35 results about "Position bias" patented technology

The invention provides a load-carrying travelling crane positioning method and device based on a gauge pile, and the method comprises the steps: recording a real-time coding value and a calibration signal of an encoder of a travelling crane at the current moment when the travelling crane receives the calibration signal of the gauge pile, and determining the actual position of the travelling crane relative to a reference point at the current moment; based on the real-time coding value and the standard coding value, determining the coding deviation of the driving at the current moment; based on the actual position and the standard positioning value, determining the position deviation of the travelling crane at the current moment; the standard deviation of the encoder is determined based on the real-time encoding value of the encoder within the past preset number of times of driving; when the absolute value of the position deviation is greater than a preset position deviation threshold value and the standard deviation is within a preset standard deviation interval, performing equivalent updating on the current equivalent between the encoder and the crane rail based on the encoding deviation and the position deviation to obtain a new equivalent; according to the invention, accurate positioning of the load-carrying travelling crane can be realized.

The invention relates to the technical field of workpiece machining, in particular to an intelligent robot incremental forming device and a machining method thereof.The intelligent robot incremental forming device comprises a robot base, a robot body, a positioning system, a multi-dimensional force sensor, a self-adaptive force compensation system, an auxiliary floating device and a forming tool head; the control system obtains a simulation coordinate # imgabs0 # of the workpiece, the positioning system measures an actual coordinate # imgabs1 # of the workpiece, and a translation distance # imgabs2 # is carried out in the normal direction of the actual coordinate to obtain a positioning coordinate # imgabs3 #; the control system controls the robot body to move to a positioning coordinate # imgabs4 #; the control system controls the positioning system to position the mark points of the workpiece so as to obtain an actual point group # imgabs5 #; and a coordinate offset rotation matrix # imgabs6 # and a translation vector # imgabs7 # are obtained, and the coordinate offset rotation matrix # imgabs8 # and the translation vector # imgabs9 # are added into a control system, so that position deviation compensation during workpiece machining is realized. According to the invention, the processing and forming quality of the workpiece is improved.

The invention discloses a part relative position detection method and system under three-axis movement and a medium, and relates to the technical field of position detection, and the method comprises the steps: carrying out the scanning of a target part, obtaining three-dimensional point cloud data, carrying out the region segmentation recognition, carrying out the clustering analysis, obtaining a point cloud cluster of a plurality of feature points, carrying out the global fitting, and determining a plurality of feature space coordinates; constructing a part local coordinate system, mapping the plurality of feature space coordinates to the part local coordinate system for position calculation, and obtaining a plurality of feature relative position relation diagrams; traversing a plurality of feature relative position relation graphs to perform deviation marking, obtaining a plurality of out-of-tolerance items, performing binding storage on the out-of-tolerance items and a target part, constructing a position detection anomaly database, performing position deviation training, and constructing a position deviation prediction channel to perform feedback optimization on a detection path. The technical problems that in the prior art, part relative position detection precision is insufficient, and efficiency is low are solved, and the technical effect of improving part relative position detection precision and detection efficiency is achieved.

The invention aims to provide a vertical shaft heading machine attitude determination method and system, and belongs to the field of underground engineering monitoring. The method comprises the following steps: establishing a fixed coordinate system and a carrier coordinate system, acquiring propagation time by using a well wall ultrasonic transmitter and a platform receiver, and resolving absolute coordinates of the receiver and absolute attitude of a platform based on spatial spherical intersection; meanwhile, a gyroscope and an accelerometer are used for collecting angular velocity and specific force information, and high-frequency and low-frequency predicted attitudes are calculated. And combining the absolute attitude with the predicted attitude through a fusion algorithm to obtain an optimal attitude angle, and constructing a conversion matrix to process specific force information so as to solve an optimal position deviation. The advantages of high penetrability of ultrasonic waves and continuity of inertial navigation are utilized, the problem that laser guidance is prone to failure in the dust and muddy water environment is solved, and high-precision and anti-interference real-time monitoring of the posture of the vertical shaft heading machine is achieved.

