A book delivery machine intelligent level correction method, medium and device

Abstract

The present application relates to the technical field of automation equipment control, and especially relates to a book delivery machine intelligent level correction method, medium and equipment, which simultaneously represents the real-time weight and the stacking loose state of the book pile through the vibration characteristic parameters, can continuously obtain the book pile height information, avoids the problems of empty grabbing and missed grabbing and book page bruising caused by poor height adaptation, extracts the book page deformation characteristic parameters through three-dimensional point cloud scanning and determines the first grabbing posture parameters, improves the adaptation ability to the book form difference, obtains the book page physical characteristic parameters through the air pressure pulse test and corrects the first grabbing posture parameters, realizes the stroke reduction and negative pressure reduction of the hard book page to avoid bruising, realizes the stroke increase and negative pressure increase of the soft book page to ensure the effect of adhesion, breaks through the limitation that pure visual detection cannot perceive the physical characteristics of the book, effectively solves the problems of air suction cup air leakage, book slipping and cover damage, realizes the adaptive level correction of the book delivery machine grabbing, and improves the grabbing precision and stability.

Description

Technical Field

[0001] This invention relates to the field of automated equipment control technology, and in particular to an intelligent horizontal correction method, medium, and device for a book feeder. Background Technology

[0002] In automated book cutting production lines, the book feeder plays a crucial role in transporting books one by one to the cutting mechanism. During continuous operation, the book feeder relies on a gripping mechanism to pick up books one by one from a continuously decreasing stack of books, ensuring that the books are delivered flat to the cutting station. However, as the height of the stack of books decreases during the gripping process, the books to be gripped generally exhibit differences in shape, such as warped covers, arched pages, and curled edges, as well as differences in physical properties such as paper material, thickness, and rigidity. This leads to multi-source uncertainties in the book feeder's gripping operation, placing extremely high demands on the adaptive capabilities of the gripping system.

[0003] Traditional book feeder gripping systems generally use a preset fixed stroke to control the downward pressure of the gripping mechanism, which cannot detect the continuous height change of the book stack as the gripping process progresses. When the book stack height is lower than the preset stroke, the suction cup cannot effectively contact the pages, resulting in empty gripping or missed gripping; when the book stack height is higher than the preset stroke, the suction cup presses down excessively, easily causing damage to the pages or even the book stack to collapse. Moreover, existing technologies mostly rely on pure visual detection or simple photoelectric sensors to obtain book status information, but visual detection can only identify geometric deformations such as warping and curling on the page surface, and cannot perceive physical characteristics such as the book's flexibility, stiffness, and airtightness. For subtle deformations such as slight warping and local dents, the vision system often has difficulty effectively identifying them, leading to frequent malfunctions such as air leakage during suction cup adsorption, book slippage, or cover damage. Existing book feeder gripping technology is difficult to meet the requirements of high-speed automated production lines for gripping accuracy, stability, and adaptability.

[0004] Therefore, how to achieve intelligent horizontal correction in the book feeder's grasping process, enabling it to adapt to dynamic changes in the height of the book stack, integrate multi-source sensor information to perceive the physical characteristics of the books, and improve the grasping accuracy and stability in high-speed automated production lines, has become an urgent problem to be solved. Summary of the Invention

[0005] To address the aforementioned technical problems, the present invention provides an intelligent leveling correction method for a book feeder. The book feeder is equipped with a conveyor belt, a gripper mechanism, and a suction cup. A stack of books to be leveled is placed on the conveyor belt. The intelligent leveling correction method for the book feeder includes the following steps:

[0006] S1. Based on the vibration characteristic parameters corresponding to the conveyor belt, the equivalent height of the book stack is obtained. The vibration characteristic parameters are used to characterize the real-time weight and loose stacking state of the book stack.

[0007] S2, determine the target position above the stack of books based on the sum of the equivalent height and the preset safety gap, and control the gripper mechanism to move to the target position.

[0008] S3 scans the pages on the upper surface of the stack of books to obtain three-dimensional point cloud data, and extracts deformation feature parameters based on the three-dimensional point cloud data.

[0009] S4. Based on the comparison results between the deformation characteristic parameters and the preset deformation judgment rules, determine the first gripping posture parameters of the suction cup.

[0010] S5, after the suction cup contacts the page according to the first gripping posture parameters, apply a test air pressure pulse to the page and collect the air pressure response signal and suction cup displacement signal.

[0011] S6. Based on the air pressure response signal and the suction cup displacement signal, the physical characteristic parameters of the book page are calculated, and the first gripping posture parameters are corrected based on the physical characteristic parameters to obtain the second gripping execution parameters. The physical characteristic parameters are used to characterize the local equivalent stiffness of the book page and the degree of adhesion and sealing between the suction cup and the book page.

[0012] S7 controls the gripper mechanism to perform book gripping operations based on the equivalent height and the second gripping execution parameters.

[0013] The present invention also provides a non-transitory computer-readable storage medium storing at least one instruction or at least one program, wherein the at least one instruction or at least one program is loaded and executed by a processor to implement the above-described intelligent level correction method for a book feeder.

[0014] The present invention also provides an electronic device, including a processor and the aforementioned non-transitory computer-readable storage medium.

[0015] This invention has at least the following beneficial effects: It utilizes vibration characteristic parameters to simultaneously characterize the real-time weight and looseness of the book stack, enabling continuous acquisition of book stack height information during conveyor belt operation, effectively avoiding missed grasps and page damage caused by poor height adaptation; it extracts page deformation characteristic parameters through 3D point cloud scanning and determines the first grasping posture parameters accordingly, transforming the page surface morphology into quantifiable maximum warp height, maximum indentation depth, and warp area ratio. This allows for automatic selection of standard suction cup gripping, offset suction cup center point gripping, or switching of gripper modes based on the actual flatness of the page, improving adaptability to differences in book shape; it obtains page physical characteristic parameters through air pressure pulse testing and corrects the first grasping posture parameters accordingly, reducing the stroke and negative pressure of hard pages to avoid damage, and increasing the stroke and negative pressure of soft pages to ensure a good fit. This overcomes the limitations of pure visual detection in perceiving the physical characteristics of books, effectively solving problems such as suction cup leakage, book slippage, and cover damage. It achieves adaptive horizontal correction throughout the entire book feeder grasping process, improving the grasping accuracy and stability in high-speed automated production lines. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of an intelligent horizontal correction method for a book feeder provided in Embodiment 1 of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It is understood that, where appropriate, the terms used to distinguish similar objects can be interchanged so that the invention can also be implemented in other embodiments besides the illustrated or described embodiments. Furthermore, the terms "including," "having," and any variations are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products, or devices.

