51results about How to "Effective correction" patented technology

The invention discloses a channel time-frequency joint extrapolation method based on a multi-layer perceptron hybrid architecture. Aiming at the problems of high acquisition cost of actual communication scene channel data, high time consumption of model online training and difficulty in covering the actual communication scene by simulation data, the method comprises the following steps of: making a pre-training model by using a large-scale simulation channel data set to extract channel change characteristics; and carrying out online rapid adaptation according to the trace data of the actual measurement scene. Aiming at the characteristics that the time delay spread of an indoor wireless channel is relatively small and the number of antennas is relatively small, a multi-layer perceptron hybrid architecture is applied to the time-frequency domain extrapolation of a single-carrier broadband channel. The model can quickly adapt to an actual communication scene through fine tuning of a small amount of data. The method has the advantages of being high in adaptation speed, small in data dependence, high in scene migration capacity, high in calculation parallelism degree and the like, and is suitable for edge deployment and rapid adaptation to actual scenes.

The invention discloses a blood oxygen signal denoising method and system based on adaptive filtering, and the method comprises the steps: synchronously collecting an original blood oxygen signal and a three-axis acceleration signal through a sensor module, and obtaining an original blood oxygen signal sequence and a three-axis acceleration signal sequence which are aligned in time; the three-axis acceleration signals are fused into one-dimensional motion artifact reference noise signals used for representing the limb motion state of the user; the method comprises the following steps: dynamically calculating an adaptive step length factor sequence through a nonlinear mapping function based on multilevel energy analysis of a motion artifact reference noise signal; constructing a time-varying step size adaptive filter, and calculating noise components in the original blood oxygen signal by using the motion artifact reference noise signal and an adaptive step size factor; and subtracting the calculation noise signal from the original blood oxygen signal, and reconstructing to obtain a pure blood oxygen signal without motion artifacts. According to the method, accurate calculation and efficient suppression of motion artifacts can be achieved, and the purification quality and processing efficiency of blood oxygen signals are remarkably improved.

The utility model relates to textile equipment technical field especially, more camera coordination's fabric weft yarn deformation correction equipment. Its technical scheme includes support foot, the top fixedly connected with crosspiece of multiple support foot, the outer wall of two crosspieces is provided with correction mechanism, the correction mechanism includes the multiple fixed knot of fixed in the crosspiece outer wall, the top fixedly connected with top support column of multiple fixed knot, multiple top support column is provided with transmission, the inner wall of crosspiece is provided with heating conduction mechanism, the transmission includes the outer wall one side fixedly connected with drive motor of multiple top support column fixed in the both sides. The utility model discloses more camera coordination's fabric weft yarn deformation correction equipment, it can realize the efficient correction to different types and degree weft yarn deformation to the advantage that avoids the performance decline caused by the uneven heating of fabric.

The invention belongs to the field of oil and gas exploration and development, and relates to an oil drilling and workover well mouth centering and aligning device and method.The device comprises an upper flange and a lower flange, the upper flange is connected with a rotary control head, the lower flange is connected with an annular blowout preventer, a sealing device is arranged between the upper flange and the lower flange, and the upper flange and the lower flange are each sleeved with a pressing plate; the pressing plate on the upper flange and the pressing plate on the lower flange are detachably connected through a fastening device, and adjusting bolts used for adjusting the upper flange to move in the horizontal direction are arranged on the pressing plate on the upper flange and the pressing plate on the lower flange. A flange position adjusting device is designed, horizontal multi-direction free adjustment can be simply and rapidly achieved, a wellhead can be rapidly, safely and effectively corrected, and vertical centering alignment of the center of wellhead devices such as a rotary control head of the device and a crown block and a traveling block is achieved.

The application provides a scene understanding-oriented open-vocabulary three-dimensional scene graph generation system and method, and a computer storage medium, and relates to the technical field of three-dimensional scene graph generation.The method comprises the following steps: acquiring an RGB image;reconstructing a three-dimensional Gaussian representation scene based on multi-view RGB images with camera poses;constructing a plurality of view selectors based on the three-dimensional Gaussian representation scene;acquiring a SAM mask based on a SAM model and the RGB image, and performing clustering;independently constructing a local K-neighbor graph for each cluster based on the clustering result and the three-dimensional Gaussian representation scene;performing smoothing processing on the K-neighbor graph;performing Gaussian codebook encoding based on the image subjected to the smoothing processing to obtain a root cluster node and a leaf cluster node;and generating a three-dimensional scene graph based on the root cluster node and the leaf cluster node.The application can alleviate the boundary feature pollution caused by alpha blending in the three-dimensional Gaussian representation, and retain semantic distribution information under different views.

