Patents
Literature
Patsnap Eureka AI that helps you search prior art, draft patents, and assess FTO risks, powered by patent and scientific literature data.

1400results about How to "Achieve lightweight" patented technology

Unmanned aerial vehicle cooperative air combat decision-making method based on GRU-MAPPO deep reinforcement learning

The invention discloses an unmanned aerial vehicle cooperative air combat decision-making method based on GRU-MAPPO deep reinforcement learning, and relates to an air combat game maneuvering strategy. The method comprises the following steps: starting from a complex air combat environment, analyzing a motion model, a maneuvering action, an air combat situation and a reward function of an unmanned aerial vehicle, and establishing a multi-unmanned aerial vehicle air combat environment model which comprises an unmanned aerial vehicle particle motion model, an unmanned aerial vehicle maneuvering action model, an air combat situation evaluation model and an unmanned aerial vehicle reward function model; for a completely cooperative multi-agent environment, a partially observable Markov decision chain framework is established, and a working mode of centralized training and distributed execution is adopted; for unmanned aerial vehicle collaborative decision making, a near-end strategy optimization PPO algorithm is expanded to the multi-agent field, and a GRU is used to replace a full-connection neural network in an original algorithm, so that a GRU-MAPPO algorithm suitable for multi-agent deep reinforcement learning is obtained; a simulation experiment shows that the method can improve the collaborative decision-making capability and the air combat winning rate of the unmanned aerial vehicle.
Owner:XIAMEN UNIV

Classroom teaching quality auxiliary evaluation method, equipment and system based on cloud edge collaboration

The invention discloses a classroom teaching quality auxiliary evaluation method, device and system based on cloud edge collaboration, and the method comprises the steps: obtaining video image data, collected by a camera, of students in a classroom in real time, and transmitting the video image data to an edge device; and in the edge device, performing student classroom behavior identification and tracking on the received video image data by using a pre-trained student classroom behavior identification model, and fusing the identification results of adjacent multiple frames with the identification result of the current frame in the tracking process to obtain the final identification result of the current frame. Evaluating the classroom teaching quality according to the student classroom behavior recognition result, wherein the student classroom behavior recognition model is obtained by training based on a multi-label target detection algorithm and by adopting a knowledge distillation structure in advance; and sending a student classroom behavior identification result and an evaluation result to a cloud video server for storage. The method has the advantages of being simple in implementation method, low in cost, high in real-time performance and precision, high in robustness and the like.
Owner:湖南工商大学

Crowd counting method based on adaptive global perception and multi-scale feature fusion

The invention belongs to the technical field of computer vision, and discloses a crowd counting method based on adaptive global perception and multi-scale feature fusion, and a crowd counting model built in the method comprises an adaptive global perception module AGP, a multi-scale Mama aggregation module and a channel perception Mama decoder CMD. The AGP realizes efficient extraction of global information of the image, and improves the reasoning speed; the multi-scale Mama aggregation module effectively fuses the features of different scales, enhances the capturing ability of detail and global features, and improves the accuracy of crowd counting; the CMD enhances the relevance of information between channels, and further improves the counting precision of the model; the composite loss function ensures rapid convergence and high-precision performance of the network in a crowd counting task. According to the method, the model reasoning speed and the counting accuracy are ensured, meanwhile, the calculation resource consumption is remarkably reduced, and the light weight of the model is achieved.
Owner:SHANDONG UNIV OF SCI & TECH

Lightweight and high-precision underwater target detection method

The invention discloses a lightweight and high-precision underwater target detection method, and relates to the technical field of computer vision and deep learning. Comprising the following steps: data acquisition: acquiring an underwater image; the collected underwater image is preprocessed; the preprocessed image is input into an improved YOLO-RDM model, and a backbone network Backbone of the model uses an RFAConv module to carry out feature extraction; a multi-scale space enhanced attention mechanism is introduced in the feature extraction process, and appropriate weights are adaptively given to targets of different scales; multiple layers of feature maps extracted from the Backbone network are further processed and fused through a Neck layer, target positioning and classification are carried out on fused features through a Head layer, and bounding boxes and category information of targets are output; a C2fDual structure optimization model is introduced to carry out lightweight processing; training and optimizing the model; and performing real-time target detection on the trained and optimized model, and processing and outputting a detection result. According to the invention, the efficiency and reliability of underwater target detection are significantly improved.
Owner:INNER MONGOLIA UNIVERSITY