[0020] Example 1

[0021] This embodiment provides an intelligent leveling correction method for a book feeder. The book feeder is equipped with a conveyor belt, a gripper mechanism, and a suction cup. A stack of books to be leveled is placed on the conveyor belt, such as... Figure 1 As shown, the intelligent level correction method for the book feeder includes the following steps:

[0022] S1. Based on the vibration characteristic parameters of the conveyor belt, the equivalent height of the stack of books is obtained.

[0023] The conveyor belt, serving as the material transport carrier for the book feeder, is horizontally or slightly inclined and mounted on the machine frame, forming a channel for transporting books from the previous process to the gripping station. The conveyor belt carries the stack of books and transports it to the designated area of ​​the gripping station, i.e., the working area of ​​the gripper mechanism. After the stack of books moves to the target position with the conveyor belt, the conveyor belt may pause or continue running depending on the specific working conditions, allowing the gripper mechanism to perform the book picking operation.

[0024] The gripper mechanism is installed above the book feeder frame and achieves displacement in three-dimensional space through linear modules, multi-axis robotic arms, or gantry structures. Its working range covers the book stack gripping station on the conveyor belt and the book conveying and cutting stations at the rear. Specifically, the gripper mechanism is equipped with a servo drive system inside or outside, and its encoder can provide feedback on the minute displacement signals of the suction cup. The gripper mechanism carries the suction cup and laser scanning equipment, moving from the initial position to the target position above the book stack. It remains stable at the target position, allowing the laser scanning equipment to complete the scanning operation. It is also used to drive the suction cup to press down, attract, and lift the book, completing the transfer of the book from the book stack to the subsequent process.

[0025] The suction cups are fixedly mounted on the end effector of the gripper mechanism and move synchronously with it. They are typically an array of multiple independent suction cups, each with its own independently controllable air path. Each suction cup's air path is connected in parallel to a high-frequency response pressure sensor and a proportional pressure regulating valve for adjusting the suction negative pressure. A microswitch or force sensor is installed at the connection between the suction cup and the gripper mechanism to confirm whether the suction cup is effectively contacting the book page. In this embodiment, the high-frequency response pressure sensor has a sampling rate ≥1kHz to collect the air pressure response signal.

[0026] Multiple books awaiting processing are stacked vertically to form a book pile, which is then placed at the gripping station on a conveyor belt. Each book is bound with several pages, with the top layer of pages being the object directly contacted by the suction cup. During continuous gripping, the height of the book pile continuously decreases. The top layer of pages in the pile generally exhibits morphological deformations such as cover warping, page arching, and edge curling, and the physical properties of the paper within the pile vary in terms of material, thickness, and rigidity. Correspondingly, before gripping, the book pile needs to be corrected for height deviations caused by the decrease in stack height, deformation deviations caused by page warping or dents, and adhesion deviations caused by differences in material rigidity and airtightness. This ensures that the suction cup conforms to the shape of the page surface, achieving stable adsorption and gripping without damaging the pages, and realizing adaptive horizontal correction of the book feeder's gripping posture.

[0027] The vibration characteristic parameters are a set of feature quantities extracted from the vibration signals of the conveyor belt support structure, including a multi-order dominant frequency amplitude sequence and vibration damping ratio, used to characterize the real-time weight and stacking looseness of the book stack. Specifically, the real-time weight is the total mass of the book stack at the current moment, which continuously decreases as the grasping process progresses. The vibration amplitude is positively correlated with the weight of the book stack; the greater the weight, the stronger the vibration response. The stacking looseness refers to the density / air gap size between the papers inside the book stack, reflecting the loose or tight characteristics of the stacking structure, and affecting the dissipation characteristics of vibration energy. Loose stacking has high damping, while tight stacking has low damping.

[0028] Correspondingly, when books are placed on a conveyor belt, their weight and the looseness of the stack alter the dynamic characteristics of the conveyor belt support structure. The greater the weight, the higher the vibration amplitude; the greater the looseness of the stack, the faster the vibration energy dissipates. Therefore, this embodiment uses MEMS triaxial accelerometers deployed at the conveyor belt support rollers to collect vibration signals in real time. Spectral analysis is then used to extract vibration characteristic parameters representing the real-time weight and the looseness of the stack. Combined with a pre-calibrated height mapping model, the stack thickness of the books is indirectly predicted, achieving a reverse calculation from vibration to height.

[0029] In one specific embodiment, S1 includes the following steps:

[0030] S11, collects the vibration time-domain signal of the conveyor belt during the conveying process.

[0031] S12 performs spectral analysis on the vibration time-domain signal to extract the multi-order dominant frequency amplitude sequence and vibration damping ratio.

[0032] S13, substitute the multi-order main frequency amplitude sequence and vibration damping ratio into the preset height mapping model to calculate the equivalent height of the stack of books.

[0033] The weight of the stack of books acts on the conveyor belt support structure, forming a dynamic load that directly affects the vibration response amplitude of the support structure. The looseness of the stack of books affects the transmission and dissipation of vibration energy between the stack and the structure. By placing acceleration sensors at the support rollers along the vibration transmission path, the time-domain vibration signal of the conveyor belt during the conveying process can be collected. The acquisition frequency is set to be sufficient to cover the dominant frequency range of the conveyor belt's mechanical vibration.

[0034] The time-domain vibration signal is converted into a frequency-domain representation using a Fast Fourier Transform (FFT). The complex time waveform is decomposed into a superposition of frequency components, resulting in a spectrum. The amplitude corresponding to each dominant frequency in the spectrum reflects the intensity of the vibration at that frequency and is positively correlated with the weight of the stack of books. By extracting the amplitude corresponding to each dominant frequency, a multi-order dominant frequency amplitude sequence is formed. The damping ratio is calculated using either the half-power bandwidth method in the frequency domain or the logarithmic decay method in the time domain, reflecting the rate of vibration energy dissipation and being related to the looseness of the stack of books. Correspondingly, for the resonance peak at the dominant frequency, the half-power bandwidth method is used, i.e., the frequency bandwidth between the half-power points on both sides of the resonance peak is divided by twice the resonance frequency to calculate the damping ratio of each order. The comprehensive damping ratio is taken as a characterization of the looseness of the stack of books; or the logarithmic decay method in the time domain is used to calculate the damping ratio based on the ratio of adjacent peak values ​​during free vibration.

[0035] The height mapping model fuses a multi-order dominant frequency amplitude sequence and vibration damping ratio into a single height value. Model coefficients can be predetermined through offline calibration experiments. Vibration data are collected under standard stacks of books with known weights and looseness, and the optimal coefficients β0 and β2 are obtained through regression fitting. i γ forms a height mapping function with a fixed constant.

