670results about How to "Solve the accuracy problem" patented technology

The invention discloses a method for an AI model maternal and infant assistant from question answering to accurate recommendation, belongs to the technical field of artificial intelligence, and comprises the steps of user demand input and analysis, user portrait construction and dynamic updating, commodity recommendation and optimization, feedback optimization and strategy adjustment. According to the method and the system, the change of user demands and interests can be flexibly coped with, the recommendation accuracy and the user satisfaction degree can be improved, accurate understanding and processing of complex demands are ensured, the real-time performance and accuracy of user portraits are ensured, the change of the user demands and interests can be flexibly coped with, and more accurate and personalized recommendation is realized. The recommendation accuracy and the user satisfaction are improved, and the user can understand what a certain commodity is recommended, so that the trust of the user to the recommendation system is enhanced.

The invention discloses a mechanical arm grabbing control method and system based on digital twinning, and relates to the technical field of industrial robot intelligent control, and the mechanical arm grabbing control method comprises the steps that target object data and grabbing scene data are collected, and a digital twinning model of a target object and a grabbing scene is constructed; dynamic changes of the target object and the grabbing scene are monitored in real time, the digital twin model is dynamically corrected, and a high-precision digital twin model is generated; based on a high-precision digital twinborn model, extracting characteristics of a target object, and identifying an optimal grabbing point; generating a grabbing strategy based on the optimal grabbing point and the high-precision digital twin model, and verifying the feasibility of the grabbing strategy; and according to a verification result, dynamically executing an intelligent grabbing strategy. According to the method, the grabbing precision and stability of the mechanical arm are improved by correcting the digital twin model in real time and recognizing the optimal grabbing point.

The invention discloses a ship classification method based on multi-source information feature level fusion, and the method comprises the steps: firstly obtaining the HRRP and SAR original radar data of a ship target, carrying out the preprocessing of the HRRP original radar data, and generating two formats of HRRP samples; secondly, performing feature extraction on the HRRP sample in the first format to form traditional features; secondly, constructing a sequential network model branch, taking an HRRP sample in a second format as input, embedding a traditional feature, and outputting a mixed feature; a two-dimensional convolution model branch is constructed, SAR original radar data passes through a lightweight CNN network, and ship target structure features are extracted. And finally, designing a feature fusion module of an attention weighting mechanism, carrying out weighted fusion on the mixed features and ship target structure features, and outputting a ship classification result. According to the method, the problem that automatic target radar recognition is limited in recognition precision and performance based on single-modal data is solved, and the classification accuracy is improved.

The invention discloses a method for predicting and suppressing harmonic waves generated by photovoltaic grid connection, which relates to the technical field of harmonic wave prediction, and comprises the following steps: detecting a three-phase current signal of a photovoltaic grid connection circuit by using an ip-iq detection method, and decomposing the three-phase current into an active component and a reactive component through Clarke conversion and Park conversion; according to the method, high-precision detection and separation of harmonic waves in the photovoltaic grid-connected system are realized by combining an ip-iq method with Clarke and Park transformation, dynamic modeling and harmonic signal prediction are performed by using the LSTM model, the defect of poor adaptability of a traditional method to nonlinear loads is overcome, and the prediction error is as low as 0.1%. Through combination of LSTM prediction and APF dynamic compensation, the compensation current opposite to the harmonic current is generated in real time, the harmonic suppression efficiency and the power grid stability are significantly improved, the electric energy quality is optimized, and the dynamic change of a complex load is adapted.

The invention provides an intelligent optimization control method and device for a mold cooling system, and relates to the field of intelligent optimization control devices. The intelligent optimization control device based on the mold cooling system comprises a hardware part, a software control system, a model calculation module and an optimization algorithm, the hardware part comprises a temperature sensor, a flow sensor, a controller, a cooling liquid pump, a cooling loop and an actuator, the software control system is used for receiving and processing real-time data from the sensors, and the model calculation module is used for calculating the model calculation module according to the real-time data. The model calculation module is used for analyzing temperature and flow changes in the cooling process. High-precision and low-delay temperature and flow sensors are utilized, and sensor data are corrected in real time through an error correction module, so that the measurement precision and the response speed are greatly improved, and the problem that the precision and the response speed of the sensors are limited is solved.