The present application relates to a kind of unmanned aerial vehicle image radiation consistency correction method under varying illumination conditions, comprising: obtaining the orthographic image of target farmland area;Orthographic image of target farmland area is positioned and corrected in space, and the spatial position of each farmland plot or test plot in orthographic image in target farmland area is obtained;According to the spatial position of each farmland plot or test plot in orthographic image, orthographic image is cut, and the radiation information affected by illumination change in cut image is corrected in radiation consistency;Based on the image data after radiation consistency correction, crop growth monitoring, farmland condition evaluation, variety comparison analysis or yield prediction are carried out.The present application can be widely applied to the field of remote sensing image processing and spectral correction.

The utility model discloses a hydrogen fuel cell carbon paper deformation detection device which comprises a universal testing machine, an experiment pressing block, an experiment cushion block, a polar plate carbon paper stacking structure, a video extensometer, a concentrated light source and a microscope, the polar plate carbon paper stacking structure is arranged between the experiment pressing block and the experiment cushion block, and the universal experiment machine is configured to provide continuously changing packaging force and uniformly transmit the packaging force to the polar plate carbon paper stacking structure through the experiment pressing block and the experiment cushion block, namely, the mechanical environment of carbon paper in a galvanic pile is simulated; the concentrated light source is configured to provide brightness, and the video extensometer and the microscope are configured to observe the form change of the carbon paper in the polar plate carbon paper stacking structure after continuous stress. The device simulates the in-pile mechanical environment through the universal testing machine, and directly observes the real form change of the carbon paper in the flow channel through the optical detection technology, thereby being beneficial to improving the cognition on the internal structure of the electric pile and guiding the design of the electric pile.

The invention provides a polishing device special for a deep hole runner of a pump body, and relates to the technical field of polishing equipment, the polishing device comprises a polishing part and a flexible shaft, the polishing part comprises a center disc and a plurality of supporting rods, and supporting rollers and polishing rollers are arranged on the two sides of the ends of the supporting rods correspondingly; when the supporting roller and the polishing roller are located in the pump body flow channel, the supporting roller makes contact with the polished part of the pump body flow channel, the area where the supporting roller is located serves as the polishing standard, the polishing roller makes contact with the unpolished part of the pump body flow channel, and when the polishing roller makes contact with local protrusions, the trend of radial displacement is generated. The first included angle is equal to the second included angle, and the supporting roller enables the grinding roller to keep the swing direction synchronous with the grinding reference so as to forcibly restrain radial displacement of the grinding roller, so that the grinding roller can apply continuous and stable active correction force to local protrusions, and the capacity of the grinding roller for effectively correcting the shape precision of the pump body flow channel is guaranteed; therefore, the grinding rigidity is ensured while the method adapts to complex runners.

The invention relates to the technical field of treatment drug monitoring, in particular to a method for simultaneously determining the concentration of multiple antiepileptic drugs in plasma. Comprising the following steps: S1, preparing a standard working solution; s2, preparing an internal standard solution; s3, mixing the standard working solution with blank plasma, adding an internal standard solution, centrifuging, and detecting to obtain a chromatogram of a standard solution; s4, based on the chromatogram, taking the ratio of the peak area of the anti-epileptic drug to the peak area of the corresponding internal standard substance as a vertical coordinate Y, taking the ratio of the concentration of the anti-epileptic drug to the concentration of the corresponding internal standard substance as a horizontal coordinate X, and fitting to obtain a standard curve equation; s5, preparing a plurality of groups of quality control working solutions, and verifying the standard curve equation obtained in the step S4; and S6, uniformly mixing the blood to be detected with the internal standard solution, centrifuging, detecting the supernate, and calculating the concentration of the antiepileptic drug in the plasma. The method can realize simultaneous, accurate and quantitative detection of multiple antiepileptic drugs in blood.