Fabricated building management method and system based on BIM and cloud service

The fabricated building management method based on the BIM and the cloud service is characterized by comprising the following steps: step 1, constructing a heterogeneous data management model based on the cloud service; step 2, realizing multi-party collaborative fabricated building management based on the data management model; step 3, performing intelligent production and distribution optimization of the fabricated components based on the BIM; step 4, cross-platform lightweight model rendering and interaction based on the BIM, wherein the step 4 comprises the following steps: 4.1, constructing a heterogeneous BIM geometrical relationship model; 4.2, carrying out adaptive streaming transmission; 4.3, multiplexing and rendering the heterogeneous grids; 4.4, on the basis of vision field-first progressive rendering, an overall solution of multi-granularity grid simplification, self-adaptive rendering, vision field-first progressive rendering and multi-terminal response type interaction is provided, geometric topology information and semantic attribute information of the BIM model are brought into full play, the data scale is simplified to the maximum extent, and the rendering efficiency is optimized.
Owner:CHONGQING GUIDANCE TECH CO LTD

Humanoid robot lower limb structure and robot

The invention relates to a humanoid robot lower limb structure and a robot. The lower limb structure comprises a hip unit, a thigh unit, a shank unit and a foot unit. The hip unit comprises a hip seat body, a first joint structure, a thigh swing drive, a thigh rotation drive and a thigh side-sway drive; the first joint structure comprises a first movable seat and a second movable seat; the thigh swing drive is mounted on the inner side of the hip seat body, and an output rotating shaft of the thigh swing drive is connected with the first movable seat; the thigh rotation drive is fixedly connected with the second movable seat, and an output rotating shaft of the thigh rotation drive is connected with the upper end of the thigh unit; the thigh side-sway drive is fixedly connected with the first movable seat, and an output rotating shaft of the thigh side-sway drive is connected with the second movable seat; the axial direction of a rotating shaft driven by thigh swing is parallel to the first direction; the axial direction of a rotating shaft of the thigh side-sway drive is perpendicular to the first direction; the axial direction of the rotating shaft of the thigh rotation drive is perpendicular to the first direction and the axial direction of the rotating shaft of the thigh side-sway drive.
Owner:NAT UNIV OF DEFENSE TECH

Self-adaptive pruning method based on priority experience playback

The invention discloses a self-adaptive pruning method based on priority experience playback. The method comprises the steps of training a to-be-pruned image classification model, converting a pruning process of a neural network into a Markov decision process, and designing a corresponding state space, an action space and a reward function; constructing and initializing a value network, a strategy network, a target value network and a target strategy network of a depth deterministic strategy gradient algorithm; pruning the to-be-pruned image classification model; storing the pruning process of the to-be-pruned image classification model into an experience playback pool, and setting a priority for the pruning process; updating the sampling probability of each sample according to the time difference error; updating the value network and the strategy network, updating the target value network and the target strategy network, and carrying out loop training until a set maximum number of iterations is reached; and carrying out fine tuning on the image classification model to be pruned of image classification. According to the method, the pruning proportion of each layer of the network can be adaptively generated, and adaptive compression of the network to-be-pruned image classification model is realized.
Owner:NO 30 INST OF CHINA ELECTRONIC TECH GRP CORP

Eddy current sensor, detection device and electric driving system

The invention provides an eddy current sensor, a detection device and an electric driving system, and the sensor comprises an eddy current rotor, one end of which is connected with a rotating shaft of a to-be-detected motor and rotates synchronously; one end of the printed circuit board assembly is an electromagnetic induction area, the exciting coil and the receiving coil are arranged in the electromagnetic induction area, the receiving coil is distributed on the inner side of the exciting coil, the other end of the eddy current rotor is close to the electromagnetic induction area to form a preset air gap distance, and the eddy current control module is arranged at the other end of the printed circuit board assembly. The eddy current control module is connected with the excitation coil to form a high-frequency oscillation magnetic field and is connected with the receiving coil to receive the induced electromotive force. The design is compact, the size is greatly reduced, light weight is achieved, and the installation space and the manufacturing cost are reduced.
Owner:UNITED AUTOMOTIVE ELECTRONICS SYST

