43results about How to "Avoid completeness" patented technology

The invention belongs to a data processing method, and provides a data enhancement method, a data enhancement system, electronic equipment and a computer storage medium aiming at the technical problems that a manual annotation mode is difficult to cope with continuously changing business scenes and the timeliness and the flexibility of corresponding model training cannot be ensured when atlas construction and relation extraction technologies are utilized at present. And marking entities and relationships in the initial sample, performing classification arrangement to obtain an entity dictionary, replacing the entities of the text in the marked sample, and marking the replaced entities to obtain a first augmented sample. Constructing a relationship template and a sentence pattern template, re-expressing the text in the replaced first augmented sample through the sentence pattern template, replacing the entity and relationship of the text in the re-expressed first augmented sample in combination with the entity dictionary to obtain a second augmented sample, and then summarizing and fusing the first augmented sample and the second augmented sample to obtain a second augmented sample; and enhanced data is obtained.

The invention relates to the technical field of mask product production, in particular to an anti-aging and firming mask and a preparation method thereof.The anti-aging and firming mask is prepared from, by mass, 5-10 parts of seaweed extract, 3-7 parts of collagen peptide, 2-6 parts of sodium hyaluronate, 4-9 parts of peptide compound and 1-4 parts of vitamin E ester. The anti-aging compact facial mask is prepared through the preparation device of the anti-aging compact facial mask, a discharging roller and a collecting roller in the collecting and tightening structure are matched with a pressing roller, in the facial mask cutting process, the discharging roller conducts discharging, the collecting roller can rapidly collect cut facial mask belt waste and collect the cut facial mask belt waste in a coiled mode, the recovery efficiency of the facial mask preparation waste is improved, and the recovery efficiency of the facial mask preparation waste is improved. And arrangement of a production environment and resource recycling are facilitated.

The invention discloses an internal space relation speculation method and system based on a rock mass structural surface. The method comprises the steps that based on rock mass structural plane recognition results of a ground plane, a joint plane and a fault plane, correlation of rock mass structural plane surface features and internal space relations of the rock mass structural plane surface features is constructed, and surface constraint conditions of the ground plane, the joint plane and the fault plane are determined; based on key control point setting, thickness constraint and geometric calculation of the triangulated irregular network, realizing thickness-invariant extension of the rock mass structural surface; based on thickness constraint, triangularization curved surface mesh deformation is achieved, and then thickness change extension between different depth layers is achieved; and the triangular irregular net is cut based on the intersecting lines of the intersecting rock mass structural surfaces, and stratum pinching is achieved. According to the method, occurrence information such as the inclination angle of the rock mass internal structure and the internal space relation of the rock mass structural plane are obtained by extending the surface of the structural plane, and then a rock mass internal space relation speculation result serves as an internal constraint to provide data support for subsequent three-dimensional modeling.

The pleating and folding integrated device comprises a base, a rotating disc and a plurality of pleating heads. The base is provided with a first central fixing piece and a second central fixing piece which are used for connecting and fixing the balloon, and the first central fixing piece and the second central fixing piece are coaxially arranged at an interval; the turntable is used for realizing rolling, holding and folding of the balloon; the pleating heads are movably mounted on the rotating disc and are circumferentially and uniformly distributed around the same circle center, so that pleating of the balloon is realized; after balloon pleating is completed, the balloon is folded by directly rotating the pleating head, the folding effect is better, the balloon transferring step is removed, the situation that the balloon is scratched can be avoided, the product pleating and folding qualification rate is increased, the operation effect is prevented from being affected, the balloon pleating and folding process is simplified, and the balloon pleating and folding efficiency is improved. And the qualified rate of balloon pleating and folding is further improved.

The invention provides an industrial internet information model modeling optimization method based on natural language processing, and relates to the technical field of computers. The method comprises the following steps: inputting a preset data source and model prompt information into a preset natural language processing model, and extracting to obtain modeling elements; the model prompt information comprises first-stage prompt information and second-stage prompt information; dividing the modeling elements into a corresponding number of modeling element clusters based on the target cluster number; constructing an information model based on the modeling element cluster; the information model is used for indicating the relationship among the modeling elements; the information model comprises a tree structure layer or a combination of the tree structure layer and a net structure layer; wherein the tree structure layer comprises an object class layer corresponding to the information model, a cluster class layer corresponding to each modeling element cluster and a representative element layer corresponding to each modeling element cluster; the problem that an information model is poor in effect and reliability can be solved. And the effect and reliability of the information model are improved.

