodiments of the present disclosure, by periodically updating the training set and the verification set according to the wake-up model and the preset corpus database during an iteration, may improve optimization efficiency and effects of the wake-up model, thereby improving stability and adaptability of the wake-up model and avoiding overfitting.

Abstract: Techniques for push-pull type gestures are described. In embodiments, a push gesture away from a cursor object that is displayed by a display device is recognized. An assisted location object is then displayed to assist a user in repositioning the cursor object based on the push gesture. In implementations, a prompt gesture is recognized effective to cause a prompt object having one or more options to be displayed. The options that are displayed are selectable to initiate a function that is operable using the cursor object. In addition, a motion select gesture that selects one of the options is recognized effective to replace the prompt object and initiate a function corresponding to the selected option. Also, selected text is automatically divided into a plurality of vertical sections that are individually selectable to include or exclude when performing a subsequent function, and then the subsequent function is performed.

Abstract: Document editing synchronization techniques are described. In one or more implementations, one or more inputs are received via a user interface to edit a first portion of a first document, the first portion of the first document corresponding to a first portion of a second document. Responsive to receipt of an input to begin editing of a second portion of the first document, correspondence in the user interface of the second portion of the first document is indicated to a second portion of the second document automatically and without user intervention.

Abstract: The invention refers to a technical field of display device, and specifically discloses a slicing device. The slicing device includes a base, a mold and a cutting device; therein, an avoiding groove for cooperating with the cutting device is provided on the base, and multiple first supporting rods are fixed on the base; the mold is fixed detachably at a setting position on the base, and the mold has an opening cooperating with the cutting device; the cutting device is assembled slidably on the multiple first supporting rods; therein, each of the first supporting rods is sheathed with a first buffer spring. In the above technical solution, by means of the detachable connection between the mold and the base, when the mold is damaged, the mold can be detached directly, and replaced with a new mold, and the base does not need to be replaced, so that reduces the cost of maintaining the slicing device.

Abstract: Techniques for push-pull type gestures are described. In embodiments, a push gesture away from a cursor object that is displayed by a display device is recognized. An assisted location object is then displayed to assist a user in repositioning the cursor object based on the push gesture. In implementations, a prompt gesture is recognized effective to cause a prompt object having one or more options to be displayed. The options that are displayed are selectable to initiate a function that is operable using the cursor object. In addition, a motion select gesture that selects one of the options is recognized effective to replace the prompt object and initiate a function corresponding to the selected option. Also, selected text is automatically divided into a plurality of vertical sections that are individually selectable to include or exclude when performing a subsequent function, and then the subsequent function is performed.

Abstract: Techniques for push-pull type gestures are described. In one or more embodiments, a push gesture away from a cursor object that is displayed by a display device is recognized. An assisted location object is then displayed to assist a user in repositioning the cursor object based on the push gesture. In implementations, a prompt gesture is recognized effective to cause a prompt object having one or more options to be displayed. The options that are displayed are selectable to initiate a function that is operable using the cursor object. In addition, a motion select gesture that selects one of the options is recognized effective to replace the prompt object and initiate a function corresponding to the selected option.

Abstract: Techniques for push-pull type gestures are described. In one or more embodiments, a push gesture away from a cursor object that is displayed by a display device is recognized. An assisted location object is then displayed to assist a user in repositioning the cursor object based on the push gesture. In implementations, a prompt gesture is recognized effective to cause a prompt object having one or more options to be displayed. The options that are displayed are selectable to initiate a function that is operable using the cursor object. In addition, a motion select gesture that selects one of the options is recognized effective to replace the prompt object and initiate a function corresponding to the selected option.

Abstract: Document editing synchronization techniques are described. In one or more implementations, one or more inputs are received via a user interface to edit a first portion of a first document, the first portion of the first document corresponding to a first portion of a second document. Responsive to receipt of an input to begin editing of a second portion of the first document, correspondence in the user interface of the second portion of the first document is indicated to a second portion of the second document automatically and without user intervention.