[0036] In this embodiment, the preset height mapping model is: ,in, For equivalent height, Let ξ be the i-th amplitude in the multi-order dominant frequency amplitude sequence, i = 1, 2, ..., m, where m is the total number of amplitudes in the multi-order dominant frequency amplitude sequence, ξ is the vibration damping ratio, and β0 is the reference height offset constant, representing the reference height value when the vibration amplitude approaches zero. i γ is the height correction coefficient corresponding to the i-th order amplitude, representing the contribution weight coefficient of each order of the main frequency amplitude to the predicted height, and γ is the damping compensation coefficient, used to adjust the damping ratio term to the equivalent height correction strength.

[0037] As described above, by collecting the vibration signal of the conveyor belt and extracting the amplitude and damping ratio through spectrum analysis, the vibration characteristics can simultaneously characterize the real-time weight and looseness of the stack of books, improving the accuracy and information content of the stack status characterization; by calculating the equivalent height through multi-order dominant frequency amplitude sequence and vibration damping ratio, the height prediction integrates weight information and looseness correction, improving the accuracy and adaptability of height prediction.

[0038] S2, determine the target position above the stack of books based on the sum of the equivalent height and the preset safety gap, and control the gripper mechanism to move to the target position.

[0039] The equivalent height represents the height of the top page of the stack of books relative to the surface of the conveyor belt. When the equivalent height of the stack is obtained, the gripper mechanism is still in its initial standby position, far from the stack. Scanning has not yet begun. If the gripper mechanism's suction cup and laser profilometer are directly against the pages, not only will complete point cloud data not be obtained, but unknown page warping could also cause equipment collisions. Therefore, the gripper mechanism's suction cup and laser profilometer cannot be pressed directly down onto the stack surface in the vertical direction. A safety gap needs to be added to the equivalent height to position the gripper mechanism at a height as close to the stack as possible while remaining absolutely safe, allowing for high-precision laser scanning at this position.

[0040] The preset safety gap is a fixed distance allowance reserved to ensure the safe operation of the equipment. It prevents accidental scraping or collision between the suction cup or laser profilometer and the pages when the gripper mechanism, carrying the scanning device, approaches the stack of books. The specific value of the preset safety gap can be set by the implementer according to the actual situation, for example, 20mm to 30mm.

[0041] The vertical height component of the target location needs to meet two conditions: the gripper mechanism must be close enough to the stack of books so that the laser profilometer can acquire high-resolution, high signal-to-noise ratio 3D point cloud data; the gripper mechanism must be far enough away from the stack of books so that even if the pages are warped or curled, the gripper mechanism and its integrated equipment will not physically collide with the stack of books.

[0042] Therefore, in this embodiment, the height of the target position is the sum of the equivalent height and the preset safety gap. The horizontal coordinates of the target position are determined based on the actual position of the stack of books on the conveyor belt, and can be provided by the conveyor belt positioning sensor. In this embodiment, they are defined as the coordinates of the center point of the stack of books on the conveyor belt.

[0043] The gripper mechanism consists of a linear module or a multi-axis robotic arm driven by a servo motor. The motion control system converts the target position command into drive signals for the motors of each joint / axis, thereby achieving spatial positioning of the end effector.

[0044] As described above, by summing the equivalent height with the preset safety gap to determine the target position, the gripper mechanism can approach the surface of the stack of books at the minimum distance while ensuring absolute safety. This not only eliminates the collision risk caused by deformations such as page warping and curling, but also creates close-range scanning conditions for high-precision laser scanning, improves the coverage integrity and measurement accuracy of the three-dimensional point cloud data, and provides a data foundation for the subsequent extraction of deformation feature parameters.

[0045] S3 scans the pages on the upper surface of the stack of books to obtain three-dimensional point cloud data, and extracts deformation feature parameters based on the three-dimensional point cloud data.

[0046] In one specific embodiment, S3 includes the following steps:

[0047] S31, the line laser profilometer fixed to the gripper mechanism performs a line scan on the upper surface of the book stack to obtain three-dimensional point cloud data of the upper surface of the book pages.

[0048] S32 preprocesses the 3D point cloud data and performs plane fitting on the preprocessed 3D point cloud data to obtain the reference plane.

[0049] S33 calculates the height residual of each sampling point in the 3D point cloud data relative to the reference plane.

[0050] S34, extract the deformation feature parameters of the page based on the height residual, wherein the deformation feature parameters include at least one of the maximum warp height, maximum indentation depth and warp area percentage of the page.

[0051] In this process, a line laser profilometer obtains several height points along a single scan. It typically performs multiple scans in conjunction with the lateral movement of a gripper mechanism or the longitudinal movement of a conveyor belt, stitching together these scan lines to form a dense 3D point cloud covering the entire surface of the book page. The 3D point cloud data consists of a large number of discrete points with 3D spatial coordinates, each point representing the measured position and height value of a sampling location on the book page's surface.

[0052] The raw point cloud data contains a large number of invalid points from non-page areas such as conveyor belt edges, spine outlines, and background debris, as well as measurement noise introduced by ambient light and vibration. In the preprocessing stage, an algorithm automatically identifies and removes invalid data, retaining only the clean point cloud within the effective page-bearing area. Then, the least squares method is used to perform plane fitting on the effective point cloud, i.e., finding a plane in three-dimensional space that minimizes the sum of squared residuals from all sampling points to this plane. This plane is considered the ideal height distribution when the page is completely flat and serves as the zero reference benchmark for subsequent deformation determination. The preprocessing includes at least one of denoising, outlier removal, and region clipping. Those skilled in the art will recognize that existing denoising, outlier removal, region clipping, and plane fitting methods fall within the scope of this invention and will not be elaborated upon here.

[0053] The reference plane represents the theoretical height distribution assuming the book pages are perfectly flat. However, actual book pages have height deviations due to warping, indentations, and other factors. The height residual is the actual height of each sampling point minus the predicted height of the reference plane, used to quantitatively describe the degree of deformation at each location. A positive height residual indicates that the sampling point is convex relative to the reference plane, while a negative height residual indicates that the sampling point is concave relative to the reference plane.

[0054] The residual sequence contains height deviation information at various locations on the page surface. By statistically analyzing the extreme values ​​and distribution characteristics in the residual sequence, deformation characteristic parameters can be extracted: the maximum warping height representing the most severe bulge, the maximum indentation depth representing the most severe depression, and the area ratio of the warped region representing the breadth of deformation coverage. This allows for a complete characterization of the page's deformation state from both the severity and coverage dimensions.

[0055] Specifically, the maximum value of the height residual is determined as the maximum warping height, and the minimum value of the height residual is determined as the maximum indentation depth. A flatness judgment threshold is set, and the number of target sampling points whose absolute height residual value is greater than or equal to the flatness judgment threshold is counted. The proportion of the number of target sampling points in the total number of sampling points is set as the proportion of the warped area. The specific value of the flatness judgment threshold can be set by the implementer according to the actual situation; for example, in this embodiment, it is set to 0.3 mm.