Vehicle battery pack mounting structure and method

The invention provides a battery pack mounting structure for a vehicle, which comprises a first battery mounting unit, and the first battery mounting unit comprises a first square frame, a second square frame and a third square frame which are sequentially arranged at intervals in the direction of a longitudinal beam; the first longitudinal supporting rod and the second longitudinal supporting rod are arranged in parallel; the third longitudinal supporting rod and the fourth longitudinal supporting rod are arranged in parallel; a first guard plate bracket; a first vertical rod, a second vertical rod, a third vertical rod and a fourth vertical rod; a plurality of rubber bushings or elastic connectors; a plurality of mounting brackets; the first connecting pieces are arranged on the first square frame, the second square frame and the third square frame; and the second connecting pieces are arranged on the first square frame, the second square frame and the third square frame. The first battery installation unit and the battery pack can be pre-installed in a factory, the first battery installation unit and the battery pack can be connected to the installation support on the longitudinal beam only through a rubber bushing or an elastic connecting piece on site, the length is adjustable, light weight is achieved, and the invention further provides an installation method of the battery pack for the vehicle.
Owner:JIANGSU TANGCHEN AUTOMOBILE PARTS

Topological optimization method for thermally stable bearing support structure

ActiveCN119203686Ahigh specific stiffnessHigh heat deformation resistanceGeometric CADDesign optimisation/simulationThermal deformationThermomechanical coupling
The invention discloses a topological optimization method for a thermal stable bearing support structure, which relates to the field of structure optimization, and comprises the following steps: defining a design domain and a non-design domain, and completing initialization setting; based on the density filtering function and the projection function, converting the design variable into a physical variable; establishing an interpolation model of the elastic coefficient and the heat conductivity coefficient; constructing and assembling a heat transfer matrix and a stiffness matrix according to the finite element mesh model; executing heat transfer simulation to obtain an equivalent thermal load of the unit; statics simulation is executed, and design response is calculated; calculating the sensitivity of the design response to the design variables; establishing an optimization model, solving the optimization model based on a moving asymptote algorithm, and updating design variables; and post-processing an optimization result to generate a support structure with a smooth surface. Aiming at the thermal-mechanical coupling optimization problem under environmental thermal fluctuation, the bearing structure with high specific stiffness and thermal deformation resistance is obtained, and the problems that a traditional method is low in stiffness and prone to failure under large environmental fluctuation are solved.
Owner:SHANGHAI JIAOTONG UNIV

Method and system for detecting target grabbed by mechanical arm

The invention relates to a method and system for detecting a target grabbed by a mechanical arm, and the method comprises the steps: collecting a target material image, and obtaining an initial data set; preprocessing the initial data set by adopting a data enhancement mode to obtain a target data set; taking the generative residual convolutional neural network model as a reference model to construct an improved generative residual convolutional neural network model as a mechanical arm grabbing target detection network model; calculating each output loss and total loss by using an improved loss calculation method, obtaining a difference between the output loss and the total loss and practical feasible capture information, and obtaining a training weight; and updating the weight of the model by using a loss calculation result to obtain an optimal training weight, and inputting a test set into the trained mechanical arm grabbing target detection network model to obtain a mechanical arm grabbing target detection result. According to the improved mechanical arm grabbing target detection network model, the grabbing target detection precision is improved, and the grabbing target detection speed is increased.
Owner:CHANGZHOU UNIV

Multi-unmanned aerial vehicle cooperative three-dimensional rapid modeling method for highway accident scene

