18results about How to "Efficient culling" patented technology

The invention provides an unmanned aerial vehicle image target detection method and system, and the method comprises the steps: carrying out the filtering of a feature map through a Gaussian filter and a Laplacian filter through the filtering and fine-grained processing of a feature map based on a frequency domain attention mechanism, and highlighting the edge features of a target; processing an input image by using improved ResNet50, introducing spatial convolution, reducing spatial redundant information through reconstruction and separation operations, and generating a feature map with expressive force; filtering low-level information through high-level semantic information by using the features subjected to selective fusion network SFCA fusion processing, and fusing the features; performing classification and regression by using a DETR-based detection network to realize target detection; and model training is completed, the test set is input into the model, and unmanned aerial vehicle image target detection is completed according to a prediction output result. The method solves the technical problems that the target detection efficiency and accuracy are low, the feature pattern quality and expression ability are poor due to insufficient fusion of high-level features and low-level features, and gradient explosion is likely to occur.

The invention relates to a compressor shell burr removing device. The compressor shell burr removing device comprises a feeding assembly, a shell clamping assembly, a burr removing assembly and a discharging assembly which are sequentially arranged in the conveying direction. According to the burr removing device for the compressor shell, automation and continuity of the burr removing process of the compressor shell are achieved by integrating multiple links such as feeding, clamping, burr removing and discharging, and the production efficiency is remarkably improved; a follow-up rod is matched with a feeding mechanism of an inclined guide rail, so that scattered shells can be orderly conveyed to a clamping station, and the problems of low efficiency and unstable posture caused by a traditional manual feeding mode are effectively avoided; meanwhile, the adjustable clamping mechanism is adopted, compressor shells of different specifications and shapes can be effectively adapted, the universality of equipment is improved, the clamp does not need to be frequently replaced, and therefore the production preparation time is shortened, and the production cost and the maintenance cost are reduced.

The present application belongs to the field of refractory material resource utilization, and particularly relates to a comprehensive treatment process for resource utilization of used refractory materials, which comprises: pretreatment, wherein the used refractory materials are sorted, surface cleaned and impurities removed to obtain raw materials for recycling; the raw materials are sequentially subjected to primary coarse crushing, secondary medium crushing, tertiary fine crushing and matched closed circuit screening to obtain crushed materials with target particle size; the qualified particle size materials are sent into a hydration and drying integrated equipment, and after hydration in a water vapor environment, vacuum drying is performed; the dried materials are sent into a wheel milling equipment through a quantitative feeder set, and selective crushing of pseudo-particles is performed under an applied pressure of 15-25 MPa, and finally, classification screening is performed to obtain regenerated materials with different particle size specifications. The process has the advantages that: the step-by-step accurate refinement from large waste materials to target particle size is realized, the particle size is accurately controlled, and the aggregate loss rate is effectively reduced. Compared with traditional single-stage or two-stage crushing, the over-crushing phenomenon is effectively controlled, and the aggregate recovery rate is improved.

The invention belongs to the field of oat grit production processes, and provides a large oat grit production process which comprises the steps of raw material selection, raw material pretreatment, material shelling treatment, elutriation treatment, thermal steaming treatment, drying treatment, color sorting and screening, delinting processing, vibration screening, secondary air screening treatment and finished product manufacturing. Production and processing are conducted through modern machines, natural food nutritional ingredients are reserved, meanwhile, the production efficiency is effectively improved, all indexes of the finished oat grits reach related standards, the finished oat grits have the advantage of being higher in quality, assembly line work is formed, time and labor can be saved, and the production efficiency is improved. Meanwhile, through mechanical process, shells, light impurities and the like in the oat grits can be efficiently removed, so that the quality of finished products is improved, the activity of oat enzyme can be reduced, enzyme decomposition products are reduced, peculiar smell is avoided, deterioration and waste of raw materials caused by manual treatment are avoided, the time cost of production is effectively reduced, and the production efficiency is improved. And the production efficiency and quality are improved.