The invention belongs to the technical field of equipment detection, and provides an anomaly detection system and method for bid evaluation base electrical equipment, and the method comprises the steps: carrying out the automatic collection and analysis of the operation data of the bid evaluation base equipment in a normal operation state in a data collection period; obtaining performance data of the abnormal data and analyzing the performance data to obtain a correction operation signal; acquiring voltage data of an output power supply connected with the PLC in the data acquisition period, performing combined analysis on the voltage data and the operation data, judging whether the low voltage fluctuation of the power supply affects the acquisition of the operation data or not according to the influence performance value, and generating an influence signal; according to the method, the abnormal data in the operation data is corrected, and the reason of the abnormal acquisition of the operation data is analyzed, so that the subsequent data acquisition and data correction are facilitated, and the accuracy of the acquired data is improved on the premise of ensuring the data integrity.

The invention provides a tissue slice patch preparation method based on a tissue chip. The tissue slice patch preparation method comprises the following steps: mixing raw materials to prepare a receptor wax block, collecting a tissue core and transferring the tissue core to the receptor wax block to form a fused wax block, cooling the fused wax block, storing the fused wax block, and carrying out operations of pre-cooling correction, slice unfolding and slice baking. According to the tissue chip-based tissue slice patch preparation method, the tissue chip can be better protected, and the integrity of the tissue chip is ensured.

This invention discloses a UAV inspection method and system for bridge bearings. The method includes: planning a strip-shaped flight path and constructing a coarse 3D point cloud model of the bridge; then planning a scanning flight path around the bridge's outer envelope to collect fine point clouds of bridge components; firstly, using a small amount of labeled data combined with pseudo-label generation and knowledge distillation to complete lightweight training of the PointNet model; then, strengthening structural association features through a contextual attention mechanism and generating virtual points using an occlusion region feature completion method to complete the occluded region and obtain a complete bearing point set; finally, extracting bearing geometric parameters by weighted noise point removal, iteratively fitting bounding boxes, and smoothing; planning a flight path based on the bearing's location to collect the bearing's surface image; then performing image cleaning and enhancement processing, and finally identifying surface defects using image recognition methods to generate an inspection report. Comprehensive technical support is provided.

This invention belongs to the field of QR code printing control technology, specifically disclosing a variable QR code printing control method. This method includes: importing order parameters containing QR code printing position, content type, and application scenario characteristics; real-time acquisition of the flatness and colorimetric features of the substrate surface; quantifying and generating integrity assurance coefficient and clarity maintenance coefficient based on flatness features and printing position; quantifying and generating color stability coefficient and readability maintenance coefficient based on colorimetric features; comparing and verifying each quality indicator with quality indicator thresholds dynamically generated based on order parameters; if verification passes, printing is executed; otherwise, printing parameters are dynamically adjusted before printing is executed. This invention, through multi-feature fusion and dynamic verification adjustment, significantly improves the printing quality, physical durability, and visual readability of QR codes in complex outdoor environments, and overall enhances the adaptability of the QR code printing process.

The invention relates to the field of ships, in particular to a platform wave compensation analogue simulation method based on UDP (User Datagram Protocol) communication, which comprises the following steps of: constructing a Stewart platform model consistent with an entity platform as an upper platform through simulation software, additionally constructing a platform model as a lower platform, and inputting a wave model for dynamic analogue simulation; calculating and analyzing to obtain compensation data of an upper platform driving electric cylinder; generating a corresponding control signal for driving an electric cylinder on the basis of the compensation data, transmitting the control signal to an entity platform through UDP communication, and receiving actual sea condition wave feedback data of the entity platform; and constructing a new wave model according to actual sea condition wave feedback data, and repeating the above steps. The characteristics of low delay and high transmission efficiency of UDP communication are applied to Stewart platform wave compensation analogue simulation, and a complex and low-efficiency communication process in a traditional analogue simulation process is avoided, so that the real-time response speed of the platform is improved.