PendingCN121810922AEfficient collaborative collectionAllocation is accurateResource allocation3D-image renderingVoxelPoint cloud
The invention relates to the technical field of multi-unmanned-aerial-vehicle cooperative operation and three-dimensional modeling, in particular to a multi-unmanned-aerial-vehicle cooperative three-dimensional rapid modeling method for a highway accident scene, and the method comprises the steps: generating a three-dimensional grid map of an accident area through the scanning of a millimeter-wave radar by a main control unmanned aerial vehicle; subareas are divided according to a load balancing strategy and are distributed to slave unmanned aerial vehicles, the slave unmanned aerial vehicles traverse grids along a snake-shaped track, laser radar point clouds and five-view-angle images are synchronously collected, data are bound through double time stamps and space coordinates, the point clouds are preprocessed through edge computing nodes, and the point clouds are stored in a database; the master control unmanned aerial vehicle evaluates quality based on density standard deviation and overlapping matching degree and instructs to reacquire, performs high-precision Poisson reconstruction on an accident core area, performs voxelization processing on a peripheral area, maps image textures, optimizes vehicle deformation details, simplifies a model and retains key element precision, and finally performs data processing. License plate coordinates, a scattered object thermodynamic diagram and an emergency lane occupation state are automatically marked, a visual model is generated, and rapid and accurate restoration of an accident scene is realized.
Owner:NINGXIA COMM TECH DEV CO LTD

Automatic driving vehicle target detection system and method based on cloud edge collaboration

The invention discloses an automatic driving vehicle target detection system and method based on cloud edge collaboration, and relates to the technical field of computer vision and artificial intelligence. According to the target detection model YOLO-Vehile provided by the invention, effective integration of text semantic information and visual features is realized through a CLIP-ViT text feature extractor and a YOLOV8 image feature extractor, and the understanding ability of the model for complex traffic scenes is remarkably improved. According to the target detection model YOLO-Vehile, high detection precision is maintained, meanwhile, light weight of the model is achieved, possibility is provided for deployment of the model on edge equipment with limited resources, and the problem that an existing large model is difficult to deploy rapidly in an actual scene is effectively solved. Through cooperative work of the edge server and the cloud server, flexible scheduling of computing resources and dynamic allocation of tasks are realized. The problem that an existing single calculation mode is unbalanced in calculation resource distribution in the face of complex and changeable traffic scenes is effectively solved, and the overall performance and reliability of the system are remarkably improved.
Owner:SHENYANG AEROSPACE UNIVERSITY

Hollow fresh soybean detection method and sorting device based on machine vision

The invention discloses a machine vision-based hollow fresh soybean detection method and a sorting device. The method comprises the following steps: acquiring a fresh soybean RGB image with transmission information and surface information; extracting a single fresh soybean image through an edge detection algorithm, and constructing a fresh soybean data set with center coordinates; marking the fresh soybean data set, and dividing the fresh soybean data set into a training set, a verification set and a test set; constructing an improved YOLOv8 classification model by using the labeled fresh soybean data set; the improved YOLOv8 classification model is used for conducting hollow fresh soybean judgment on the fresh soybeans on the conveying line, and the hollow fresh soybeans are separated out. The sorting device comprises a material conveying mechanism, an illumination mechanism, an image acquisition unit, a central controller, a sorting mechanism and the like, an improved YOLOv8 classification model is embedded in the central controller, light-weight and high-accuracy classification of fresh soybeans is realized, a multi-target detection task is converted into a single-target classification task, characteristics of the fresh soybeans are maximized, and the sorting efficiency of the fresh soybeans is improved. And the classification accuracy is improved.
Owner:CHINA NAT PACKAGING & FOOD MACHINERY +1

Omnidirectional all-drive robot trajectory tracking method based on constraint simplification

The invention provides an omni-directional all-drive robot trajectory tracking method based on constraint simplification, and belongs to the technical field of ground wheel type mobile robot trajectory tracking. According to the method, a robot kinematic model is established on the basis of rigid body dynamics, the transformation relation between each set of wheel speed vectors and robot speed information in a world coordinate system is represented, and multiple sets of hard constraint problems caused by motor physical characteristics are converted into soft constraint or unconstrained problems in a target function. And the program running speed and the solving success rate are improved. In addition, tracking errors caused by mechanical friction and noise interference are compensated for on the control outermost ring based on high-frequency speedometer information, and it is ensured that the robot accurately executes according to a control instruction. Through experimental verification, the solving speed of each frame of the method is within 50ms, and the tracking error evaluation of a reference trajectory is controlled within 5cm.
Owner:DALIAN UNIV OF TECH