The application provides a positioning adjustment method, device and equipment of a track plate and a readable storage medium, relates to the technical field of measurement and control, and comprises the following steps: obtaining three-dimensional design coordinates of all measuring devices; establishing a plane equation based on the three-dimensional design coordinates; obtaining three-dimensional measurement coordinates of all measuring devices; determining elevation adjustment amounts of all measuring devices; calculating three-dimensional projection coordinates of all measuring devices; determining lateral adjustment amounts and longitudinal adjustment amounts of all measuring devices based on the three-dimensional projection coordinates and the three-dimensional design coordinates; and adjusting the position of a target track plate based on the elevation adjustment amounts, the lateral adjustment amounts and the longitudinal adjustment amounts of all measuring devices. The application can accurately obtain the position deviation of the current position relative to the plate plane by establishing the plate plane equation, calculating the deviation between the measurement position coordinates and the design coordinates to the directions parallel to the plate plane and perpendicular to the plate plane, and quickly and accurately adjusting the track plate to the design position by using the fine adjustment mechanism.

In order to accurately calculate the amount of deviation between a main shaft and a slave shaft disposed obliquely with respect to the main shaft, a calculation device is configured so as to be capable of: acquiring a first value indicating the positional deviation of the main shaft and a second value indicating the positional deviation of the slave shaft disposed obliquely with respect to the main shaft by a predetermined angle; performing at least one of conversion of the acquired first value to a third value indicating a positional deviation of the main axis in a predetermined coordinate system and conversion of the acquired second value to a fourth value indicating a positional deviation of the slave axis in the predetermined coordinate system; calculating a deviation amount between the positional deviation of the main axis and the positional deviation of the slave axis in the predetermined coordinate system on the basis of the first value and the fourth value, the second value and the third value, or the third value and the fourth value; and outputting the calculated deviation amount.

This calculating device is designed for the precise calculation of a deviation between a master axis and a slave axis arranged at an inclination relative to the master axis, such that it is capable of performing any of the following actions: first, obtaining a first value representing a positional deviation of the master axis; second, obtaining a positional deviation of the slave axis arranged at an inclination at a predetermined angle relative to the master axis; third, converting the obtained first value into a third value representing the positional deviation of the master axis in a predetermined coordinate system; and fourth, converting the obtained second value into a fourth value representing a positional deviation of the slave axis in the predetermined coordinate system.to calculate a deviation magnitude in the specified coordinate system between the positional deviations of the master axis and the slave axis based on the first and fourth values, the second and third values, or the third and fourth values; and to output the calculated deviation magnitude.

The application discloses a feature map upsampling method and application, and belongs to the technical field of image processing, and comprises the following steps: S1, mapping an input feature map into a position bias map; wherein, elements (Delta x i , Delta y j ) of the i-th row and the j-th column of the position bias map represent position bias at a position ([i / s], [j / s]) of the input feature map; s is a scaling ratio; H and W are respectively the length and the width of the input feature map; S2, adding the elements (Delta x i , Delta y j ) of the i-th row and the j-th column of the position bias map to the position ([i / s], [j / s]) to obtain a sampling position map of the input feature map; S3, performing interpolation on the input feature map based on the sampling position map to obtain a high-resolution feature map, so that the upsampling operation of the input feature map is realized. The application can realize the upsampling with low complexity and fast calculation speed on the basis of ensuring the upsampling performance.

The invention discloses an error compensation tapping method and device based on bottom hole deviation and a medium. The method comprises the steps that the compensation limit value delta lim of a target threaded bottom hole is obtained; establishing a target coordinate system, and acquiring actual coordinates and expected coordinates of the target threaded bottom hole on the target coordinate system; based on the actual coordinates and the expected coordinates, the x-axis direction position deviation delta x and the y-axis direction position deviation delta y of the target threaded bottom hole are obtained; based on the position deviation delta x in the x-axis direction, the position deviation delta y in the y-axis direction and a specified deviation threshold value delta, whether the position of the screw tap needs to be compensated or not is judged; and when the position of the screw tap needs to be compensated, the compensation direction and the compensation amount are obtained based on the compensation limit value delta lim. The technical problem that the threaded hole is unqualified due to the position error of the threaded bottom hole can be effectively solved, and the percent of pass of threaded hole machining can be remarkably improved.