The application discloses a kind of unsupervised exposure correction methods, method includes: constructing semantic prompt guide's exposure perception module, the exposure level of image is perceived at visual semantic level;Exposure parameter modulation is constructed pseudo-supervision generation module, according to image exposure level adaptive exposure modulation is carried out, and synthesis is used to guide model training pseudo-supervision image;Construct color degradation compensation exposure correction module, by joint execution luminance correction and color compensation, the luminance distortion in abnormal exposure image and color degradation are recovered;Unsupervised exposure correction network is trained using the network optimization mechanism of multiple loss functions;Based on the unsupervised exposure correction network after training, input image is handled, and image luminance level and color fidelity are effectively improved.The application generates pseudo-supervision image according to image exposure level adaptive to guide model training, and compensates the color information of image while luminance correction, to realize the collaborative recovery of luminance and color.

The utility model provides a kind of flat foot hallux valgus correction shoes for correcting hallux valgus and limiting foot sole simultaneously, to solve above technical problem, the utility model provides following technical scheme, including the shoe sole for supporting foot and the upper for limiting foot position, shoe sole includes heel support part, foot sole support part and toe support part, the heel support part is downward concave structure, and the toe support part is provided with the traction piece that protrudes upwards at the position corresponding with the inside of big toe, traction rope for the traction of big toe outward is connected to the upper end of traction piece, the winding device for winding traction rope is equipped on the outside of upper, and the side of foot sole support part and the position corresponding with big toe is provided with the guide boss that protrudes upwards for limiting the inside of foot sole.

The utility model relates to the technical field of cast-in-place pile construction, in particular to a cast-in-place pile latticed column positioner which comprises a latticed column and a positioning table, adjusting supporting feet are arranged on the side edge of the positioning table, a strip-type gradienter and a lifting lug are arranged on the positioning table, and a rotary table is rotationally arranged on the upper side of the positioning table. A plurality of groups of oil cylinders are arranged on the inner side of the round hole of the rotary table and correspond to the latticed column, an anti-skid pad is arranged at one end of each oil cylinder, the anti-skid pads abut against the side wall of the latticed column, a brake is arranged on the upper side of the positioning table, and the brake comprises a guide rod and a magnetic sheet. The positioner comprises the positioning table and the rotary table, adjusting supporting legs are arranged at the bottom of the positioning table, the positioning table is corrected through the strip-type gradienter, the positioning table is prevented from inclining, the latticed column is clamped through the multiple sets of oil cylinders, telescopic control is carried out, and deviation and angle deviation of the latticed column can be rapidly and effectively corrected in cooperation with the rotary table.

The application discloses a student homework correction method based on submission history information, which comprises the following steps: firstly, processing the submission history information, extracting the latest and initial code changes of the student, extracting the submission message and the code changes of the reference program each time, segmenting the code changes extracted from the student and the reference program, and establishing code change segments; then, finding the code change segments that may be matched between the student program and the reference program, calculating the similarity after finding the code change segments, and establishing a similarity set; then, finding the best matching code segment by using the maximum weight algorithm of the bipartite graph; then, performing text similarity calculation on the submission message corresponding to the matched code segment; and finally, integrating the similarity of the code segment and the corresponding message according to the proportion.

The present application provides a kind of microscopic image distortion correction and reconstruction method, system, equipment and storage medium, preset multi-shape calibration slide, including: regular grid slide, concentric circle slide, irregular curve slide and multi-layer depth slide;The image of multi-shape calibration slide under different field of view, different magnification and different focal plane is collected by microscope, and the geometric distortion and blur characteristics of image are quantified by combining edge detection, contour analysis and blur measurement method, and distortion feature vector is generated;Point spread function model is constructed based on distortion feature vector, and point spread function model is optimized using convolutional neural network;The optimized point spread function model is applied to the deconvolution processing and geometric correction of microscopic image, and finally the clear image after reconstruction is obtained.The present application can effectively correct geometric distortion and edge blur, restore the details and clarity of microscopic image.It has wide application prospect in scientific research, clinical diagnosis and industrial detection field.

The utility model belongs to kitchen appliance technical field discloses a kind of stable operation's food processor, including comminuting device and motor, motor includes stator, rotor, the rotating shaft protruding from rotor, the upper end of rotating shaft is transmission cooperation with comminuting device, motor further includes the balance weight being arranged above rotor and the fan being arranged below rotor, the density of fan is less than the density of balance weight, fan includes integrally formed first balancing portion, second balancing portion and fan blade, first balancing portion is close to rotor and with balance weight opposite arrangement, second balancing portion extends downward from the lower end of first balancing portion and projects outward along radial direction, fan blade is arranged in second balancing portion and is arranged along circumferential direction interval. The initial unbalance of rotor is reduced on the basis of reducing cost, reducing assembly error, the balanced operation of rotor is realized, motor vibration is improved, and rotation noise is reduced.