Digital twinning efficient virtual-real synchronization method for inter-face coal pillar digging and supporting integrated equipment

The invention belongs to the technical field of coal mine safety monitoring, and relates to a digital twinning efficient virtual-real synchronization method of inter-face coal pillar digging and supporting integrated equipment. Comprising the following steps: step 1, lightening a digital twin three-dimensional model of the equipment; 2, establishing an equipment digital twin motion behavior model; step 3, establishing an equipment trajectory prediction model: predicting the twinborn body posture of the tunneling and supporting equipment based on Self-Attention-LSTM (Long Short Term Memory); step 4, establishing a virtual space equipment driving and correction model, and a digital twin platform dynamic updating method based on pose correction of a quadratic spline interpolation method and state synchronization; and 5, associating the models to establish virtual-real synchronous interaction of the digital twin tunneling and supporting equipment. The method combines compressed twinning three-dimensional model data and a trajectory prediction model to reduce the driving data volume required by a digital twinning system, more efficiently drives the movement of virtual space equipment, can stably work under the conditions that sensing data is difficult to obtain and data communication is limited, and has higher reliability and practicability.
Owner:XIAN UNIV OF SCI & TECH

Actuating mechanism for motorcycle engine and motorcycle engine

The utility model discloses an actuating mechanism used for a motorcycle engine and the motorcycle engine, the actuating mechanism comprises a driving assembly, an output shaft assembly, a transmission assembly located between the driving assembly and the output shaft assembly, a shell and a shell cover, the driving assembly is fixed on the shell cover, the output shaft assembly is fixed on the shell, and the transmission assembly is fixed on the shell cover. The transmission assembly is located in a cavity formed by the shell and the shell cover. The driving assembly comprises a first motor and a second motor, and the output ends of the first motor and the second motor are both connected with the transmission assembly. Compared with an electric control hydraulic mechanism, the actuating mechanism cancels a compression pump or a hydraulic pump in the electric control hydraulic mechanism, so that the whole system is simpler in structure. Meanwhile, the motor is small in structure and convenient to arrange, and the overall weight of the executing mechanism is reduced. In addition, the response speed is far higher than the transfer speed of the liquid medium, and the action accuracy of the executing mechanism is also improved. And the motor is not highly dependent on the environment, so that the adaptability and reliability of the whole system to the environment are enhanced.
Owner:CHONGQING BEIDA LANDAI AUTOMOBILE TRANSMISSION CO LTD

Hybrid power device and aircraft

The invention provides a hybrid power device and an aircraft, and relates to the field of aircraft power systems. The hybrid power device comprises a power supply system and a control system. The power supply system comprises a solid oxide fuel cell module, a storage battery module and a super capacitor module. The solid oxide fuel cell module comprises an integrated plate type metal support solid oxide fuel cell or electric pile and has the advantages that the fuel gas sealing performance is good, the number and the weight of structural parts such as metal connecting bodies, end plates and sealing materials are reduced, the light weight can be realized, and the power-to-weight ratio is favorably improved. The control system comprises an operation control module and an energy management module; the energy management module is connected with the power input end of the aircraft to obtain the power requirement of the aircraft and calculates the output power of the solid oxide fuel cell module, the storage battery module and the super-capacitor module according to the requirement, the weight of a power system is reduced, the flight endurance time is prolonged, and the flight safety is improved. And the output performance and the service life of the hybrid power system are greatly improved.
Owner:GUANGDONG INST OF NEW MATERIALS

Gateway data processing method based on plug-in mechanism, electronic equipment and medium

The invention belongs to the technical field of software, and discloses a gateway data processing method based on a plug-in mechanism, electronic equipment and a storage medium, and the method comprises the steps: obtaining a network request for collecting data; starting a main program according to the network request, and loading a plug-in; configuring an event type according to the plug-in; and performing data processing according to the event type. According to the invention, the gateway is divided into the main program and the extensible plug-in, the main program only pays attention to the configuration of the equipment and the management of the data flow, the plug-in pays attention to the specific network protocol processing and data acquisition, and only related interfaces defined in the main program need to be realized, so that the implementation is convenient. The main program can load the plug-ins with the unlimited number in a hot loading mode, collection and data forwarding of various kinds of data can be met, the data collection efficiency is improved, light weight is achieved, and expansibility is improved.
Owner:山东浪潮智能生产技术有限公司