The present invention belongs to the field of cryptographic technology, and discloses a method, system, storage medium and terminal for implementing a soft cryptographic module. The method for implementing the soft cryptographic module includes: using a random number generation unit to manage entropy sources and generate random numbers; using a key management unit to generate a key pair, generating a KEK based on a PIN code derivation function KDF, and securely storing CSPs such as working keys and MAC verification of PSPs; using an SSL communication unit to establish an SSL secure channel with a security service platform and obtain multiple working keys from the security service platform; using a cryptographic service unit to provide cryptographic data encryption and decryption, MAC calculation, signature verification and entity authentication services. Aiming at the two core problems faced by the soft cryptographic module, the method and SDK for implementing the soft cryptographic module provided by the present invention particularly involve the selection of random number entropy sources and the generation of random numbers, secure key storage and cryptographic services, etc.

The present application relates to the field of video slicing, and discloses a method, system, device and storage medium for video slicing of an engineering inspection process, the method comprising: passing a video through a multi-target detection model to obtain the position information of the equipment, the tester and the number display screen of the inspection object; transforming the position information and cropping the image with the transformed position information to obtain the first information and identify it to obtain the second information; determining the start and end time of the slice according to the appearance and disappearance of the inspection object; selecting the second information of the corresponding time node according to the start and end time to verify the authenticity of the test process, if it is true, slicing the video to be sliced, and binding the second information of the corresponding time node and the authenticity to the slice. The present application determines the start and end time of the slice by the appearance and disappearance of the inspection object, ensures that the dynamic changes of the inspection object are fully recorded, binds the second information to the slice, realizes the traceability of the inspection process, and avoids the problem of false experiments through authenticity verification.

The utility model discloses a vehicle ID information box automatic inputting and conveying device, the vehicle ID information box automatic inputting and conveying device comprises a rack, a material storage mechanism, a conveying mechanism and an inputting platform, the material storage mechanism is used for storing ID information boxes, and a conveying assembly can convey the ID information boxes to the inputting platform through a first material outlet. According to the technical scheme, the conveying platform is arranged to convey the ID information boxes to the input platform, it is ensured that a computer can accurately and completely input vehicle information into each ID information box, and the accuracy of input results is greatly improved; the driving assembly is arranged to move the ID information boxes to the conveying assembly one by one, so that the plurality of ID information boxes can be prevented from interfering with each other when information is input into a computer; and the ID information boxes are conveyed to the input platform by arranging the storage mechanism and the conveying mechanism, reciprocating motion of workers is not needed, the labor intensity is reduced, the working time is saved, and the working efficiency is effectively improved.

The present invention discloses an abnormal data monitoring method and system based on artificial intelligence, which relates to the field of data monitoring technology. The method generates a standardized feature matrix by real-time monitoring and preprocessing data from different sources, effectively reduces data noise and scale differences between features, and analyzes the deviation of each sample relative to the overall distribution, and combines the critical value of the chi-square distribution table to make abnormal judgments on the samples, thereby ensuring the accuracy of abnormal data detection. Then, artificial intelligence technology is used to enhance the abnormal samples, and by introducing distribution constraints, several groups of supplementary abnormal samples are dynamically generated, while ensuring that the generated samples are close to the real distribution, and then by identifying and screening outliers in the abnormal group, the unreasonable samples introduced by noise or deviation in the enhancement process are further reduced, thereby improving the quality of the abnormal group data.

This invention relates to the field of intelligent battery swapping recommendation technology, and discloses an intelligent battery swapping recommendation method and system based on historical battery data. The method includes: a vehicle central control system identifying the vehicle from a user-triggered battery swapping request signal to obtain parameter information of the current vehicle; querying historical battery matching data corresponding to the parameter information from a cloud database, and calculating the compatibility score between each battery model and the current vehicle; a battery swapping cabinet control system performing multi-objective collaborative recommendation decision-making with the compatibility score and the inventory batteries to obtain the optimal recommended battery identifier; and the battery swapping cabinet opening the corresponding compartment door according to the optimal recommended battery identifier to execute the battery swapping process and generate a battery swapping operation record. This invention improves the accuracy of battery compatibility assessment, can discover performance correlations between different battery models, and thus improves the automation and intelligence level of the entire battery swapping process.