A mirror surface defect detection system imaging simulation method based on a structured light field comprises the steps that S1, system parameters are calibrated and serve as simulation model input; and S2, deducing a light propagation model of the light field camera, performing light tracing from a sensor, performing parallel calculation on the light field sub-graph array, and solving an incident light vector of the light field camera. And S3, designing a three-dimensional surface shape of a function fitting deformation defect, solving an incident light vector at the defect, and tracing light to obtain a light source point coordinate and illumination intensity. And S4, based on a Lambert diffuse reflection model, introducing a coefficient representing scratch roughness / stain opacity, and establishing a diffuse reflection non-deformation defect gray model. And S5, generating a simulation image and a label thereof. And comparing the simulation image with the real shot image, and evaluating the reliability of the simulation model. And evaluating the operation efficiency of the simulation model. According to the method, the precision and speed of mirror surface defect light field detection imaging simulation are effectively improved, and the method can be applied to parameter optimization, simulation image data set generation and the like before construction of a detection system.

The invention discloses a social network user recommendation method and system. Relates to the field of social network systems. According to the social network user recommendation method and system, the social network user recommendation method and system for analysis and prediction based on user vectors, multiple behavior vectors and interests are provided specially for scenes of a social network platform, and according to collected behavior information data of the user vectors, a deep neural network model is adopted for analysis and learning; the method comprises the following steps: establishing a cascade relationship of various behavior vectors of user vectors by using a graph convolutional network, modeling influence values of different factor vectors on different user preference scores through a decoupling representation technology, screening potential friends through two stages of Pearson similarity calculation and a hierarchical clustering method, and recommending the friends. The method aims to accurately recognize and recommend potential social relations which a user vector may be interested in through data modeling, so that the user vector experience and the activeness of a social network platform are improved.

The invention discloses a metasurface reverse design method of a generative adversarial network based on anchor point network control. The method comprises the following steps: collecting metasurface design parameters and spectral response data, preprocessing, and encoding the data into a format which can be processed by a deep learning model; and analyzing the training data by using a K-means clustering algorithm to generate a structured control vector. And constructing an AcGAN framework, and integrating the spectrum similarity measurement SOC into a loss function of a generator and a discriminator to enhance the spectrum matching precision. Through an adversarial training method, a generator is trained to generate a design conforming to target spectral characteristics, and meanwhile, a discriminator and an anchor network AncorNet are optimized to quickly predict and evaluate spectral response. Candidate designs are generated using the trained generator in combination with control vectors and random noise. And screening the candidate design to select the optimal design. According to the method, the spectrum regulation and control precision and the design efficiency of the metasurface design are effectively improved, and the challenge faced by a traditional method when a complex spectrum requirement is processed is solved.

The invention provides a large eddy simulation method and system based on a mechanism-only model. The method comprises the following steps: acquiring a current dictionary database of a target turbulent flow field; the current dictionary database comprises multi-dimensional flow field features in a current time step; performing feature screening on the current dictionary database to obtain key flow field features of the target turbulent flow field; inputting the key flow field features into a mechanism-only model, and reconstructing to obtain the sublattice turbulence viscosity of the target turbulence field in the current time step; wherein the mechanism-only model is obtained by training an initial neural network model based on direct numerical simulation data of a sample turbulence field and a sublattice turbulence viscosity label of the sample turbulence field. According to the method, the simulation precision of small-scale turbulence vortex motion in large vortex simulation is improved, so that the problems of poor performance of a bottom layer model and low simulation calculation precision caused by calculation grid scale limitation and sub-grid stress simplification hypothesis are effectively solved.

The invention discloses a large indoor direct shear tester for geotechnical engineering, and relates to the technical field of geotechnical shear tests.The large indoor direct shear tester for geotechnical engineering comprises a sample container system, a shear loading system, a data acquisition control system and a base protection system.The sample container system is formed by splicing a plurality of modular sample containers; the modularized sample container is connected through a detachable connecting fastener, a replaceable lining plate is arranged in the container and is used for adjusting the size and interface conditions of the sample container, and an embedded porous pressure plate is arranged at the bottom of the container and is used for measuring the pore pressure change in the sample seepage process. By installing the base, the protection system and the normal loading system, the function of uniformly applying normal force in the shearing process is achieved, the problem that the test result is affected by eccentric errors in the shearing process is solved, and the precision and stability of the shearing test are improved.