The application discloses a wide-spectrum catadioptric objective. The wide-spectrum catadioptric objective comprises a first mirror group, a second mirror group and a third mirror group arranged in sequence along the optical axis direction; the first mirror group comprises a meniscus positive lens, a meniscus negative lens and a plano-convex lens, and the edge region of the first mirror surface of the meniscus positive lens and the second mirror surface of the meniscus negative lens is provided with a reflecting surface region; the incident light is reflected and refracted in the meniscus positive lens and the meniscus negative lens for multiple times, and then forms an intermediate image after passing through the plano-convex lens; the first mirror group has positive focal power, increases the numerical aperture and corrects aberration; the second mirror group and the third mirror group both have negative focal power; the intermediate image is emitted to an image plane in parallel light after passing through the second mirror group and the third mirror group, and satisfies 0.35<|R min / f|<1.5; f is the combined focal length of the meniscus positive lens and the meniscus negative lens, and R min is the minimum curvature radius of the reflecting surface. The wide-spectrum catadioptric objective has the characteristics of wide spectrum, high resolution, high numerical aperture, large field of view and long working distance simultaneously.

The invention relates to the field of space remote sensing measurement, in particular to a double-line-scan digital camera optical system based on multiplexing of a primary mirror and a secondary mirror, which comprises the primary mirror, the secondary mirror, a first third mirror, a second third mirror, a rear line-of-sight array focal plane and a front line-of-sight array focal plane, for a rear-view array imaging light path, a light beam is reflected by a first reflection area of a primary mirror, a secondary mirror, a second reflection area of the primary mirror and a first third mirror in sequence and then enters a rear-view array focal plane; for a front-sight array imaging light path, a light beam is reflected by the second reflection area of the primary mirror, the secondary mirror, the first reflection area of the primary mirror and the second third mirror in sequence and then enters a front-sight array focal plane. Through multiplexing of the primary mirror and the secondary mirror, two independent three mirrors are adopted to respectively adjust observation angles corresponding to a double-linear-array imaging light path, the system can flexibly adapt to asymmetric observation angles, and is excellent in image quality, compact in system structure and light in weight under the condition of a large relative aperture. And the method has important engineering significance for promoting the miniaturization and lightweight development of the double-linear-array surveying and mapping camera.

The application discloses a kind of pipeline product image correction method, device, equipment and storage medium, belong to the field of automated production technology.The application obtains the real-time image of the loading tray on the pipeline, and the real-time image of the loading tray is split into correction reference image and image to be corrected;The correction reference image is split into multiple correction reference sub-images by equal width;Obtain the line feature parameter of each correction reference sub-image, and obtain the correction parameter sequence according to the line feature parameter;According to the correction parameter sequence, the image to be corrected is corrected, and the corrected product image is obtained.The above-mentioned method can effectively correct the product image by linear camera shooting and the method based on correction reference image, eliminate the distortion caused by the speed of the pipeline, so as to obtain accurate image data, realize the automatic identification, positioning and quality detection of the product on the pipeline, and improve the production efficiency and production observation precision.

The application discloses a method for real-time adjustment of diamond wire cable-stayed by image recognition, comprising: installing an industrial camera on a servo adjustment mechanism fixing piece on the outside of the diamond wire, above the wire roller and below the synchronous guide wheel, and adjusting to obtain a final shooting area range; installing a light source on the back plate of the take-up chamber on the inside of the diamond wire, above the wire roller and below the synchronous guide wheel; starting the slicing machine to cut the silicon rod, real-time monitoring the diamond wire to obtain a real-time diamond wire monitoring picture, and determining a left and right allowable deviation angle; the diamond wire starts to appear cable-stayed, and gradually deviates from the left or right allowable deviation angle; when it is judged that the diamond wire deviates from the left allowable deviation angle, a left deviation pulse signal is continuously output to correct the diamond wire; when it is judged that the diamond wire deviates from the right allowable deviation angle, a right deviation pulse signal is continuously output to correct the diamond wire. The application achieves the purpose of effectively correcting the diamond wire cable-stayed.