The application discloses a cutting method and a cutting device, and the method comprises the following steps: mapping and calibrating the position of a positioning plate and the position on a numerical control cutting machine by using a reference plate with marked points cut through; establishing a coordinate system xoy by using the reference plate; placing a printed material with the back side upward on the positioning plate, so that two right-angle edges of the printed material are coincident with the x-axis and the y-axis of the coordinate system xoy; reading the position deviation of the marked points relative to the cut-through holes and the angle deviation of the connecting line of the two marked points relative to the connecting line of the two cut-through holes by using a camera; synchronously grabbing and placing the printed material on the numerical control cutting machine; and correcting and outputting a cutting pattern according to the position deviation and the angle deviation by the numerical control cutting machine. The method establishes a reference coordinate system by using the reference plate, determines the position deviation and the angle deviation of the marked points by using the camera, synchronously transmits the deviation information to the control software of the numerical control cutting machine, corrects and outputs the cutting pattern, realizes high-precision cutting and high-efficiency cutting.

A three-axis machine tool track contour error calculation method considering geometric and motion characteristics comprises the following steps: firstly, establishing a geometric error equivalent mapping model of a three-axis machine tool, and uniformly describing a positioning error, a straightness error, an angle error and an inter-axis perpendicularity error of each motion axis; the geometric error is mapped to a tool nose space position error through homogeneous coordinate transformation; secondly, a three-ring servo linkage control equivalent model containing a position loop, a speed loop and a current loop is established, a dynamic response relation between interpolation instruction displacement and actual axis displacement is obtained through transfer function derivation, and a three-axis machine tool servo linkage control equivalent model is established; integrating the geometric error mapping result with the three-axis linkage control dynamic position deviation to realize the prediction of the actual track of the tool nose of the tool; and finally, based on the interpolation instruction position point and the predicted actual position point, a contour error is calculated by adopting an interpolation line segment perpendicular foot method. The method has the advantages of unified modeling, high calculation efficiency and high engineering applicability.

The application discloses a kind of information fusion methods suitable for multi-architecture heterogeneous cooperative navigation, belong to cooperative navigation technical field.The application includes respectively receiving the carrier inertial navigation position data that inertial navigation system real-time sends, the carrier position data that global positioning system real-time sends and the carrier relative position data that multi-carrier cooperative positioning system real-time sends;The position deviation amount is obtained by comparing the received position data with the inertial navigation position data at corresponding moment;According to the current moment existing global positioning system data or cooperative positioning system data, the position deviation amount is used as input, and the inertial navigation solution result at the current moment is corrected by filtering mode;Finally, the corrected carrier navigation information is output, the GPS position information and cooperative positioning relative position information can be simultaneously accessed into combined navigation system by the application, and effective fusion is carried out.

To provide an aligner device configured to be capable of reducing a time of transfer including processing of positional deviation correction of a tabular work-piece.SOLUTION: An aligner device comprises: a robot hand including a hand body that mounts and holds a work-piece; a work-piece lift mechanism that lifts a work-piece transferred by the robot hand and mounted on the hand body; a sensor 3 that has a downward sensor surface and is disposed so as to be capable of detecting an outer shape of a work-piece made to be close to the sensor surface by the work-piece lift mechanism; positional deviation amount calculation means 6 for calculating amounts of positional deviation in X, Y, and θ directions from a reference position of the work-piece, from the outer shape of the work-piece acquired by the sensor 3; X-Y direction positional deviation amount correction means 81 for correcting the X direction positional deviation amount and the Y direction positional deviation amount of the work-piece on the basis of the amounts of positional deviation in the X-Y directions; and θ direction positional deviation amount correction means 82 for correcting the θ direction positional deviation amount of the work-piece on the basis of the amount of positional deviation in the θ direction.SELECTED DRAWING: Figure 3

The present application relates to the technical field of device positioning, in particular to a large rotating device full-position intelligent positioning system, comprising: determining a three-dimensional coordinate set of each search direction according to a reference coordinate system of a current rotating device; calculating an attitude angle deviation of a current search direction and a preset search direction, correlating a position deviation through the attitude angle deviation, and configuring a position deviation matrix of each reference point; calculating a deviation probability of each reference point under a combination of multiple search directions, and combining into a deviation probability matrix; correlating the deviation probability of each reference point with each search direction according to the deviation probability matrix, coordinating and optimizing the probability distribution of the search direction through the relative spatial coordinates between each reference point; mapping the optimized probability distribution of each search direction into the reference coordinate system, correcting the three-dimensional coordinates of each search direction, and obtaining the final output positioning result. The accuracy and efficiency of rotating device positioning are realized.