The invention relates to the technical field of concrete crushing, in particular to concrete crushing and recycling equipment. According to the technical scheme, the hammer comprises a bottom plate and a motor, a main plate is arranged on one side of the upper surface of the bottom plate, a main shaft is rotationally arranged at the upper end of the main plate, a hammer head is arranged at one end of the main shaft through a shaft rod, a shaft head is arranged at the other end of the main shaft, the motor is arranged on the main plate through a sliding assembly, and a square head is arranged at the output end of the motor. The square head is inserted into the shaft head, a braking assembly is arranged on the main plate and acts on the main shaft, roll shafts are rotatably arranged on the bottom plate at equal intervals, and a push plate is slidably arranged on the other side of the upper surface of the bottom plate through a push assembly. The rotary hammer head is adopted for hammering and crushing, concrete blocks can be effectively crushed into small blocks, reinforcing steel bars in the small blocks are removed, the device is suitable for crushing reinforced concrete blocks with different sizes and shapes, and concrete and the reinforcing steel bars can be conveniently recycled.

This invention discloses a method for detecting encrypted communication of malicious software based on autonomous learning, characterized by the following steps: Step S1: Preprocessing network traffic data, including traffic reassembly, protocol filtering, and structured log generation, to extract SSL / TLS encrypted sessions; Step S2: Extracting session statistical features, including TCP connection layer features, SSL / TLS protocol layer features, and certificate layer features, forming a multi-dimensional feature vector; Step S3: Constructing a host session heterogeneous graph, which includes two types of heterogeneous nodes: host nodes and session nodes, and two types of edge types: session-related edges and master-slave-related edges; Step S4: Using the hierarchical aggregation algorithm HSGraphSAGE to embed nodes into the heterogeneous graph, and obtaining host node representations containing spatiotemporal features through multi-hop neighbor aggregation; Step S5: Inputting the host node representations into a classifier, and outputting the detection results of malicious communication behavior of the host. This invention achieves accurate detection of encrypted malicious traffic through a five-step collaborative mechanism. The entire process effectively improves the detection rate of APT attacks and significantly reduces the false positive rate while ensuring privacy compliance (without decrypting the original data).

The invention relates to a method for removing burrs of a compressor shell, which realizes automation and continuity of the burr removal process of the compressor shell by integrating a plurality of steps of feeding, clamping, burr removal, discharging and the like, remarkably improves the production efficiency and reduces manual intervention. According to the method, the feeding step that a follow-up rod is matched with an inclined guide rail is adopted, scattered compressor shells can be conveyed to a clamping station in order, and the problems that traditional manual feeding is low in efficiency and unstable in posture are effectively solved; the adjustable clamping mechanism can flexibly adapt to compressor shells with different specifications and shapes, so that the universality of equipment is improved, and the production preparation time and cost caused by frequent replacement of clamps are reduced; according to the burr removing step of the rotatable gauze wire wheel, burrs at the knife edge part of the compressor shell can be efficiently and uniformly removed, the problems of excessive grinding, insufficient grinding or secondary burrs and the like caused by a traditional mode are solved, and the machining precision and the surface quality are remarkably improved.

The invention discloses a power system safety assessment method, device and equipment and a medium, and relates to the technical field of power system management, and the method comprises the steps: constructing an initial explicit constraint set with unit output as a decision variable; screening the initial explicit constraint set to obtain an effective domain boundary set; establishing a probability distribution model of the uncertain power factors, and extracting a first probability feature of each uncertain power factor by using the probability distribution model; transmitting the uncertainty of each power grid node to a corresponding effective domain boundary based on the effective domain boundary set and the first probability feature to obtain a second probability feature of each effective domain boundary; and obtaining a probabilistic feasible region boundary based on each second probability feature, and evaluating the power system according to the probabilistic feasible region boundary. The power system security risk probabilistic assessment with clear physical meaning is realized by transferring the source load uncertainty from the power grid node to the effective boundary and calculating the probability density function.

The application discloses a biological age evaluation method and system based on whole life cycle DNA methylation and application, and belongs to the technical field of bioinformatics. First, the obtained original DNA methylation data is subjected to quality control and standardization processing; then, differential methylation sites significantly related to calendar age are screened as features; next, taking calendar age as a target variable, a regression model is constructed by adopting a LightGBM gradient boosting framework to obtain a methylation clock model; subsequently, the model is used to predict the DNA methylation age of an individual, and an epigenetic age acceleration value is calculated based on the deviation of the DNA methylation age from the calendar age; finally, the age acceleration value is subjected to correlation analysis with health or physiological indexes in different life stages, so as to evaluate the biological aging state of the individual and predict related health risks. The application covers the whole life cycle, is suitable for Chinese population, and can provide an effective tool for clinical disease risk prediction, health management and anti-aging intervention effect evaluation.