The application discloses a three-way catalyst reducer shaping tool composite die set, which comprises a fixed die seat, the upper end face of which is provided with a lower die seat in parallel, two symmetrically arranged side die frames, which are symmetrically fixed on the fixed die seat and located on the two sides of the lower die seat, an upper stamping seat arranged above the fixed die seat, a first forming die set and a second forming die set, each of which is composed of two model plates arranged in an up-down mode, a die closing mechanism arranged below the lower die seat, and a plurality of reducer units, each of which is arranged at two reducer shaping ports of the first forming die set and the second forming die set. The double die set design of the application combines pre-reducing with accurate reducing, step-by-step forming is realized, size out-of-control caused by single large deformation is avoided, and a micro dynamic spinning shaping force is applied to the inside of the shell, so that the rebound is significantly reduced and the size stability after reducing is ensured.

The application discloses a glass aspheric wide-spectrum imaging lens to solve the imaging requirement of a wide waveband. The lens comprises four glass aspheric lenses and a protective glass coaxially arranged in sequence from an object side to an image side, wherein the first lens is a positive lens; the second lens is a negative lens; the third lens is a positive lens; and the fourth lens is a negative lens. The total length of the lens is 5.98 mm, the F# is 3.11, and the lens satisfies the requirement of the axial aberration being less than 50 microns in the confocal imaging in the waveband range of 340 nm-1550 nm. The lens system effectively corrects various aberrations in the waveband range of 340 nm-1550 nm by using the mode of matching the high-Abbe-number positive lens with the low-Abbe-number negative lens.

The application relates to the technical field of computer vision and artificial intelligence, and provides a remote sensing image target detection method based on an improved YOLOv9s, which comprises the following steps: acquiring a remote sensing image and performing a pretreatment operation; introducing a kernel selection feature fusion (KSFF) module into a neck network of a YOLOv9s model, and replacing at least one Concat structure in the original model; adding a cross-space multi-scale attention (CSMA) module in the neck network; replacing at least one SPPELAN structure in the original model with a parallel pooling feature modulation (PPFM) module in the neck network; and training, testing and evaluating the improved YOLOv9s network. Through the comprehensive application of network structure improvement, data pretreatment optimization, a semi-supervised learning strategy and an advanced training method, the improved YOLOv9s model has higher detection precision in a remote sensing image target detection task.

The application discloses a binocular structured light 3D measurement method based on two-step phase shift, and relates to the technical field of computer vision and optical three-dimensional measurement. Firstly, a nonlinear correction method based on four-value defocus stripe optimization and cubic filtering is proposed, four-value defocus stripes are projected, the optimal four-value gray combination is screened by taking the residual sum of squares (SSE) and the coefficient of determination (R²) as quantitative indexes, and the system nonlinear correction is completed by combining the cubic increasing kernel rectangular filtering and the polynomial correction model. Secondly, a two-step phase shift algorithm with multi-step and multi-direction difference is proposed, the cosine component is extracted through stripe preprocessing, the difference operation of nine different steps and directions is completed by using the difference method to obtain the sine component, the sine and cosine component coefficients are calibrated by combining energy matching, and the final phase is output through phase level alignment and statistical fusion. Finally, the binocular stereo vision technology is fused, the phase, modulation degree and phase gradient joint matching features are combined, and the SGM algorithm is used to realize three-dimensional point cloud reconstruction.

The present application relates to a kind of synthesis method and electronic equipment of coronary artery CCTA image, comprising: obtaining the image of paired NCCT and CCTA, after pretreatment, obtain high consistency image training pair;LDM model suitable for adapting coronary artery image synthesis is constructed, the LDM model includes frozen variational autoencoder, the denoising U-Net of trainable based on cross attention mechanism, and decoder, the denoising U-Net integrates ResNet2D module, Transformer2D module and multi-scale cross attention layer;After pre-processing, the NCCT image to be processed is input into the LDM model trained, and the latent representation of CCTA image is generated by the reverse diffusion denoising process in latent space, and finally the high-resolution Syn-CCTA image is output by decoder.Compared with prior art, the present application has the advantages of no contrast agent to realize coronary artery CCTA image with auxiliary diagnostic value.

The application discloses a distributed high-speed railway mileage correction method assisted by account data, and comprises the following processes: reading detection data and account data of high-speed railway track geometry, finding matched detection curves of the account curves corresponding to each account data through positioning and matching, and then performing translation correction; the non-curve segments are corrected through an equal-division interpolation method, and the complete corrected data are obtained through alternate splicing. The application can assist railway technical personnel in performing the mileage correction of track geometry data, and solves the problem of slow mileage correction.