The application relates to the technical field of buoy positioning, and discloses a digital buoy positioning method and system based on satellite data. The method comprises the following steps: acquiring satellite data and preprocessing the satellite data, and extracting feature information of a buoy; dividing a buoy position partition based on the feature information, and constructing a mapping relationship between the feature information and the position partition; subsequently, extracting a key partition, calculating a combination matching degree of position data to obtain a position bias combination; then, updating positioning information in the position partition based on the position bias combination; finally, generating a buoy positioning management library according to the updated mapping relationship. Through the steps of accurately extracting feature information, scientifically dividing a position partition, and establishing a reasonable mapping relationship, the method improves the accuracy and timeliness of buoy positioning, and the generated positioning management library can provide reliable position support for buoy-related applications.

The application discloses a marine mobile channel adaptive coding modulation transmission method and system, relates to the technical field of wireless communication, and comprises the following steps: acquiring dynamic sensing data, spatial position and channel state information history sequence of a mobile platform; reconstructing the absolute motion velocity vector of an antenna in a reference navigation coordinate system based on the antenna installation position bias feature; extracting an effective radial velocity scalar in combination with a line-of-sight direction vector; determining the dynamic channel coherence time boundary according to the effective radial velocity scalar, performing truncation processing on the channel sequence to obtain a target prediction input sequence; generating a reference signal-to-noise ratio prediction value by using a time series prediction model; determining a nonlinear rollback compensation amount based on the angular acceleration obtained by differentiating the angular velocity, and then correcting the reference prediction value to obtain a target signal-to-noise ratio and performing physical layer resource mapping; and the application can effectively cope with the prediction failure caused by the dramatic mutation of the carrier attitude, and improve the robustness and throughput of the marine communication link.

The invention discloses a workpiece positioning method and a workpiece positioning device. A correction block is arranged in a machine table, and for any inner side wall of the correction block, a probe is adopted to obtain the actual distance between the probe and the inner side wall; adjusting the position of the probe in the machine table according to the actual distance corresponding to each inner side wall; acquiring position coordinates of each preset point location of the target workpieces by adopting a probe, and obtaining a set of the target workpieces according to the left sides of the positions of the plurality of preset point locations; and compensation values of the target workpiece in the width direction, the length direction and the height direction are obtained according to the position coordinate set, and the initial position of the target workpiece is compensated according to the compensation values. By arranging the correction block, the position of the probe is adjusted based on the actual distance of the inner side wall of the correction block, and the installation error and calibration deviation of the probe can be effectively eliminated; compensation values of a workpiece in the length direction, the width direction and the height direction are calculated through a collected point position coordinate set, and the position deviation generated in the clamping link can be comprehensively covered.

The application discloses a part relative position detection method and system under three-axis movement and a medium, relates to the technical field of position detection, and comprises the following steps: scanning a target part to obtain three-dimensional point cloud data, performing region segmentation and identification, and performing clustering analysis to obtain point cloud clusters of a plurality of feature points for global fitting to determine a plurality of feature space coordinates; a part local coordinate system is constructed, the plurality of feature space coordinates are mapped to the part local coordinate system for position calculation to obtain a plurality of feature relative position relationship graphs; deviation marks are added to the plurality of feature relative position relationship graphs, a plurality of out-of-tolerance items are obtained, and the target part is bound and stored to construct a position detection anomaly database, and position deviation training is performed to construct a position deviation prediction channel to feed back and optimize a detection path. The application solves the technical problems of insufficient part relative position detection precision and low efficiency in the prior art, and achieves the technical effects of improving part relative position detection precision and detection efficiency.

The application provides a method and system for determining position deviation, wherein the method comprises the following steps: obtaining first photoelectric data and second photoelectric data corresponding to a conveyed object; bringing the first photoelectric data into a corresponding first mathematical model to obtain an actual horizontal coordinate value, and bringing the second photoelectric data into a corresponding second mathematical model to obtain an actual vertical coordinate value; subtracting a standard horizontal coordinate value from the actual horizontal coordinate value to obtain a horizontal coordinate deviation, and subtracting a standard vertical coordinate value from the actual vertical coordinate value to obtain a vertical coordinate deviation; and taking the horizontal coordinate deviation and the vertical coordinate deviation as the position deviation of the conveyed object. The first photoelectric data and the second photoelectric data are respectively brought into the corresponding first mathematical model and the second mathematical model to determine the actual horizontal coordinate value and the actual vertical coordinate value of the conveying mechanism, and then the horizontal coordinate deviation and the vertical coordinate deviation are determined, so that the technical problem that the position deviation of the conveyed object cannot be determined is solved, and the technical effect of improving the conveying accuracy is achieved.