The invention relates to the technical field of federated learning, enhanced gradient attack resistance and privacy protection, in particular to a safe and efficient federated learning method for resisting enhanced gradient attack. According to the method, on the premise that privacy of participants is not leaked, combined defense is carried out through constraint of the component upper bound and the L2 norm of the integer update vectors of the participants, and enhanced gradient poisoning attacks of malicious participants can be efficiently eliminated. According to the method, Boolean sharing and Diffie-Hellman key exchange protocols are combined, the privacy of model update vectors of participants is protected, meanwhile, the privacy protection operation of the participants is also lightweight Boolean operation, expensive encryption operation is not needed, and the privacy protection operation does not further increase the communication overhead of the participants. The invention provides a method for using an alternative casual transmission protocol, Boolean sharing owned by a server is converted into corresponding arithmetic sharing, and the server uses the arithmetic sharing to aggregate a federal learning global model.

A method for removing outliers of three-dimensional high-density electrical data, comprising: arranging high-density original data in columns, respectively, coordinates of electrodes A, B, M and N, natural electric potential SP, voltage VP, current IP and resistivity RO; calculating electrode AB distance AB, electrode AB midpoint distance XAB, electrode MN distance MN, electrode MN midpoint distance XMN, normalized electric potential LOGUI and normalized electric potential and distance ratio LOGUIMNAB according to the high-density original data; establishing a mapping relationship between the spatial positions of data points and each measurement value according to the calculated values of electrode AB distance AB, electrode AB midpoint distance XAB, electrode MN distance MN, electrode MN midpoint distance XMN, normalized electric potential LOGUI and normalized electric potential and distance ratio LOGUIMNAB, and drawing scatter plots of different dimensions; removing outlier data in the scatter plots according to the drawn scatter plots of different dimensions, and saving the data after removing outliers. The present application considers the spatial relationship of each data point and the physical relationship of each measurement value, and has the advantages of high efficiency and high precision.

This invention relates to the field of aquaculture equipment technology, and particularly to a dual-circulation breeding tank that prevents fish eggs from clogging and has a self-cleaning function. The tank includes: a tank body, circulation pipes, and a circulation pump; the circulation pipes are connected to the circulation pump and are connected to the upper and lower ends of the tank body; two sets of circulation pipes are installed on the tank body, forming a dual-circulation design from bottom to top and from top to bottom; a filter box is fixedly connected inside the tank body; filter cotton is placed inside the filter box; an egg-guiding screen is installed inside the tank body; a collecting funnel is located below the egg-guiding screen and is used to collect fish eggs; this invention, by setting two sets of circulation pipes and a circulation pump on the tank body, forms a dual-circulation system consisting of a rearing cycle and a hatching cycle, which not only ensures the water quality of the broodstock in the upper rearing area but also provides a uniform upward flow of water to the lower hatching bed, creating a stable high-dissolved oxygen microenvironment for fish egg development and improving the hatching rate.

The utility model discloses a paper die cutting technical field's a kind of defective product paper automatic rejection die cutting device, to solve the problem of inconvenient rejection of defective product generated in the paper die cutting process in prior art. It includes paper feeding mechanism, cutting mechanism and material receiving mechanism, at least one group of sensors is provided between paper feeding mechanism and cutting mechanism, both sides of material receiving mechanism are equipped with mounting bracket, feed shaft is rotatably arranged between two mounting brackets, adhesive tape roll is installed on feed shaft, the outside of adhesive tape roll is adhesive surface, feed shaft can move in the vertical direction of mounting bracket, material receiving shaft is also installed between two mounting brackets, and material receiving roll is installed on material receiving shaft;The utility model is used to detect the position of defective product in paper die cutting work, and automatically rejects corresponding paper according to the position of defective product paper, since adaptive automation equipment, convenient operation, can avoid secondary damage, with excellent application prospect.