The application discloses a video slicing method and device based on multi-dimensional features. The method comprises the following steps: obtaining a video to be sliced; extracting multi-dimensional features of the video to be sliced according to a preset interval, and adaptively determining dynamic fusion weights of the multi-dimensional features based on an attention mechanism to fuse the multi-dimensional features and obtain comprehensive features of the video to be sliced; determining the similarity of the comprehensive features of adjacent intervals of the video to be sliced according to the preset interval, determining key nodes of the video to be sliced according to the similarity, and completing video slicing of the video to be sliced based on the key nodes to obtain a target sliced video. By extracting multi-dimensional features, determining dynamic fusion weights of the multi-dimensional features for fusion to obtain comprehensive features, the content-independent defects caused by fixed-length slicing are solved, the limitations of single-dimensional feature slicing are overcome, the slicing integrity and accuracy are improved, and manual intervention is not required, thereby improving the operation efficiency.

The application discloses an AI model-based RSS subscription source intelligent screening and automatic article generation method and related device, relates to the technical field of artificial intelligence, and comprises the following steps: acquiring a subscription configuration containing an RSS source basic URL, a push cycle and the like, processing tasks in batches through timed scheduling and monitoring states, and improving service reliability; intelligently identifying RSS sources and reducing user configuration thresholds; preprocessing heterogeneous data into a standard format and guaranteeing data flow efficiency; filtering expired content according to a time window calculated according to the push cycle and relieving information overload; dynamically grouping according to an AI screening model Token upper limit and improving AI call success rate; implementing structured screening by using a preset JSON Schema combined with screening prompt words, improving precision compared with a traditional scheme; secondarily grouping according to an AI generation model Token upper limit after recombining results and guaranteeing generation quality; automatically generating articles combined with generation prompt words, reducing the time consumed by manual work to the minute level; generating a matching title, saving a notification and improving experience. The method realizes the upgrade of RSS from passive aggregation to active intelligent processing and is suitable for multi-field content services.

The utility model relates to the technical field of soil sampling, and particularly discloses a soil sampling device for soil fertilizer informatization, which comprises a soil sampler main body and a handheld handle, a soil sample ejection mechanism is inserted into the top end of the soil sampler main body, a display is fixedly arranged at the top end of the soil sample ejection mechanism, and a steering driving part is fixedly arranged on the outer side of the soil sampler main body. A sampling opening in the bottom end of the sampling barrel after sampling can be plugged through the plugging module, so that a soil sample can be prevented from falling off from the sampling barrel in the process of upwards lifting the soil sampling device for soil and fertilizer informatization; through a humidity sensor, a temperature sensor and a PH value sensor on the soil fertilizer detection assembly, the humidity, the temperature and the PH value of soil fertilizer can be detected in the sampling process, data detected by the humidity sensor, the temperature sensor and the PH value sensor can be digitally displayed through a displayer, and detected values can be visually read conveniently; and the specific information of the soil fertilizer can be known in detail.

The invention provides a pumped storage power station monitoring system control flow standardization test method and system, and the method comprises the steps: generating a standard test case set covering all key control branches based on a pumped storage power station control operation sequence and an equipment state constraint condition, and obtaining a standard test case library; based on the standard test case library, designing a multi-frequency vibration suppression equipment state feedback simulator, simulating an equipment state response signal through the equipment state feedback simulator, and outputting an equipment response data stream; performing step-by-step verification on the control logic of the monitoring system based on the equipment response data flow to obtain a control flow verification result; and based on the control flow verification result, performing result analysis to obtain a test result. According to the invention, the standardized test of the control flow of the pumped storage power station monitoring system is realized, and the comprehensiveness, consistency and repeatability of the test are improved.

The embodiment of the present invention provides an image data processing method, device, electronic device and medium, the method is applied to a scanning module, the scanning module is provided with a medium transmission channel, and a second sensor, an image sensor and a third sensor are sequentially arranged on the medium transmission channel; the method comprises: when the medium triggers the second sensor to generate the first edge information, controlling the image sensor to collect initial image data; when the initial image data meets the preset detection condition, judging whether the initial image data has edge information; if not, retracting the medium to the position when the second sensor is triggered to generate the second edge information, and transferring the medium to the image sensor; if so, when the medium triggers the third sensor to generate the second edge information, generating target image data based on the collected initial image data. The embodiment of the present invention can ensure the quality and integrity of the image data obtained by scanning the medium.