[0056] The above-mentioned method, by calculating the height residual point by point and extracting the maximum warping height, maximum indentation depth and the area ratio of the warped region, fully describes the deformation state of the page from two dimensions: severity and coverage. This provides a clear and quantifiable basis for the accurate determination of the gripping posture in the future.

[0057] S4. Based on the comparison results between the deformation characteristic parameters and the preset deformation judgment rules, determine the first gripping posture parameters of the suction cup.

[0058] The process involves substituting the values ​​of the deformation characteristic parameters into the preset deformation judgment rules for comparison. Based on which judgment branch the comparison result falls into, the method and posture by which the suction cup should contact the page are determined.

[0059] In one specific embodiment, the first grasping posture parameters include grasping mode, grasping center point position, suction cup extension stroke, and suction negative pressure value, and the preset deformation judgment rules include:

[0060] When the maximum warping height is less than the first threshold and the maximum indentation depth is greater than the second threshold, the page flatness is determined to be good. The first gripping posture parameter is configured to suction cup gripping mode, the suction cup is kept vertically pressed down and the rated negative pressure is output.

[0061] When the maximum warping height is greater than or equal to the first threshold, or the maximum indentation depth is less than or equal to the second threshold, it is determined that there is local deformation of the page, and the first gripping posture parameter is configured to suction cup gripping mode, the gripping center point is offset, and the suction cup extension stroke is pre-adjusted.

[0062] When the area of ​​the warped region is greater than the preset area threshold, it is determined that there is a large area deformation of the page, and the first gripping posture parameter is configured to gripper gripping mode.

[0063] Specifically, when the maximum warpage height is less than the first threshold and the maximum indentation depth is greater than the second threshold, the page flatness is considered good. Although there is slight deformation, its amplitude is within the adaptive capability range of the suction cup. In this case, no special compensation is required, and the suction cup can form an effective seal within the contact area by pressing vertically downwards. Correspondingly, the gripping mode is the suction cup gripping mode, the gripping center point is the default geometric center of the page, the suction cup extension stroke is the default standard stroke, and the suction negative pressure value is the rated negative pressure.

[0064] When the maximum warpage height is greater than or equal to the first threshold, or the maximum indentation depth is less than or equal to the second threshold, it is determined that the page has localized bulges or indentations. Bulges can cause air leakage at the suction cup edges, while indentations can cause the suction cup center to be suspended. Therefore, it is not suitable to align the suction cup center with the core deformation area. Instead, the gripping center point is offset 5mm to 10mm outward from the deformation area, allowing the suction cup to adhere to a relatively flat area. Simultaneously, the suction cup extension stroke is pre-adjusted to allow for subsequent fine-tuning based on physical property parameters. Correspondingly, the gripping mode is suction cup gripping mode, the gripping center point is offset 5mm to 10mm from the geometric center of the page, the suction cup extension stroke is the pre-adjusted stroke, and the suction negative pressure value is the rated negative pressure.

[0065] When the area of ​​warped region exceeds a preset area threshold, the page is deemed to have undergone large-area deformation. In this case, the suction cup gripping mechanism faces a significant risk of air leakage. Neither offsetting the center point nor adjusting the suction cup extension stroke can guarantee an effective seal between the suction cup and the page over a sufficient area. This exceeds the physical effective boundary of the suction cup gripping mechanism, necessitating a switch to flexible gripper gripping mode. Correspondingly, the gripping mode is a gripper gripping mode, with the gripping center point located at the geometric center of the page. The suction cup extension stroke and suction negative pressure value are not applicable to this gripper gripping mode.

[0066] The specific values ​​of the first threshold, the second threshold, and the preset area threshold can be set by the implementer according to the actual situation. For example, in this embodiment, the first threshold is set to 3mm, the second threshold is -3mm, and the preset area threshold is 50%.

[0067] As described above, by comparing the continuously quantized deformation characteristic parameters with the preset judgment rules of discrete grading through multi-condition logic, the automatic grading judgment of the page deformation state is realized. By configuring the gripping posture parameters corresponding to different working conditions in a graded manner, a differentiated control strategy is realized, which improves the rationality of the configuration of the first gripping posture parameter and the adaptability of working conditions.

[0068] S5, after the suction cup contacts the page according to the first gripping posture parameters, apply a test air pressure pulse to the page and collect the air pressure response signal and suction cup displacement signal.

[0069] Among the first grasping posture parameters, the suction cup extension stroke gives the distance the suction cup extends downward from the end of the gripper mechanism. The gripper mechanism carries the suction cup and begins to descend from the target position. The suction cup first extends the pre-adjusted stroke. Under the combined motion of the overall descent of the gripper and the extension of the suction cup, the end of the suction cup gradually approaches and finally contacts the upper surface of the book page.

[0070] Traditional gripping methods immediately create a vacuum after the suction cup makes contact, neglecting the direct impact of the book's physical properties on the vacuum adsorption effect. This results in situations where the book is too rigid, failing to form an effective seal during vacuuming, or too flexible, potentially causing excessive deformation. Therefore, this embodiment applies a low-pressure, short-duration test pressure pulse before initiating the vacuum. Specifically, the normal adsorption negative pressure is typically 50kPa to 80kPa, and the normal evacuation time is typically 100ms to 300ms. This embodiment applies a 5kPa, 20ms test pressure pulse first to ensure that the excitation process does not damage the book while being sufficient to elicit the book's mechanical response characteristics. Correspondingly, the operation of not immediately creating a vacuum ensures that there is no negative pressure interference from active evacuation in the air path. The test pressure pulse reflects the passive sealing characteristics of the book in its natural fit.

[0071] After the test air pressure pulse is applied, the pressure in the air circuit does not instantly reach the target value or disappear instantly. A dynamic fluid response process occurs within the air circuit: the pressure rise rate is affected by the air circuit volume and sealing characteristics; the peak value is affected by the stability of the air source; and the decay phase is affected by the degree of air circuit leakage. The entire response process contains a wealth of information about the physical characteristics of the book page: the rise slope reflects airtightness, the deviation between the peak value and the command value reflects sealing performance, and the fall rate reflects the degree of leakage. Simultaneously, after the suction cup contacts the book page, the tiny compression displacement (on the order of millimeters) generated by the suction cup contact force and the minute lifting force produced by the air pressure pulse is fed back in real time by the servo encoder. All sensor data are uniformly collected by the control system, and after timestamp alignment, the air pressure response signal and suction cup displacement signal are obtained as the data basis for judging the physical characteristic parameters of the book page.