Lightweight myoelectricity gesture recognition method based on multi-domain feature fusion

According to the method, a VanilaNet network structure in the field of computer vision is used for reference and improved and then is applied to the gesture recognition problem based on electromyographic signal processing, and the VanilaNet network is difficult to be directly applied to electromyographic signal processing in consideration of the difference of image signals and electromyographic signals in data dimension and feature complexity, so that the recognition accuracy is improved. According to the invention, the VanilaNet network structure is improved in terms of data input and structure simplification, so that the VanilaNet network structure can be applied to electromyographic signals. And finally, under the condition that the recognition precision is ensured, model lightweight can be realized. Furthermore, in the data preprocessing stage, the mode decomposition mode and the data multi-domain fusion are improved, the problem of mode aliasing is avoided, the accuracy is further improved, and the robustness is improved; meanwhile, in the real-time control process, a data caching method is further designed to solve the delay phenomenon caused by the intelligent bionic hand device.
Owner:SHENYANG UNIVERSITY OF TECHNOLOGY

Model data processing method and model data processing device

The invention provides a model data processing method and device, and the method comprises the steps: obtaining original model data of a target model; determining repeated vertexes, deleting vertex data corresponding to the repeated vertexes to obtain de-duplicated vertex data, and performing quantization processing on the de-duplicated vertex data to obtain processed vertex data; according to the data type of the map data, compressing the map data to obtain processed map data; according to the material data, model components with the same material attribute are determined and combined, combined model components are obtained, and the material data of the combined model components serve as processed material data; and compressing and uploading the processed vertex data, the processed mapping data and the processed material data as new model data. By reducing redundant data, optimizing a data structure and adopting an efficient compression algorithm, light weight of model data is realized, so that the rendering efficiency and the transmission rate are improved.
Owner:ZHEJIANG SUPCON INFORMATION TECH CO LTD

Split type hidden outer sealing strip

The invention provides a split type hidden outer water cutting which comprises a vehicle door metal plate and an outer water cutting assembly installed in the vehicle door metal plate, and the outer water cutting assembly comprises an outer water cutting fixing support, a sealing component and a clamping accessory. An outer water cutting installation hole is formed in an automobile door outer plate reinforcing plate of the automobile door metal plate, an outer water cutting assembly installation hole is formed in the outer water cutting fixing support, and the outer water cutting fixing support is installed on the automobile door reinforcing plate under the action of the clamping accessory, the outer water cutting installation hole and the outer water cutting assembly installation hole. A groove is formed in a framework of the outer water cutting fixing support, a base of the sealing component is clamped in the groove, a first sealing feature is connected to the base, a first contact medium is arranged on the first sealing feature, and the first contact medium makes contact with automobile door glass. The outer water cutting assembly is of a split structure, the effect of firm connection can be achieved through cooperation of the outer water cutting fixing support, the sealing component and the clamping accessory, meanwhile, the cost is low, and sealing can be achieved through the sealing component.
Owner:PENG LING AUTOMOTIVE RESEARCH & DEVELOPMENT (SHANGHAI) CO LTD

Integrated thin-wall lattice cooling plate and electronic component

The invention discloses an integrated thin-wall dot matrix cooling plate and an electronic component, the integrated thin-wall dot matrix cooling plate comprises a hollow cold plate body, a cooling flow channel, a liquid inlet and a liquid outlet are arranged in the cold plate body, the liquid inlet and the liquid outlet are communicated with the cooling flow channel, the cooling flow channel is filled with a hollow heat exchange body, and the hollow heat exchange body is filled with a liquid outlet. The positions, except for the cooling flow channels, of the cold plate body are arranged to be of the hollowed-out structure, so that the whole cold plate body can achieve light weight, the hollowed-out heat exchange bodies are arranged in the cooling flow channels, the contact positions of cooling liquid and the heat exchange bodies have the larger specific surface area, and therefore the heat exchange effect of the cold plate is better.
Owner:BEIJING INST OF RADIO MEASUREMENT