The invention discloses an abnormal data monitoring method and system based on artificial intelligence, and relates to the technical field of data monitoring. According to the method, data from different sources are monitored and preprocessed in real time, a standardized feature matrix is generated, the scale difference between data noise and features is effectively reduced, and the accuracy of abnormal data monitoring is improved. And by analyzing the deviation condition of each sample relative to the overall distribution and combining the critical value of the chi-square distribution table, abnormal judgment is performed on the samples, so that the detection accuracy of abnormal data is ensured. Then, the artificial intelligence technology is used for enhancing the abnormal samples, distribution constraints are introduced, a plurality of groups of supplementary abnormal samples are dynamically generated, meanwhile, it is ensured that the generated samples are close to real distribution, outliers in the abnormal groups are recognized and screened out, unreasonable samples introduced due to noise or deviation in the enhancement process are further reduced, and the enhancement accuracy is improved. And the quality of the abnormal group data is improved.

The embodiment of the invention provides a data processing method and device based on a scoring system, electronic equipment and a medium, and belongs to the technical field of financial science and technology. The method comprises the following steps: obtaining an investment portfolio, and obtaining research result data of a plurality of fund companies according to a preset period; in response to the credit scoring instruction, performing data analysis on index data in the investigation result data through a scoring system to obtain a company score corresponding to the fund company; screening all company scores to obtain score quantiles; for each fund company, setting a position scale parameter according to the score quantile, and setting a rating grade according to the position scale parameter and the company score; generating a position monitoring condition according to the rating grade and the position scale parameter, and performing data monitoring on the investment portfolio based on the position monitoring condition; and when the portfolio does not meet the position monitoring condition, generating prompt information. According to the embodiment of the invention, the position scale of a company can be tracked, and potential risk of position of the company can be prompted in time.

The invention belongs to the field of machining, and discloses a complex part hole system machining scheme generation method and related equipment, the method can significantly simplify the programming process for a complex shell part hole system structure, and the hole system target features and feature attributes are automatically recognized and matched with a pre-constructed machining rule library and a pre-constructed machining strategy library, so that the machining efficiency is improved. And a complete processing link is formed step by step in a reverse calling mode, so that the problems of long consumed time, easiness in causing fatigue of programmers and possible human errors in a traditional programming mode are solved. By the adoption of the method, the overall efficiency of programming work is greatly improved, the accuracy and stability of the machining scheme are ensured, the quality of the finally-machined part is effectively improved, and powerful support is provided for quality control in the production process.

The utility model discloses a plastic package sealing device, which relates to the technical field of sealing devices and comprises a sealing table, a connecting frame is fixedly connected to the top of the sealing table, and a position calibration mechanism is arranged on one side of the connecting frame. An electric cylinder drives a connecting rod to move downwards, so that L-shaped plates are located on the two sides of a packaging bag, a second electric push rod drives the L-shaped plates to move, the packaging bag is pushed to move through the L-shaped plates, position calibration of the packaging bag is achieved, the packaging bag is pressed and limited through a limiting piece, position deviation of the packaging bag during sealing is reduced, alignment precision during hot pressing is guaranteed, and deviation or incomplete sealing is avoided. The position, above the conveying rollers, of the packaging bag is limited through the guide frame, the fourth gear drives the third gear to rotate, the using direction of the draught fan can be adjusted, and cooling of the sealing position of the packaging bag can be accelerated through the draught fan and the cooling box.

The invention discloses a multi-angle scanning and curved surface unfolding weld defect detection method, system and device, and the method comprises the steps: controlling each group of triangular laser contour sensors to scan a weld region of a target welding object, and obtaining a multi-view weld point cloud; point cloud fusion processing is carried out on the multi-view welding seam point cloud to obtain a three-dimensional surface point cloud of the welding seam area; performing curved surface expansion processing on the three-dimensional surface point cloud to obtain a two-dimensional expansion graph; performing fusion processing on the three-dimensional surface point cloud and the two-dimensional expansion graph according to a preset fusion algorithm to obtain a fused image; and inputting the fused image into a preset defect detection model to obtain a defect detection result. According to the method, the multi-view weld point cloud of the weld area is obtained, the problems that data are incomplete and detection blind areas are prone to occurring are avoided, curved surface expansion is conducted on the three-dimensional surface point cloud to obtain the two-dimensional expansion view, then the three-dimensional surface point cloud and the two-dimensional expansion view are fused, and the geometric feature utilization rate of the two-dimensional expansion view to the three-dimensional surface point cloud can be increased.