[0072] As described above, by actively applying a test air pressure pulse of known amplitude and duration after the suction cup contacts the book page and before vacuuming, and collecting the air pressure response signal and suction cup displacement signal, the physical response characteristics of the book page can be detected before actual adsorption is performed. This avoids adsorption failure or book page damage caused by directly vacuuming on pages that are too rigid or too flexible, and provides complete data support for calculating the physical characteristic parameters of the book page.

[0073] S6. Based on the air pressure response signal and the suction cup displacement signal, calculate the physical characteristic parameters of the page, and correct the first gripping posture parameters based on the physical characteristic parameters to obtain the second gripping execution parameters.

[0074] Among them, physical characteristic parameters are used to characterize the local equivalent stiffness of the book page and the degree of adhesion and sealing between the suction cup and the book page. Local equivalent stiffness is the ability of the book page to resist compression deformation in the local area of ​​the suction cup contact point, and the unit is N / mm. It is used to quantitatively characterize the softness or hardness of the book page. High stiffness indicates that the page is hard, such as coated paper or thick cover, and the downward stroke needs to be reduced to avoid crushing. Low stiffness indicates that the page is soft, such as newsprint or thin inner pages, and the downward stroke needs to be increased to ensure effective adhesion. The degree of adhesion and sealing quantitatively characterizes the airtight adhesion quality between the suction cup and the book page, with a value range of [0,1]. It reflects whether there is an air leakage channel at the contact interface between the suction cup and the book page. The closer the value is to 1, the better the sealing. If it is less than the preset adhesion threshold, it indicates that there is serious air leakage, and it is necessary to trigger re-suction or switch the gripping scheme.

[0075] Correspondingly, this embodiment converts the air pressure response signal and the suction cup displacement signal into physical characteristic parameters of the page, and then adjusts the control parameters of the suction cup in reverse according to the physical characteristic parameters to achieve the final correction of visual geometric coarse adjustment and tactile physical fine adjustment, thereby improving the accuracy and rationality of the second gripping execution parameters.

[0076] In one specific embodiment, the air pressure response signal includes pressure change and equilibrium pressure, the suction cup displacement signal includes compression displacement, and S6 includes the following steps:

[0077] S61, based on the pressure change, the effective adsorption area of ​​the suction cup, and the compression displacement, the local equivalent stiffness of the page is calculated.

[0078] S62 calculates the degree of adhesion and sealing between the suction cup and the book page based on the balance pressure and the preset target adsorption pressure.

[0079] The pressure change ΔP is the difference between the air pressure and the peak value during the application of the test air pressure pulse, from the reference value (usually 0). Under the same air pressure excitation, the pressure change is related to the ability of the book page to resist deformation.

[0080] The effective adsorption area A of the suction cup suctionIt is the effective area where the suction cup opening actually contacts the book page and forms a sealed area, which is determined by the suction cup model and size.

[0081] Compression displacement Δz contact This refers to the axial compression of the suction cup after it contacts the page, caused by the page's reaction force. This compression is fed back with high precision by the servo motor encoder, with an accuracy of 0.01mm. Under the same contact force, a larger compression displacement results in a softer page, while a smaller compression displacement results in a harder page.

[0082] Specifically, when the suction cup applies a test air pressure pulse to the page, the pressure in the air path generates a downward force F=ΔP×A on the effective area of ​​the suction cup. suction The force applied to the page surface causes a slight compressive deformation Δz in the page. contact According to Hooke's Law, F = k × x, within the elastic deformation range, the force is directly proportional to the deformation, and the proportionality constant is the stiffness k. Therefore, the local equivalent stiffness k... page =F / Δz contact =ΔP×A suction / Δz contact The higher the stiffness value, the more difficult it is to compress the instruction manual page, requiring a reduction in downward pressure; the lower the stiffness value, the easier it is to compress the instruction manual page, requiring an increase in downward stroke to ensure that the suction cup lip deforms sufficiently to form a seal.

[0083] Preset target adsorption pressure P command This is the target set value for testing the air pressure pulse, such as 5 kPa, which is the pressure value corresponding to the proportional pressure regulating valve command. It serves as an ideal reference value for calculating the degree of sealing. The closer the actual equilibrium pressure is to the target value, the better the sealing performance.

[0084] Balance pressure P eq This is the final pressure value after the air pressure stabilizes following the dynamic response following the application of a test air pressure pulse. It reflects the static sealing capability of the interface between the suction cup and the book page. A balanced pressure close to the target value indicates a good seal, while a pressure significantly lower than the target value indicates air leakage.

[0085] Specifically, the degree of sealing requires a comprehensive assessment of both static sealing capability and dynamic leakage. Static sealing capability is characterized by the closer the equilibrium pressure is to the preset target adsorption pressure, indicating a tighter fit between the suction cup opening and the book page surface, with no significant leakage channels. Dynamic leakage is characterized by a smaller pressure decay rate λ, indicating better airtightness at the contact interface and preventing rapid gas escape. The pressure decay rate λ is the slope of the air pressure decrease over time after the test pressure pulse is removed or during its maintenance phase, calculated from the pressure response curve, reflecting the leakage rate. A faster decay indicates a more obvious leakage channel between the suction cup and the book page.

[0086] The overall sealing performance, including static sealing capability and dynamic leakage, is then used to determine the degree of fit and seal, Q. seal =(P eq / P command )×e^(-α×λ). Where, P eq / P command The static sealing capability reflects the sealing performance and indicates whether the expected pressure has been achieved. The empirical correction coefficient α is an empirical coefficient that adjusts the weight of the attenuation rate term in the formula for calculating the degree of fit and seal. Its value range is [0.1, 0.5], and it is used to balance the relative contributions of sealing performance and leakage degree to the degree of fit and seal. e^(-α×λ) reflects the dynamic leakage degree of sealing performance and indicates whether the pressure can be maintained without attenuation.

[0087] In one specific embodiment, the second gripping execution parameters include gripping mode, suction cup compensation stroke, gripping center point position, and suction negative pressure value. S6 further includes the following steps:

[0088] When the degree of adhesion and sealing is less than the preset adhesion threshold, the suction cup is triggered to perform a shake-release and re-suction operation or switch to the gripper grasping mode.

[0089] When the degree of adhesion and sealing is greater than or equal to the preset adhesion threshold, the suction cup compensation stroke and adsorption negative pressure value are adjusted according to the local equivalent stiffness, and the second gripping execution parameters are determined according to the adjusted suction cup compensation stroke and adsorption negative pressure value.