The invention relates to the field of electric power automation, and discloses an intelligent switch cabinet control system and method, and the system comprises sensing equipment which is used for collecting the voltage, current and temperature multi-dimensional operation parameters of a power grid; the edge calculation module is in communication connection with the sensing equipment and is configured to generate a switch control strategy through dynamic modeling and an optimization algorithm based on the collected operation parameters; a control execution module which is in communication connection with the edge calculation module and is used for receiving the switch control strategy and adjusting the operation state of the switch cabinet; and the cloud platform is in communication connection with the edge computing module and is used for storing operation parameters, fault response records and an optimization control model and providing global data analysis and optimization services. By adopting the technical scheme of combining the sensing equipment and the edge calculation module, real-time acquisition and dynamic modeling of multi-dimensional parameters of the power grid are realized, and the technical effect of quickly capturing the running state change of the power grid is achieved.

The invention relates to the technical field of rock mass fracture recognition, in particular to a rock mass fracture recognition method based on a boundary detection neural network model. According to the method, an innovative deep learning framework is provided for boundary recognition, the characteristics of a refinement strategy and sub-pixel convolution are combined, and a leading result is obtained on a data set. The research shows that the CED model can improve the performance of middle-level and high-level visual tasks including optical flow estimation, target suggestion generation, semantic segmentation and the like. According to the method, a novel architecture is provided to solve the precision problem of a CED algorithm, a reverse refining path is supplemented for an HED network to improve the recognition precision, and the path uses efficient sub-pixel convolution to merge features step by step; the CED provided by the invention can generate an edge image which is well aligned with an image boundary.

The invention discloses a topographic data updating method, a topographic data updating device and electronic equipment. The method comprises the following steps: receiving initial topographic data of an operation scene of the transportation equipment and map boundary data of an operation area of the transportation equipment in the operation scene; based on the initial topographic data and the map boundary data, determining a work attention area; in response to running of the transportation equipment in the operation area, scanning the operation attention area through a sensing system of the transportation equipment to obtain current topographic data of the operation attention area; based on the current topographic data and the initial topographic data, determining a topographic change region in the work attention region; and controlling the transportation equipment to carry out topographic data acquisition on the region with the topographic change, and updating the initial topographic data of the work attention region and the map boundary data of the work region based on the acquired topographic data. According to the invention, the technical problem of poor freshness and precision of map updating of the mine operation scene is solved.

The invention relates to the technical field of knowledge graph construction, in particular to an enterprise knowledge graph construction method, system and device and a storage medium, and the method comprises the steps: collecting enterprise data, and obtaining a test data set which comprises the enterprise data of each field and a corresponding identified entity; training a neural network model by taking the enterprise data of each field as input and the corresponding identified entity as output, and generating a first entity identification model of each field enterprise; fusing the first entity recognition model corresponding to each domain enterprise into a second entity recognition model in a knowledge distillation mode; inputting the enterprise data into a second entity recognition model to recognize entities of the enterprise data and extracting a relationship between the entities; and constructing the enterprise knowledge graph based on the entities and the relationships. According to the method and the device, the problem of insufficient entity recognition precision caused by field difference in the prior art can be effectively solved, and the adaptability of the model to different field data is enhanced, so that the efficiency and the quality of knowledge graph construction are improved.

The invention relates to the technical field of general image data processing or generation, in particular to a soybean germplasm resource high-throughput measurement method based on image recognition, and the method comprises the steps: obtaining a correction image of a target measurement object; based on the corrected image, pod contour image parameters and grain contour image parameters are extracted, and size correction is carried out on the image; calculating a pod length parameter, a grain length parameter, a pod width parameter and a grain width parameter based on the corrected image; obtaining a pod width change curve based on the pod width parameter; calculating the pod number based on the pod width change curve; extracting colors based on the corrected image to obtain pod color data and grain color data; and verifying a pod length parameter, a grain length parameter, a pod width parameter, a grain width parameter, pod color data and grain color data based on the decision coefficient and the root mean square error. The problems that in the prior art, time and labor are consumed, and precision is low are solved.