A high-precision digital printing heat transfer printing machine for children's garments is characterized by comprising a base support, a workbench is arranged at the upper end of the base support, a refrigerant net is arranged at the lower end of the workbench, a cooling assembly is arranged at the lower end of the refrigerant net, a transfer paper winding assembly is arranged on the left side of the workbench, and a transfer paper unwinding assembly is arranged on the right side of the workbench; a pressing support is arranged in the center of the rear end of the workbench, a pressing air cylinder is arranged in the center of the front end of the pressing support, a transmission rod is arranged at the lower end of the pressing air cylinder, and a pressing plate is arranged at the lower end of the transmission rod. The printing device has the advantages that the printing effect can be guaranteed through secondary scraping, manual operation is reduced, the working efficiency is improved, and operation is easy and convenient.

This invention relates to the field of spinning machine technology, and more particularly to an intelligent control method and system for a fully servo hydraulically driven spinning machine. The method includes the following steps: fixing the workpiece to the spindle chuck and tightening it by the tailstock mechanism; a pressure sensor acquiring real-time load force information of the radial cylinder of the spinning machine, and a displacement sensor acquiring real-time radial displacement information of the piston rod of the radial cylinder; a PLC calculating the radial cylinder position deviation, speed deviation, and pressure deviation in real-time; the PLC integrating the radial cylinder position deviation, speed deviation, and pressure deviation to calculate the desired control strength percentage of the fully servo hydraulic drive mechanism; and the fully servo hydraulic drive mechanism integrating the nonlinear correction, dead zone compensation, and zero-bias compensation functions of the hydraulic pump to determine the final output flow rate or output speed, thereby controlling the spinning process of the spinning machine. The method and system provided by this invention effectively avoid workpiece tearing due to overpressure or incomplete forming due to underpressure, significantly improving surface quality and dimensional accuracy.

The utility model relates to the technical field of lamp production, and particularly discloses an automatic printing structure which comprises a printing table, a containing box, a printing mechanism and a controller, a grabbing mechanism is installed at one end of the containing box, a finished product box is arranged between the containing box and the printing table, and the printing table comprises a base, an operation panel, an operation box and a printing base. The printing mechanism comprises an inkpad taking frame, a lamp holder positioning column and a printing head, an adjusting mechanism is arranged between the side portion of the printing head and the operation box, a dirt removing mechanism used for removing dirt on the surface of the printing head is further arranged on the top of the base, and an inkpad pressing mechanism used for flattening the surface of inkpad is further arranged on one side of the bottom of the operation box. The problems that the surface of a printing head used for printing of a traditional lamp is prone to being stained with dust, inkpad and other dirt, storage and follow-up use of the printing head are not facilitated, and the printing effect and the quality of a finished product are prone to being affected are solved.

The utility model discloses a kind of rear handle bottle blowing structure, belong to plastic container forming die technical field, the structure aims at solving the problem of traditional handle forming mechanism complex, low reliability, it is integrally arranged cylinder in mould, driving slider is slidably arranged in cylinder, driving slider is connected with movable buckle, the head of movable buckle can be extended into or exit handle cavity of mould.Gas port and second gas port are equipped on cylinder, by alternately passing into high-pressure gas to different gas port, movable buckle is reciprocated between extension position and retraction position.In bottle blowing, movable buckle is extended to prop open handle inner side space;After bottle blowing is completed, movable buckle is retracted to complete handle forming, the structure of the utility model is compact, integrated degree is high, action is reliable and control is accurate, effectively replace traditional complex mechanical linkage mechanism, improve production efficiency and product quality.

This utility model relates to the field of machining technology and discloses an improved expansion rolling wheel, including a wheel body. The rolling surface of the wheel body has a trapezoidal groove in the middle, and the inner wall of the trapezoidal groove is symmetrically provided with isosceles surfaces. The cross-section of the trapezoidal groove is an isosceles trapezoid. This utility model effectively solves the problem of incomplete expansion caused by traditional flat rolling wheels by setting a trapezoidal groove in the middle of the rolling surface of the wheel body. The groove has an upper opening of 20mm, a lower opening of 13mm, and a depth of 1.5mm. It promotes the uniform expansion of metal material to the outer diameter, avoids the outer diameter depression or incomplete forming caused by traditional flat wheels, and significantly improves the outer diameter forming quality. It has the advantages of simple structure, low modification cost, and strong adaptability. It makes metal flow more controllable, eliminates surface wrinkles, cracks and other defects caused by material accumulation or missing material during the expansion process, and improves the surface smoothness of the product.