[0090] When the sealing degree is less than the preset sealing threshold, it indicates a significant defect in the airtight seal between the suction cup and the book page, whether due to excessive deformation of the book page, an overly rough surface, or rigid material preventing effective sealing. In this case, even fine adjustments to the stroke and negative pressure cannot fundamentally solve the sealing problem. This embodiment employs two remedial approaches: one is a shake-and-release re-suction method, which involves briefly shaking the suction cup to reposition the book page or the suction cup opening, potentially breaking the current poor seal and establishing a new contact relationship; the other is switching the gripper, which means abandoning suction cup gripping and switching to a gripper gripping method when re-suction fails or the deformation is so severe that the suction cup solution is completely unusable, ensuring that the gripping task can still be completed.

[0091] When the seal is greater than or equal to the preset sealing threshold, it indicates that the basic seal between the suction cup and the page is good, providing the prerequisite for fine-tuning. At this point, the local equivalent stiffness of the page determines how to adjust: pages with high stiffness are less prone to deformation when the suction cup is pressed down, and excessive stroke or negative pressure can easily cause cover damage or page creases, so the compensation stroke should be appropriately reduced and the suction negative pressure lowered; pages with low stiffness are more prone to deformation with the suction cup when pressed down, so the compensation stroke needs to be appropriately increased to ensure that the suction cup lip deforms sufficiently to form a good seal, while the negative pressure should be appropriately increased to compensate for the insufficient support of the soft page.

[0092] Specifically, if the local equivalent stiffness k page Greater than the preset stiffness threshold k th Then, a first compensation amount ΔH is added to the basic suction cup compensation stroke. offset1 =-φ×(k page -k th This is to reduce the depth of pressure application and to increase the first negative pressure amount ΔH based on the rated negative pressure. offset1 =-ω×(k page -k th This is to protect the book pages. Where φ is the preset stroke adjustment coefficient and ω is the preset negative pressure adjustment coefficient.

[0093] If the local equivalent stiffness k page Less than or equal to the preset stiffness threshold k th Add a second compensation amount ΔH to the basic stroke. offset2 =+φ×(k th -k page This is to increase the depth of downward pressure and to add a second negative pressure ΔH based on the rated negative pressure. offset2 =+ω×(k th -k page To ensure adsorption force.

[0094] The adjusted suction cup compensation stroke and suction negative pressure value, combined with the gripping mode and gripping center point position, are integrated into the second gripping execution parameters.

[0095] The specific values ​​of the preset fit threshold, preset stiffness threshold, preset stroke adjustment coefficient, and preset negative pressure adjustment coefficient can be set by the implementer according to the actual situation. For example, in this embodiment, the preset fit threshold is set to 0.7, the preset stiffness threshold is set to 2N / mm, the preset stroke adjustment coefficient is set to 0.5mm, and the preset negative pressure adjustment coefficient is set to 0.02kPa.

[0096] As described above, by substituting the pressure change in the air pressure response signal and the compression displacement in the suction cup displacement signal into the stiffness calculation formula, the accurate calculation of the local equivalent stiffness of the book page is achieved. By integrating the ratio of the equilibrium pressure to the preset target adsorption pressure and the exponential decay term of the pressure decay rate into a single adhesion and sealing degree index, a quantitative assessment of the airtightness of the interface between the suction cup and the book page is achieved, integrating two dimensions: static sealing capability and dynamic leakage degree, thus improving the accuracy of the assessment of the adhesion and sealing degree. By setting a preset adhesion threshold as a prerequisite for parameter correction, when the adhesion quality is unreliable, the system actively triggers shaking and re-suction or switching of grippers, ensuring that the gripping action is performed on a reliable adhesion basis. Under the condition that the adhesion quality is qualified, the suction cup compensation stroke and adsorption negative pressure value are adjusted in reverse according to the stiffness of the book page, realizing the dual fusion correction of visual geometric coarse adjustment and tactile physical fine adjustment, thus improving the accuracy and rationality of the second gripping execution parameters.

[0097] S7 controls the gripper mechanism to perform book gripping operations based on the equivalent height and the second gripping execution parameters.

[0098] In this process, the gripper mechanism aligns the center of the suction cup assembly with the designated landing area on the book page based on the gripping center point position in the second gripping execution parameters. The gripper mechanism descends from the determined target position, with the final target position set as the equivalent height of the book stack minus a preset gripping offset, ensuring that the suction cup tip makes flexible contact with the book page surface rather than a rigid impact. The specific value of the preset gripping offset can be set by the implementer according to actual conditions; for example, in this embodiment, it is set to 8mm.

[0099] Simultaneously, according to the suction cup compensation stroke in the second gripping execution parameters, the suction cup extends the pre-adjusted compensation stroke amount at the end of the descent action. Rigid pages require a smaller compensation stroke to reduce the pressing depth and protect the page, while soft pages require a larger compensation stroke to increase the pressing depth and ensure a proper fit. This allows the suction cup to contact pages of different materials with varying stroke amounts.

[0100] Once the suction cup is in position and makes contact with the pages, it will perform an adsorption or clamping action according to the gripping mode specified in the second gripping execution parameter:

[0101] If the gripping mode is suction cup gripping mode, the vacuum generator will be activated to create a vacuum. The suction force is determined by the suction negative pressure value in the second gripping execution parameter. Rigid book pages require a lower suction negative pressure value to avoid damaging the cover; soft book pages require a higher suction negative pressure value to ensure sufficient suction force.

[0102] If the gripping mode is the gripper gripping mode, vacuuming is not performed, and the gripper directly performs the gripping action to ensure that the gripping task can still be completed.

[0103] After the suction cups pick up the pages and begin lifting, the pages may experience slight displacement or wobbling due to the combined effects of gravity, inertia, and air resistance. This can cause a momentary change in the seal between the suction cup opening and the pages, resulting in fluctuations in the vacuum level. Without compensation, a slight momentary drop in vacuum can lead to page slippage. In this embodiment, a vacuum sensor monitors the vacuum level of each suction cup in real time. When the vacuum level of a certain suction cup is detected to be lower than a preset maintenance threshold, a proportional pressure regulating valve instantaneously increases the negative pressure value of that suction cup. This is a dynamic upward adjustment based on the negative pressure value in the second gripping execution parameters to compensate for the momentary deterioration of the seal. During the dynamic compensation process, if a single suction cup cannot establish a stable vacuum despite repeated adjustments, the suction cup is recorded as abnormal, and related parameter optimizations are triggered in subsequent batch iterations. The specific value of the preset maintenance threshold can be set by the implementer according to the actual situation; for example, in this embodiment, it is set to 85% of the rated negative pressure.

[0104] Once the vacuum level reaches the preset adsorption confirmation threshold and is maintained for a predetermined time, typically 50ms, the adsorption state is considered stable. The gripper mechanism then performs a lifting action to separate the book from the stack. After being lifted into position, the gripper mechanism moves to the feeding position of the rear cutting process, releases the book to the cutting station, and closes the vacuum or opens the grippers, completing the gripping task.