The invention belongs to the technical field of three-dimensional measurement and weld joint processing, and discloses a weld joint feature recognition and path extraction method based on line laser measurement. According to the full-automatic welding seam shape extraction method, an automatic measurement part before welding seam polishing and other machining steps can be achieved, after a robot measures the position of a welding seam, welding seam shape extraction can be achieved, the position of the center line of the welding seam can be calculated, and therefore a machining path is obtained; and the robot can be directly guided to process the welding seam through the hand-eye matrix. In addition, the welding width and surplus height characteristics before and after welding seam machining can also be calculated and compared. After the welding width and the excess weld metal are calculated, the machining parameters can be further calculated, and the whole method provides technical support for automatic welding seam machining.

The invention belongs to the technical field of inductance parameter extraction, discloses an efficient extraction and layout optimization method and system for parasitic inductance parameters of a power semiconductor module, and realizes efficient interaction of simulation data among software based on powerful simulation capabilities of commercial software ANSYS and MATLAB and a joint simulation interface. And convenience is provided for extracting module parasitic inductance parameters and current sharing characteristics. An effective algorithm for analyzing parasitic inductance parameters and current sharing characteristics is applied, scientific guidance is provided for optimizing module layout, and the method has the following remarkable advantages that compared with a method of'artificial experience + hardware debugging ', an efficient simulation development means is adopted, frequent hardware iteration in manufacturing and verification links can be avoided, and the method has the advantages that development cost is reduced, and efficiency is improved. Therefore, the research and development efficiency of the multi-chip silicon carbide power module is improved, and the advantages of higher quality and flexible development of the multi-chip silicon carbide power module are better supported.

The invention relates to the technical field of eutectic, in particular to a binding head assembly, eutectic equipment and a chip mounting method.The binding head assembly comprises a driving assembly, a suction nozzle assembly and a visual assembly, and the suction nozzle assembly is arranged at the output end of the driving assembly and comprises a suction nozzle base, a suction nozzle, a first transparent part and a second transparent part; a through hole is formed in the suction nozzle seat, and the visual assembly is aligned with the through hole; the suction nozzle is arranged at one end of the through hole, and the first transparent part is arranged at the other end of the through hole and seals a port of the through hole; a penetrating adsorption channel is formed in the suction nozzle and communicated with the through hole, and an air path connecting hole is formed in the inner wall of the through hole; the suction nozzle is further provided with a perspective observation window, the observation window and the adsorption channel are arranged in a spaced mode, projection is conducted in the axial direction of the through hole, and the projection of the observation window is located in the projection of the through hole. The binding head assembly provided by the invention can be used for carrying out high-precision positioning on a plurality of chips in the eutectic mounting process.

The invention relates to an efficient magneto-rheological chemical machining device and method for the inner face of a pipe fitting and relates to the technical field of machining, the efficient magneto-rheological chemical machining device comprises a base, the upper surface of the base is fixedly connected with a fixing rack, and a chuck is rotationally arranged on the right side of the fixing rack and used for clamping and rotating the pipe fitting; the right side of the fixed rack is rotationally connected with a guide lead screw. The composite grinding technology of magnetorheological machining and chemical liquid polishing is introduced, the spraying assembly and the magnetic grinding assembly are installed on the surface of the machining rod at the same time, the magnetic grinding assembly is adjacent to the spraying assembly, and synchronous operation of chemical polishing liquid spraying and magnetic grinding is achieved; the inner wall of the pipe fitting can be ground after the spraying assembly is used for spraying, the polishing effect and efficiency of the inner surface of the pipe fitting are improved, the problems of low precision, microcracks and the like of current magnetic grinding machining of the pipe fitting are solved, and the time and cost of high-precision machining are effectively reduced.

The invention relates to a deep pseudo video detection method based on multi-modal comparative learning, and is suitable for the field of deep counterfeit detection. The method comprises the following steps: inputting a to-be-detected video into a trained video authentic identification model, and outputting a detection result of the to-be-detected video; the video authentic identification model comprises a visual encoder which is taken from a trained audio and video contrast learning model and is used for extracting visual features of a face sequence in a to-be-detected video; the audio encoder is taken from the trained audio and video comparison learning model and is used for extracting audio features of the to-be-detected video; the cross-modal feature fusion module is used for fusing the visual features extracted by the visual encoder and the audio features extracted by the audio encoder to obtain a feature fusion result; the spatio-temporal feature extraction module is used for extracting spatio-temporal features from the feature fusion result; the fusion features not only comprise visual features, but also comprise audio features, so that the features are richer; and the classifier is used for classifying the to-be-detected video as a real video or a forged video based on the spatial-temporal characteristics.