High-precision lightweight servo tool turret tool wear monitoring method and system

The invention discloses a high-precision lightweight servo tool turret tool wear monitoring method and system, and the method comprises the steps: calculating the time domain, frequency domain and time-frequency domain features of an acceleration signal, obtaining a multi-domain feature matrix after horizontal splicing, and carrying out the normalization of the multi-domain feature matrix; calculating a correlation coefficient between each feature and the tag after normalization; calculating the reconstruction weight of each feature to obtain a reconstruction feature weight set; the reconstruction features in each domain are automatically added and horizontally spliced to obtain a reconstruction multi-domain feature matrix; establishing a physical model, and combining the physical model to construct a PINN model; establishing a physical loss weight adaptive adjustment function; integrating the data driving loss and the physical loss to construct a total loss function; the obtained reconstructed multi-domain feature matrix serves as input of a PINN model, training is conducted in combination with a total loss function, and the trained PINN model is used for predicting tool wear. By means of the method, the requirement for high-precision and light-weight tool abrasion monitoring on an industrial site can be met.
Owner:ZHEJIANG UNIV

Method, device and system for evaluating self-humidifying performance of fuel cell

The invention relates to a method, a device and a system for evaluating the self-humidifying performance of a fuel cell. The method comprises the following steps: controlling a to-be-tested fuel cell to operate; controlling the steam-water separator to obtain gas output by the fuel cell at the gas outlet of the stack anode, and controlling the steam-water separator to perform gas-liquid separation treatment on the gas to obtain liquid water and saturated gas; acquiring an anode inlet parameter and an anode outlet parameter of the fuel cell to be tested; the anode outlet parameters comprise anode outlet humidity corresponding to gas and anode outlet liquid water weight corresponding to liquid water; determining inlet water content of the galvanic pile anode according to the anode inlet parameters; determining outlet liquid water quantity and outlet gaseous water quantity of the galvanic pile anode according to the anode outlet parameters; determining a purified water transfer coefficient by utilizing the water content, the outlet liquid water quantity and the outlet gaseous water quantity; and determining that the fuel cell has the self-humidifying performance under the condition that the purified water transfer coefficient is a negative value. According to the scheme, whether the fuel cell has the self-humidifying performance or not can be accurately judged.
Owner:STATE POWER INVESTMENT CORP HYDROGEN ENERGY CO LTD

Topological configuration of ion rocket engine based on carbon fiber composite material

The invention relates to the technical field of aerospace propulsion, in particular to an ion rocket engine topological structure based on a carbon fiber composite material, which comprises a grid shell, a discharge shell, a gas circuit shell, a cover plate, a discharge cylinder and a neutralizer outer cover, wherein the grid shell, the discharge shell, the gas circuit shell, the cover plate, the discharge cylinder and the neutralizer outer cover are all prepared based on a carbon fiber composite material; the grid shell is fixedly connected with the front end of the discharge shell; the rear end of the discharge shell is fixedly connected with the front end of the gas circuit shell; the cover plate is fixed at the rear end of the gas circuit shell; the discharge cylinder is arranged in the discharge shell; and the neutralizer outer cover is fixed on the side surface of the discharge shell. The outstanding advantages of good mechanical property, excellent conductivity, small expansion coefficient and low density of the carbon fiber composite material are fully utilized, and the technical problems of high-temperature working reliability, performance level satisfaction and stability of the medium-high power ion rocket engine are solved.
Owner:LANZHOU INST OF PHYSICS CHINESE ACADEMY OF SPACE TECH

Superconducting coaxial cable for quantum computer and manufacturing method thereof