[0105] After the grabbing is complete, return to step S1 to start the loop for the next book. At this time, the height of the stack of books has decreased, and the system re-predicts the equivalent height based on the new vibration signal, driving the entire process to adaptively update.

[0106] In one specific embodiment, the intelligent level correction method for the book feeder further includes the following steps:

[0107] S8, obtain the cutting quality inspection data of the back-end cutting process, wherein the cutting quality inspection data includes at least one of the following: the straightness error of the edge of the book after cutting, the cutting offset, and the area of ​​edge burrs and wrinkles.

[0108] S9, calculate the quality loss value based on the cutting quality inspection data, and iteratively update the second grabbing execution parameters based on the quality loss value.

[0109] In one specific embodiment, S9 includes the following steps:

[0110] S91, with a preset batch as the iteration period, calculates the average edge straightness error, average cutting offset, and average cutting burr and wrinkle area of ​​all books in the current batch.

[0111] S92, based on the error proportions of average edge straightness error, average cutting offset, and average edge burr and wrinkle area, determine the first dynamic weight corresponding to edge straightness error, the second dynamic weight corresponding to cutting offset, and the third dynamic weight corresponding to edge burr and wrinkle area, respectively.

[0112] S93. Construct a quality loss function based on the edge straightness error and its corresponding first dynamic weight, the cutting offset and its corresponding second dynamic weight, and the area of ​​edge burrs and wrinkles and its corresponding third dynamic weight.

[0113] S94. Calculate the gripping posture correction factor based on the partial derivative of the quality loss function with respect to the second gripping execution parameter. The gripping posture correction factor is used to characterize the degree and direction of the gripping posture deviation of the current batch of books on the back-end cutting quality.

[0114] S95, based on the grasping posture correction factor, iteratively update at least one of the deformation judgment threshold, suction cup compensation stroke and adsorption negative pressure value in the preset deformation judgment rules.

[0115] Among them, the cutting quality inspection data is the quantitative measurement data obtained by the back-end cutting process to inspect the quality of the cut book. It is acquired by setting up a high-resolution line scan camera at the material outlet of the cutting mechanism. It includes the straightness error of the edge, the cutting offset, and the area of ​​the edge burrs and wrinkles, which serve as the basis for measuring whether the front-end gripping posture is reasonable and whether there are any transmissible defects.

[0116] Side straightness error E straight This refers to the degree of deviation of the trimmed edge of the book from the ideal straight line, reflecting the straightness of the trimmed edge. If the book is not kept horizontal when gripped, the trimmed edge will appear non-straight. Trimming offset E offset This is the offset distance of the actual cut position of the book after trimming relative to the preset cut position. It is used to quantify the inaccuracy of the cut position caused by the gripping positioning deviation. If the book is offset in the XY plane during gripping, the cut line will deviate from the preset position. The area of ​​trimmed burrs and wrinkles is A. burr It is the area of ​​burrs and wrinkles near the cut edge caused by poor cutting after cutting. It is used to quantify the deterioration of cutting quality caused by uneven gripping. If the suction cup has too much negative pressure during gripping, causing local pressure deformation of the page, the cutting blade cannot apply force evenly during cutting, which will produce burrs and wrinkles.

[0117] The defect distribution can change between different batches; the previous batch might mainly have offset issues, while the next batch might have burr problems. Using fixed weights makes it impossible to dynamically adjust priorities. Therefore, this embodiment normalizes the error proportions to obtain first, second, and third dynamic weights, ensuring that the defect with the largest average error in the current batch receives the highest weight, thus achieving adaptive control.

[0118] After obtaining the dynamic weights, the three types of defect data are linearly weighted and summed according to their respective weights to obtain the quality loss value L. cut This represents the overall evaluation of the cutting quality of this batch: the greater the quality loss value, the worse the cutting quality, and the more urgent the need to adjust the capture parameters.

[0119] By calculating the partial derivative δ of the quality loss function with respect to the second grabbing execution parameter grip This allows us to obtain the sensitivity of the second grabbing execution parameter to quality loss: a positive partial derivative indicates that increasing the parameter will increase the loss, while a negative partial derivative indicates that increasing the parameter will decrease the loss.

[0120] The deformation judgment threshold, suction cup compensation stroke, and suction negative pressure value are determined based on the corresponding δ. grip The direction and magnitude are adjusted synchronously in the same direction according to their respective correction coefficients, δ grip The positive and negative signs indicate the adjustment direction, and the absolute value indicates the adjustment range. The updated parameters are stored in the system and applied to the grasping control of the next batch of books.

[0121] The above-mentioned progressive closed loop of capturing parameter fine-tuning, cutting quality improvement, and reverse fine-tuning enables the book feeder control system to have the self-optimization capability to continuously adapt to material batch differences throughout its entire life cycle, eliminating the batch quality drift problem caused by parameter fixation in traditional solutions.

[0122] This embodiment utilizes vibration characteristic parameters to simultaneously characterize the real-time weight and looseness of the book stack, enabling continuous acquisition of book stack height information during conveyor belt operation. This effectively avoids missed gripping and page damage caused by poor height adaptation. By extracting page deformation characteristic parameters through 3D point cloud scanning and determining the first gripping posture parameters accordingly, the page surface morphology is transformed into quantifiable maximum warp height, maximum indentation depth, and warp area ratio. This allows for automatic selection of standard suction cup gripping, offset suction cup center point gripping, or switching of gripper modes based on the actual flatness of the page, improving adaptability to differences in book shape. Furthermore, by obtaining physical characteristic parameters of the page through air pressure pulse testing and correcting the first gripping posture parameters, the travel distance for hard pages is reduced and negative pressure is lowered to avoid damage, while the travel distance for soft pages is increased and negative pressure is raised to ensure a better fit. This overcomes the limitations of purely visual detection, which cannot perceive the physical characteristics of the book, effectively solving problems such as suction cup leakage, book slippage, and cover damage. It achieves adaptive horizontal correction throughout the entire gripping process of the book feeder, improving gripping accuracy and stability in high-speed automated production lines.

[0123] Example 2

[0124] Embodiment 2 of the present invention provides a non-transitory computer-readable storage medium, which can be disposed in an electronic device to store at least one instruction or at least one program related to implementing a method in the method embodiment. The at least one instruction or at least one program is loaded and executed by the processor to implement the intelligent level correction method for a book feeder provided in the above embodiment.

[0125] Example 3

[0126] Embodiment 3 of the present invention provides an electronic device, which includes a processor and the non-transitory computer-readable storage medium of Embodiment 2 of the present invention.