The invention discloses a visibility prediction method and system, and the method employs a dynamic window division and abnormal value processing technology based on data variability, solves a problem that meteorological data abnormal values are difficult to accurately recognize, achieves the precise recognition and processing of the abnormal values, and improves the data quality. By adopting the missing value interpolation method based on the weighted distance, the problem that the missing value of the meteorological data is difficult to fill is solved, the accurate interpolation of the missing value is realized, and the integrity and the reliability of the data are enhanced. According to the method, a dynamic normalization processing method is adopted, the problems of different scales of meteorological factor data and local characteristic differences are solved, adaptive normalization of the meteorological factor data is achieved, important fluctuation information is reserved, and the sensitivity of a model to data details is improved.

The invention provides an automatic production stamping method for automobile warm air heat exchanger accessories with reinforcing ribs, which belongs to the field of automobile processing and comprises the steps of die debugging, processing parameter setting, sensor mounting, processing flow execution, processing position control, processing quality control, finished product inspection and processing parameter optimization. Important machining procedures such as notching, reinforcing rib stamping, bending, cutting-off and material returning are continuously carried out, that is, multiple procedures are completed in one die, high integration of the machining procedures is achieved, all the procedures can be closely connected, transferring and repositioning of semi-finished products among all the procedures are reduced, and the production efficiency is improved. And meanwhile, the die replacement and debugging time is shortened, the production period is shortened, the optical sensor used for monitoring the position deviation of the material belt and the piezoresistive flexible sensor used for monitoring the pressure distribution of the material belt are additionally arranged, accurate positioning of the material belt in the stamping process is achieved, and the stability of the stamping quality is ensured.

The invention provides a magnetic field observation method and system suitable for carrying out high-density deployment, and relates to the technical field of data processing, and the method comprises the steps: obtaining observation planning data, determining the number, interval and position distribution of observation nodes, setting a plurality of observation nodes to designated positions in a target region, magnetic field data in a target area is collected through observation nodes, preliminary screening is performed on the collected data according to background noise generated by equipment characteristics, secondary screening is performed according to spatial distribution characteristics among the observation nodes, and the observation data of different observation nodes are aggregated according to spatial position association rules. Performing cross validation on data consistency among observation nodes of the target area, performing optimization through a precision improvement algorithm to obtain a magnetic field distribution model, identifying a magnetic field abnormal area, performing multi-level detailed analysis on abnormal area data, and generating a magnetic field distribution result and abnormal characteristic data of the target area; the magnetic field observation precision is improved.

The invention relates to the technical field of quality detection, in particular to a meat dish quality detection system which comprises a data acquisition module, a first determination module, a second determination module, a judgment module, a correction module and an alarm sending module. According to the method, the quality control level of the meat ball production line is effectively improved through accurate multi-parameter analysis and a step-by-step judgment mechanism, potential quality problems such as cracks, over-wet or over-dry defects can be accurately recognized by monitoring the moisture content, the sphericity and the deformation quantity of the meat balls in real time, the clamping pressure is dynamically adjusted, and the quality of the meat balls is improved. The system can adapt to different production conditions in real time, the influence of too large or too small clamping force on the shape and taste of meatballs is avoided, a real-time feedback mechanism of the correction module ensures that the production process can be rapidly adjusted during fluctuation, the defective rate in production is effectively reduced, and the production efficiency is improved. The problems of low detection precision and slow response speed caused by dependence on static data are effectively solved.

The invention discloses a multi-mode hydraulic motor low-speed heavy-load friction torque parameter identification method and device, and the method comprises the steps: analyzing the torque balance condition of a hydraulic motor, and building a low-speed heavy-load friction torque model suitable for the hydraulic motor based on the operation condition of the hydraulic motor; designing a group of constant-pressure variable-rotating-speed frequency sweeping tracks and a group of constant-rotating-speed variable-pressure tracks, and carrying out steady-state friction parameter identification; and controlling the hydraulic motor to operate at a variable speed, designing a constant-pressure variable-rotating-speed excitation track meeting a continuous excitation condition, and carrying out dynamic friction parameter identification. According to the method, a low-speed heavy-load friction torque model suitable for the hydraulic motor is established, and a model parameter mixed identification framework is provided based on the model. The framework is composed of a steady-state friction parameter identification link and a dynamic friction parameter identification link, excitation tracks are designed under the links respectively, friction torque parameter identification can be completed, and compared with a traditional method, the method has higher systematicness and engineering performance.