The application relates to the technical field of water level prediction, and particularly discloses a basin water level prediction system and method for complex terrains, wherein the method is characterized in that: a feature screening algorithm is used to screen a preliminary model feature set, and effective features which have a significant influence on the transit of a flood peak and the risk of backflow are reserved; the effective features are scientifically divided into a training set and a test set; the training set is used to enable the model to fully learn the mapping relationship between the buffer state of a buffer confluence unit, the flood discharge, the movement of the flood peak and the change of the confluence water level under complex terrains; and the fusion of hydrological data and topographic image data covers key influence factors in the time and space dimensions on the data level; the water flow lagging effect is captured through an upstream water level lagging feature set on the feature level; the discharge position is accurately positioned in combination with the discharge channel and the buffer state of the buffer confluence unit; and the transit path of the flood peak and the peak water level are simulated, so that the accurate simulation of the transit of the flood peak in the complex terrain and the reliable prediction of the backflow risk can be realized.

The invention discloses an automatic arranging and sorting device for oranges and tangerines. The automatic arranging and sorting device comprises a sorting rack, a visual fruit sorting assembly and a fruit diameter sorting assembly. The visual fruit sorting assembly comprises a detection box fixedly installed on the surface of the sorting rack and a feeding roller rotationally installed on the surface of the sorting rack, and an industrial camera is arranged in the detection box and used for conducting visual detection on fruits and judging the quality of the fruits through a visual analysis system. The fruit diameter sorting assembly comprises a flow dividing guide seat, a rotating disc and a motor, the rotating disc is driven by the motor to rotate on the flow dividing guide seat, and fruits are subjected to fruit diameter grading through through holes of the rotating disc and branch guide holes of the branch guide seat. By integrating visual inspection and specification sorting functions, automatic detection, classification and grading of fruits are realized. And a flexible clamping mechanism is adopted, so that mechanical damage to the fruits in the sorting process is avoided. The device is compact in overall structure, small in occupied space and suitable for large-scale production line installation and multi-specification fruit sorting requirements.

The utility model discloses a flexible rejecting device and rejecting detection equipment, the rejecting device comprises a shell, a rotating shaft and a driving part, the rotating shaft and the driving part are arranged in the shell, and an accommodating space is configured in the shell; the rotating shaft is arranged in the width direction of the shell and penetrates through the shell, one end of the rotating shaft extends out of the shell, and a removing part is arranged on the rotating shaft extending out of the shell; the driving piece is arranged in the length direction of the shell, and the driving piece is connected with the rotating shaft through a connecting piece; wherein a control switch assembly is arranged on the driving piece and used for controlling the driving piece to stretch out and retract back respectively; when the driving piece extends out, the removing part is driven to rotate to a first position; when the driving piece retracts, the rejecting part is driven to rotate to a second position, and more efficient and stable rejecting can be ensured by arranging two control switches; by arranging the connecting piece, the linear motion of the motor is converted into the rotary motion of the removing part, so that the removing part forms a removing curve, and the bottle falling phenomenon can be effectively avoided.

This invention relates to a non-contact liquid level detection method and system, belonging to the field of industrial detection technology. The method includes: receiving a ranging command; simultaneously triggering a millimeter-wave radar sensor and an ultrasonic sensor to measure the liquid level in the same container, obtaining raw echo signals respectively; performing signal processing and liquid level calculation on the raw echo signal from the millimeter-wave radar sensor to obtain a first liquid level value from the millimeter-wave radar sensor, and performing signal processing and liquid level calculation on the raw echo signal from the ultrasonic sensor to obtain a second liquid level value from the ultrasonic sensor; and determining the liquid level in the container based on the first and second liquid level values. This method achieves rapid and efficient removal of invalid data, preventing contamination from flowing into subsequent data systems, and solves the problem that a single sensor cannot identify invalid data in complex engineering environments in the liquid industry.

The application belongs to the field of materials, and particularly relates to a crystal structure generation method based on a diffusion model and a self-attention attenuation graph network, aiming to solve the problems of low matching degree of crystal structure generation and target properties and limited prediction accuracy. The method comprises the following steps: obtaining real crystal structure data, and performing feature extraction to construct a crystal structure dataset; inputting the crystal structure dataset into a diffusion model for training under the condition of target material properties; converting the crystal structure dataset into a crystal graph, and inputting the crystal graph into a graph neural network for training; in the message passing process, combining a multi-head self-attention mechanism and a weight attenuation function based on atomic distance to update node features and edge features, so as to obtain a prediction model; inputting target material properties into the trained diffusion model to generate candidate crystal structures; inputting the candidate crystal structures into the trained prediction model to output property prediction values; comparing the property prediction values with the target material properties, and screening to obtain a target crystal structure.