The invention discloses an intelligent logistics vehicle control method and device, and the method comprises the following steps: controlling an intelligent logistics vehicle to execute a task based on pre-shared map data; when the task is executed, information about a map matching rate determined based on the map data is received from the intelligent logistics vehicle; and based on the received information about the map matching rate, sending position initialization information, so that the intelligent logistics vehicle executes relocation.

The invention belongs to the technical field of micro-nano optics, and particularly relates to a large-view-field multi-wavelength achromatic double-piece cascade super-structure lens device and a design method and application thereof. The device comprises two cascaded super-structure lenses which are sequentially arranged from an object plane to an image plane, each cascaded super-structure lens is composed of achromatic super-structure atoms which are periodically arranged, incident light can be cooperatively regulated and controlled, and aberration correction is achieved while focusing is conducted. By optimizing the phase distribution of the two cascaded super-structure lenses and the distance between the two cascaded super-structure lenses and selecting adaptive achromatic super-structure atoms, the device can be adaptive to multiple wavelengths or specific single wavelengths. The design method of the device comprises the following steps: determining device parameters according to requirements, optimizing the phase distribution of the two cascaded super-structure lenses and the distance between the two cascaded super-structure lenses, and finally selecting adaptive achromatic super-structure atoms so as to complete the design. The device is suitable for multi-wavelength or specific single-wavelength, has the effects of large view field and high image quality, and can be applied to the fields of imaging modules, near-to-eye display, laser beam shaping, laser radar and the like.

The present application relates to a kind of unmanned ship combination navigation method based on IMU / DVL / magnetometer / LiDAR, for the hardware condition of MEMS IMU, Doppler velocity meter and magnetometer and other sensors.Point cloud distortion correction refers to the point cloud in the same data frame in different coordinate systems is transferred to the coordinate system of the starting point cloud in the frame.MEMS IMU, DVL, magnetometer all have high data update frequency, before point cloud data update, the motion of carrier is predicted and estimated.MEMS IMU outputs angular velocity and acceleration in carrier coordinate system, using the above information to update state;Magnetometer outputs magnetic field intensity in b system, DVL outputs velocity in the coordinate system of DVL, using the above information to update measurement, deduce simplified 9-dimensional state Kalman filter, estimate the motion of carrier, and then effectively correct point cloud distortion.Realize the point cloud in the same data frame is transferred to the coordinate system of the starting point cloud in the frame, complete the goal of point cloud distortion correction, improve the precision of LiDAR navigation.

The present application belongs to the technical field of agricultural machinery, and particularly relates to a double-ridge four-row carrot harvester, which comprises a frame, a power device arranged in the frame, a walking device arranged below the frame, and oblique clamping and conveying devices arranged on both sides of the frame. Each of the oblique clamping and conveying devices comprises a conveying support, two clamping and conveying monomer groups arranged on the conveying support, a clamping driving wheel, a clamping driven wheel, a clamping conveying belt, and a plurality of tensioning wheel assemblies. The front ends of the two oblique clamping and conveying devices are provided with feeding devices, the upper sides of the two feeding devices are provided with supporting devices, the lower parts of the two supporting devices are fixedly connected with folding assemblies away from the frame, and the lower parts of the two feeding devices are provided with digging devices. The present application has a reasonable structure and can solve the problems of poor folding effect, easy lodging, and difficulty in guiding and feeding two rows of carrots into the oblique clamping and conveying device for conveying during harvesting.

The utility model discloses a full -automatic hydraulic pressure step -feed equipment, including drive mechanism, including a plurality of rotating electrical machines, rotating electrical machine drive screw rod autorotation, guide mechanism, including a plurality of mutually parallel arrangement's screw rod, and screw rod at least has two moving plate, chuck, each moving plate all are fixed with mutually coaxial arrangement's chuck, and the chuck has the passage hole that can hold or release to axle rod spare, wherein, moving plate and screw rod all are movable assembly, and the outer wall of screw rod includes outer thread section and light axle section, and adjacent two moving plates are movable assembly with same screw rod through a screw nut and a linear bearing respectively, and different chuck alternately holds and releases to axle rod spare and alternately axial motion, and the feed inlet of the chuck in upstream has a plurality of round table cavities, and the feed inlet of the chuck in downstream has a round table cavity. Adopt the utility model guide accurate, prevent skew, adapt to the characteristic of long axle rod spare.