The embodiment of the application provides an AGV parameter calibration method, device, equipment, medium and product, relates to the field of engineering machinery control, and the method replaces the cumbersome process that manual data collection, deviation recording in the prior art use physical tools by automatically collecting the attitude angle difference of AGV at the preset starting point and terminal point and the deviation distance of the actual terminal point of AGV and the preset straight line path, automatically optimizes the rudder wheel angle bias value based on the attitude angle difference, and automatically optimizes the zero position bias coefficient based on the deviation distance, replaces the manual parameter adjustment process of manual dependence on empirical formula or trial-and-error method in an algorithm-driven parameter optimization mode, does not need manual repeated participation in testing and adjustment, reduces manual intervention in the calibration process, shortens the calibration period, and the parameter optimization process is accurate and controllable, avoids the uncertainty caused by human experience, solves the low AGV parameter calibration efficiency problem in the prior art, and realizes the rapid and accurate calibration of AGV parameters.

The invention provides an RAG answer quality evaluation method based on a large language model. The method comprises the following steps: inputting questions into two to-be-compared RAG systems to generate an answer pair corresponding to each question; comparing the lengths of the answer pairs, and if the length difference exceeds a preset threshold value, enabling the RAG system to which the short answers belong to generate answers again through an answer length weak alignment method; if the length difference of the regenerated answers still does not meet the preset threshold value, aligning the lengths of the shorter answers with the lengths of the longer answers through an answer length strong alignment method; scoring the answer pairs with the aligned lengths respectively; exchanging an answer sequence for re-scoring, and taking the sum of the two scores as a final score; and 5, comparing the answer scores of the two RAG systems to be compared, and drawing a box plot based on multiple test results. According to the method, the problems of length prejudice, position prejudice and test instability in the original scheme are solved by adopting the modes of answer alignment, position exchange, multiple tests and the like, and the evaluation problem of the RAG system capability is more accurately solved.

The application aims to provide a shaft heading machine posture determination method and system, and belongs to the underground engineering monitoring field. The method establishes a fixed coordinate system and a carrier coordinate system, uses a shaft wall ultrasonic transmitter and a platform receiver to obtain a propagation time, and calculates an absolute coordinate of the receiver and an absolute posture of the platform based on a spatial spherical intersection. Meanwhile, the gyro and accelerometer are used to collect angular velocity and specific force information, and high-frequency and low-frequency predicted postures are calculated. The absolute posture and the predicted posture are combined to obtain an optimal posture angle through a fusion algorithm, and a conversion matrix is constructed to process the specific force information to calculate the optimal position deviation. The application uses the advantages of strong ultrasonic penetration and continuous inertial navigation to solve the problem of laser guidance failure in a dusty and muddy environment, and realizes high-precision, anti-interference and real-time monitoring of the shaft heading machine posture.

The invention discloses a coordinate measuring machine based on laser tracking, positioning and calibration, relates to the technical field of coordinate measurement, and aims to solve the problem that the measurement precision of measurement data is influenced by the difference of working environments on the basis of a basic technical principle of three-coordinate laser measurement. According to the basic content, a position deviation value is obtained through rough calculation in cooperation with real-time position coordinates generated by a workpiece to be measured along with temperature changes, the key part is a set active deviation assembly, and an active deviation instruction is generated through cooperation of the influence of manual intervention and the position deviation value in essence. According to the method, the position change and the temperature change in the variable-temperature special-shaped ball head in the active deviation assembly are changed to generate an error coefficient associated with the whole to-be-measured workpiece in the measurement process, and the error coefficient is used for carrying out secondary modification on the real-time position coordinates, so that more accurate measurement coordinates are obtained, and the measurement precision is improved. Measurement errors caused by working environment differences are effectively eliminated, and the overall measurement precision and stability of the coordinate measuring machine are improved.

Systems, apparatuses, and methods are described for minimizing prompt order bias in a large language model (LLM). Using an original input prompt for an LLM, that may include instructions and ordered list, a plurality of different LLM input prompts may be generated. A plurality of LLM outputs may be determined, for example, by providing the plurality of LLM input prompts comprising the original instructions but with the order of the list permutated. A positional bias of the LLM may appear differently in the plurality of LLM outputs, for example, based on the differing list orders of the plurality of LLMs. A final LLM output may be generated, for example, by aggregating the LLM outputs to minimize the effects of positional bias.