[0127] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for intelligent horizontal correction of a book feeder, characterized in that, The book feeder is equipped with a conveyor belt, a gripper mechanism, and a suction cup. A stack of books to be leveled is placed on the conveyor belt. The method includes the following steps: S1, based on the vibration characteristic parameters corresponding to the conveyor belt, the equivalent height of the stack of books is obtained, wherein the vibration characteristic parameters are used to characterize the real-time weight and loose stacking state of the stack of books; S2, determine the target position above the stack of books based on the sum of the equivalent height and the preset safety gap, and control the gripper mechanism to move to the target position; S3, scan the pages on the upper surface of the stack of books to obtain three-dimensional point cloud data, and extract deformation feature parameters based on the three-dimensional point cloud data; S4. Based on the comparison result between the deformation characteristic parameters and the preset deformation judgment rules, determine the first gripping posture parameters of the suction cup; S5, after the suction cup contacts the page according to the first grasping posture parameters, a test air pressure pulse is applied to the page, and the air pressure response signal and suction cup displacement signal are collected; S6. Based on the air pressure response signal and the suction cup displacement signal, calculate the physical characteristic parameters of the book page, and correct the first gripping posture parameters based on the physical characteristic parameters to obtain the second gripping execution parameters. The physical characteristic parameters are used to characterize the local equivalent stiffness of the book page and the degree of adhesion and sealing between the suction cup and the book page. S7, based on the equivalent height and the second gripping execution parameters, control the gripper mechanism to perform a book gripping operation.

2. The intelligent horizontal correction method for a book feeder according to claim 1, characterized in that, S1 includes the following steps: S11, Collect the vibration time-domain signal of the conveyor belt during the conveying process; S12, Perform spectral analysis on the vibration time-domain signal to extract the multi-order main frequency amplitude sequence and vibration damping ratio; S13, Substitute the multi-order dominant frequency amplitude sequence and the vibration damping ratio into the preset height mapping model to calculate the equivalent height of the stack of books.

3. The intelligent horizontal correction method for a book feeder according to claim 1, characterized in that, S3 includes the following steps: S31, the line laser profilometer fixed to the gripper mechanism performs a line scan on the upper surface of the book stack to obtain three-dimensional point cloud data of the upper surface of the book pages; S32, preprocess the three-dimensional point cloud data, and perform plane fitting on the preprocessed three-dimensional point cloud data to obtain a reference plane, wherein the preprocessing includes at least one of denoising, outlier removal and region clipping. S33, Calculate the height residual of each sampling point in the three-dimensional point cloud data relative to the reference plane; S34, extract the deformation feature parameters of the page based on the height residual, wherein the deformation feature parameters include at least one of the maximum warping height, maximum indentation depth, and percentage of warped area.

4. The intelligent horizontal correction method for a book feeder according to claim 3, characterized in that, The first grasping posture parameters include grasping mode, grasping center point position, suction cup extension stroke, and suction negative pressure value. The preset deformation determination rules include: When the maximum warping height is less than the first threshold and the maximum indentation depth is greater than the second threshold, the page flatness is determined to be good, and the first gripping posture parameter is configured to suction cup gripping mode, keeping the suction cup pressed vertically downward and outputting rated negative pressure. When the maximum warping height is greater than or equal to the first threshold, or the maximum indentation depth is less than or equal to the second threshold, it is determined that the page has local deformation, and the first gripping posture parameter is configured to suction cup gripping mode, the gripping center point is offset and the suction cup extension stroke is pre-adjusted. When the area of ​​the warped region is greater than a preset area threshold, it is determined that the page has a large area deformation, and the first gripping posture parameter is configured as a gripper gripping mode.

5. The intelligent horizontal correction method for a book feeder according to claim 1, characterized in that, The air pressure response signal includes pressure change and equilibrium pressure, and the suction cup displacement signal includes compression displacement. S6 includes the following steps: S61, the local equivalent stiffness of the page is calculated based on the pressure change, the effective adsorption area of ​​the suction cup, and the compression displacement. S62, calculate the degree of adhesion and sealing between the suction cup and the book page based on the balanced pressure and the preset target adsorption pressure.

6. The intelligent horizontal correction method for a book feeder according to claim 5, characterized in that, The second gripping execution parameters include gripping mode, suction cup compensation stroke, gripping center point position, and suction negative pressure value. S6 also includes the following steps: When the degree of adhesion and sealing is less than the preset adhesion threshold, the suction cup is triggered to perform a shake-release and re-suction operation or switch to the gripper grasping mode. When the degree of adhesion and sealing is greater than or equal to the preset adhesion threshold, the suction cup compensation stroke and the adsorption negative pressure value are adjusted according to the local equivalent stiffness, and the second gripping execution parameters are determined according to the adjusted suction cup compensation stroke and adsorption negative pressure value.

7. The intelligent horizontal correction method for a book feeder according to claim 1, characterized in that, The method further includes the following steps: S8, Obtain cutting quality inspection data of the back-end cutting process, wherein the cutting quality inspection data includes at least one of the following: the straightness error of the edge of the book after cutting, the cutting offset, and the area of ​​edge burrs and wrinkles. S9, calculate the quality loss value based on the cutting quality detection data, and iteratively update the second grabbing execution parameters based on the quality loss value.

8. The intelligent horizontal correction method for a book feeder according to claim 7, characterized in that, S9 includes the following steps: S91, with a preset batch as the iteration period, calculates the average edge straightness error, average cutting offset, and average edge burr and wrinkle area of ​​all books in the current batch. S92, based on the error ratios of the average edge straightness error, the average cutting offset, and the average edge burr and wrinkle area, determine the first dynamic weight corresponding to the edge straightness error, the second dynamic weight corresponding to the cutting offset, and the third dynamic weight corresponding to the edge burr and wrinkle area, respectively. S93, construct a quality loss function based on the edge straightness error and its corresponding first dynamic weight, the cutting offset and its corresponding second dynamic weight, and the area of ​​the cutting burrs and wrinkles and its corresponding third dynamic weight; S94, calculate the gripping posture correction factor based on the partial derivative of the quality loss function with respect to the second gripping execution parameter, wherein the gripping posture correction factor is used to characterize the degree and direction of the gripping posture deviation of the current batch of books on the back-end cutting quality. S95, based on the grasping posture correction factor, at least one of the deformation judgment threshold, the suction cup compensation stroke, and the adsorption negative pressure value in the preset deformation judgment rule is iteratively updated.

9. A non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores at least one instruction or at least one program segment, characterized in that, The at least one instruction or the at least one program segment is loaded and executed by the processor to implement the intelligent horizontal correction method for the book feeder as described in any one of claims 1-8.

10. An electronic device, characterized in that, Includes a processor and the non-transitory computer-readable storage medium as described in claim 9.

Images