The invention relates to the technical field of valve control, in particular to a self-adaptive valve control method and system suitable for complex working conditions, and the method comprises the steps: installing a sensor group on an installation pipeline of a valve body, and collecting the opening characteristic data of the valve body in real time through the sensor group; and inputting the collected opening characteristic data into a pre-constructed first machine learning model, outputting valve opening data, and generating a valve body opening instruction based on the valve opening data to control the opening of the valve. According to the self-adaptive valve control method suitable for the complex working conditions, the pipeline vibration data and the medium fluctuation data are introduced, the cross influence between the pipeline vibration data and the medium fluctuation data is combined, the valve opening compensation instruction is generated, errors caused by vibration or medium fluctuation can be effectively compensated, it is ensured that the response of the valve is more accurate, and the valve opening compensation accuracy is improved. And secondly, in combination with the self-learning capability of the first machine learning model, the system can continuously optimize the compensation algorithm so as to adapt to wider working condition changes.

The invention provides a water outlet head and a water outlet head mounting method. The water outlet head comprises an assembling seat, a switching assembly, a valve seat, a switch seat and a nut. The assembling base is fixedly arranged at a water outlet of the faucet in advance and provided with a mounting cavity. The switching assembly and the valve seat are sequentially installed in the installation cavity, the switching assembly is provided with a valve rod which is installed in a rotating mode and penetrates through the valve seat, the valve rod is sleeved with the switch seat, the switch seat is in butt joint with the valve seat, a limiting structure for limiting rotation flexibility is arranged between the switch seat and the valve rod, and the nut is locked at the end of the valve rod. The invention provides a water outlet head which solves the problems that when the installation position of a water outlet head is directly arranged on an existing faucet, forming is not easy, precision is poor after forming, and installation errors are large, and therefore a valve seat and a switch seat can be installed in a faucet water outlet in the wanted state or angle subsequently.

The invention provides a humanoid robot single-leg gait transition action design and a control method thereof, and through an algorithm module of a finite-state machine and the single-leg gait transition action design, the problems of center-of-gravity shift and insufficient control precision of a biped robot during gait transition can be effectively solved. By means of the method, the gait control capacity of the robot in the complex environment is remarkably improved, especially in the single-foot supporting and single-leg jumping stage, the robot can be stably switched to the corresponding gait, and the possibility of inclination and unbalance is reduced.

The invention belongs to an automatic control technology, mainly relates to the field of agricultural vehicle automation, and particularly relates to a vehicle front wheel turning angle detection method and application. A method of estimating a tire angle based on a rate of change of a steering wheel angle is provided. According to the method, the change rate of the steering wheel turning angle is used as an input quantity, a mathematical model of a vehicle steering system is established, and the vehicle body absolute turning angle calculated through vehicle kinematics is used as a measurement quantity to calculate the tire turning angle. Accurate and high-stability front wheel steering angle detection can be provided under the condition of high-speed operation of an agricultural vehicle.

The invention discloses a turbine blade surface field rapid prediction method based on a multi-channel self-attention network, and relates to the technical field of data prediction. The technical problems that in the prior art, an existing turbine blade surface field prediction technology is long in consumed time, high in calculation cost and insufficient in precision, and obvious temperature and pressure distribution nonuniformity exists are solved. According to the technical scheme, the turbine blade surface field rapid prediction model building method based on the multi-channel self-attention network comprises the steps that turbine blade surface field data are collected; collecting a preset multi-channel self-attention network architecture; performing pretreatment; performing feature extraction and modeling on the turbine blade surface field according to the processed multi-channel self-attention network architecture; and according to the feature extraction and modeling results, establishing a'mapping relation from boundary conditions to a temperature field and a pressure field on the surface of the turbine blade '. The method is suitable for predicting the temperature field and the pressure field on the surface of the turbine blade.