The invention discloses a real-time fitness posture recognition method based on local and global feature fusion. The method comprises the following steps: firstly, extracting preliminary features of an input image through a CSPNeXt backbone network; afterwards, by utilizing an improved AFT-RPB reasoning module, by introducing relative position bias and fusing a Sigmoid weighting mechanism, the modeling capability of the model for a complex dependency relationship between key points is enhanced, and local details and global context information are effectively balanced; further, probability graph model optimization is carried out on the spatial relation between the key points through an NCRFs reasoning module, and the consistency of attitude estimation is improved; meanwhile, a lightweight depth block formed by depth separable convolution is adopted to replace traditional large kernel convolution, so that the complexity and parameter quantity of the model are reduced; finally, key point coordinate prediction is converted into a classification task based on a SimCC algorithm, quantization errors are reduced, and high-precision and real-time fitness posture recognition is achieved. The reasoning speed and the lightweight level of the model are remarkably improved, and the method is suitable for real-time application scenes with limited resources.

The invention relates to the technical field of equipment positioning, in particular to a large rotating equipment all-position intelligent positioning system which comprises the following steps: determining a three-dimensional coordinate set of each search direction according to a reference coordinate system of current rotating equipment; calculating an attitude angle deviation between the current search direction and a preset search direction, associating the attitude angle deviation with the position deviation, and configuring a position deviation matrix of each reference point; calculating the deviation probability of each datum point under the multi-search direction combination, and combining the deviation probability into a deviation probability matrix; according to the deviation probability matrix, the deviation probability of each datum point is associated with each search direction, and the probability distribution of the search directions is coordinated and optimized according to the relative space coordinates between the datum points; and mapping the optimized probability distribution of each search direction into a reference coordinate system, and correcting the three-dimensional coordinate of each search direction to obtain a final output positioning result. And the accuracy and efficiency of positioning of the rotating equipment are realized.

The application discloses a method and device for secondary positioning of a seabed node based on quadrant division. 2 The method comprises the following steps: establishing an orthogonal coordinate system with the current position of the seabed node as a coordinate origin; positioning a target window in the orthogonal coordinate system with a set shot point offset distance as a limit; dividing the target window into n 2 quadrants based on the coordinate axes, wherein n is greater than or equal to 2; judging whether the difference amplitude of the dynamic corrected shot point data first arrival wave between the quadrants meets the error requirement; if not, determining the offset of the current position of the seabed node relative to the real position according to the difference amplitude, adjusting the current position, returning to the above step until the difference amplitude of the dynamic corrected shot point data first arrival wave between the quadrants is small enough, and determining that the current position of the seabed node is the real position. The method can verify the correctness of the node position in marine OBN exploration, and accurately position the node with a position deviation.

The present invention provides a method for measuring and correcting the spatial position error of holes in aviation structural parts, comprising the following steps: S10, selecting a reference hole from pre-connected holes of a skin and a long stringer; S20, acquiring an image of the reference hole through a visual system, and using an algorithm to calculate the pixel position coordinates of the reference hole in the image coordinate system; S30, based on the visual system and a sensor, and through the pixel position coordinates and camera parameters of the visual system, calculating the position coordinates of the reference hole in the camera coordinate system, and calculating the position coordinates of the reference hole in the aircraft coordinate system according to the hand-eye relationship and the robot base coordinate system; S40, calculating the position error of the reference hole and the position deviation of the fastening hole to be made according to the position coordinate value of the reference hole in the aircraft coordinate system and the theoretical coordinate value of the reference hole in the aircraft coordinate system.

The embodiments of the present disclosure provide a model training method, medium, apparatus and computing device for removing position bias. For any task, either a click task or a completion task, during the model training process, based on the main model, the user characteristics and object characteristics of the sample object in the task are used to obtain a simulated prediction value corresponding to the task; at the same time, based on the first slave model, the object position characteristics of the sample object are used to obtain the position influence value corresponding to the sample object; and the main model and the slave model are iteratively trained using the obtained position influence value, so that the trained main model can remove the position bias, so that when the main model is used to recommend an object, the influence of the object's location on the recommendation result can be eliminated, and the objects that potential users are really interested in can be obtained, thereby improving user satisfaction and further improving the platform's recommendation conversion rate.