The invention relates to the technical field of cable manufacturing, in particular to a superconducting coaxial cable for a quantum computer and a manufacturing method thereof, and the method comprises the steps: coating an inner conductor with an inner insulating layer to form an inner conductor layer, and electroplating a niobium compound coating on the inner wall of a stainless steel tube to form an outer conductor; forming a cable conductive layer; acquiring an advancing speed change curve of a single inner conductor layer; determining a first velocity mutation type and a second velocity mutation type; determining the vertical stirring speed of the electroplating stirring rod at the position of the outer conductor; electroplating the inner wall of the stainless steel pipe; according to the friction degree of the inner conductor layer and the outer conductor, determining an inner conductor motion adjusting mode, including adjusting the moving speed of the inner conductor layer in the outer conductor, or determining the initial rotating speed of the inner conductor layer when the outer conductor starts to pull a wire; and performing wire drawing on the sleeved outer conductor to form the cable. According to the invention, the stability of cable assembly is improved.
Owner:嘉兴翼波电子有限公司

Lightweight mung bean drought stress identification method based on SGL-YOLOv8s

The invention discloses a lightweight mung bean drought stress identification method based on SGL-YOLOv8s. The method comprises the following steps: firstly, acquiring mung bean drought stress fluorescence image data, and labeling each image by using LabelImg software; performing preprocessing and data enhancement processing on the labeled picture, and then segmenting an image data set after data enhancement into a training set, a verification set and a test set; a YOLOv8s model is improved, an original backbone network is replaced with StarNet, C2fGLFF is introduced into a neck module, and common convolution is replaced with a lightweight LAWDS module; inputting the obtained mung bean drought stress picture into the improved YOLOv8s model for training, and obtaining a pre-trained SGL-YOLOv8s model; the method is used for detecting to-be-recognized mung bean drought stress pictures. The SGL-YOLOv8s model provided by the invention can be used for detecting mung bean fluorescence pictures of different drought stress levels, the parameter quantity and the floating-point number of the model are respectively reduced by 61% and 53%, and the lightweight of a model algorithm is realized.
Owner:HUAIYIN INSTITUTE OF TECHNOLOGY

Production method of green sand as-cast high-plasticity nodular cast iron QT800-5

The invention relates to the field of metal material casting, and discloses a production method of green sand as-cast high-plasticity nodular cast iron QT800-5, which comprises the following steps: preparing pig iron, waste steel and foundry returns, and using auxiliary materials such as a silicon carbide pretreating agent, a high-temperature graphitization carburant, a rare earth magnesium silicon iron nodulizing agent, a silicon barium nucleating agent and a silicon strontium nucleating agent to prepare the green sand as-cast high-plasticity nodular cast iron QT800-5. The purity and component stability of molten iron are ensured; in the smelting process, purification and adjustment are carried out through temperature control and segmented addition of auxiliary materials, then spheroidizing is carried out through a pouring method, and the microscopic structure of a casting is optimized through ladle-to-ladle inoculation and stream inoculation; the final casting has excellent mechanical properties including the tensile strength larger than or equal to 800 MPa, the yield strength larger than or equal to 480 MPa, the elongation larger than or equal to 5% and the matrix pearlite content larger than or equal to 80%. The production process disclosed by the invention is simplified, is suitable for large-scale industrial production, and has a wide application prospect in lightweight application requiring high-strength and high-plasticity materials.
Owner:SHANDONG HUXIWANG GRP FOUNDRY CO LTD

Light truck floor structure

The utility model discloses a light truck floor structure which comprises a front floor assembly, a middle floor assembly, a rear floor sleeping berth assembly and a floor framework assembly, and the floor framework assembly comprises a floor front cross beam assembly, a floor left longitudinal beam assembly and a floor right longitudinal beam assembly. The front floor assembly is arranged in front of the floor front cross beam assembly, the floor left longitudinal beam assembly and the floor right longitudinal beam assembly, the two ends of the middle floor assembly are connected with the middle sections of the floor left longitudinal beam assembly and the floor right longitudinal beam assembly, and the rear floor sleeper assembly is fixed to the rear ends of the floor left longitudinal beam assembly and the floor right longitudinal beam assembly. The front floor assembly, the middle floor assembly, the rear floor sleeping berth assembly and the floor framework assembly are welded into an integral structure. A metal sleeping berth can be arranged, so that the driving comfort is improved, and the increasingly diversified requirements of passengers are met; the structural strength and safety are high, and effective transmission of loads is guaranteed; and the light weight is realized on the premise of ensuring the performance.
Owner:SINO TRUK JINAN POWER CO LTD