A method and system for real-time adjustment of screed height based on industrial vision

Abstract

The present application relates to the technical field of height control, and discloses a leveling machine height real-time adjustment method and system based on industrial vision, the method comprising: coaxially tracking a target working device and a working object to obtain a transient splashing image stream; dynamically inverting the transient splashing image stream to obtain an adaptive attenuation factor; based on the adaptive attenuation factor, analyzing the deformation potential of a current action point of the target working device to obtain an estimated deformation amount; comparing the estimated deformation amount with the elevation of a reference surface of the working object, mapping the deviation of the subsequent working height of the target working device, and obtaining a feedforward compensation coefficient; data coupling and coding the feedforward compensation coefficient and the current execution elevation of the target working device to obtain a driving instruction sequence; and based on the driving instruction sequence, accurately controlling the target working device to obtain a real-time adjusted height; the present application can improve the efficiency of leveling machine height real-time adjustment based on industrial vision.

Description

Technical Field

[0001] This invention relates to the field of height control technology, and in particular to a method and system for real-time height adjustment of a leveling machine based on industrial vision. Background Technology

[0002] In the field of height control technology, traditional real-time height adjustment methods lack dynamic visual perception of the contact area between the working equipment and the working object. They cannot effectively capture and analyze the working status information such as transient splashes generated during the operation, and it is difficult to accurately predict the deformation trend of the working object based on real-time data of the operation process. Height adjustment is carried out by relying solely on preset parameters or simple contact detection, which leads to deviations in the response to the actual deformation of the working object, and the accuracy of the leveling machine height adjustment is difficult to guarantee.

[0003] Existing height adjustment compensation strategies lack analytical models related to dynamic parameters of the operation process. They cannot deduce the deformation potential of the work object through key technical parameters such as energy transfer attenuation and material plastic flow during the operation. The comparison method between the estimated deformation and the elevation of the work object's reference surface is relatively crude, failing to achieve refined zoning and quantitative analysis of deviations. This results in a lack of scientific technical basis for calculating the feedforward compensation coefficient, poor adaptability of the height adjustment compensation mechanism, and ultimately low efficiency of real-time height adjustment, failing to meet the technical requirements of dynamic height adjustment in industrial leveling operations. Therefore, how to improve the efficiency of real-time height adjustment of leveling machines based on industrial vision has become an urgent problem to be solved. Summary of the Invention

[0004] This invention provides a method and system for real-time height adjustment of a leveling machine based on industrial vision, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a method for real-time height adjustment of a leveling machine based on industrial vision, comprising:

[0006] S01. Perform coaxial tracking on the target work equipment and the work object to obtain the transient splash image stream of the work object;

[0007] S02. Perform dynamic trend inversion on the transient splash image stream to obtain the adaptive attenuation factor of the transient splash image stream;

[0008] S03. Based on the adaptive attenuation factor, perform deformation potential analysis on the current action point of the target working equipment to obtain the estimated deformation amount of the current action point.

[0009] S04. Discretely compare the estimated deformation with the reference elevation of the work object, and based on the comparison results, perform deviation mapping on the subsequent working height of the target work equipment to obtain the feedforward compensation coefficient of the target work equipment.

[0010] S05. Data coupling encoding is performed on the feedforward compensation coefficient and the current execution elevation of the target working equipment to obtain the drive command sequence of the target working equipment;

[0011] S06. Based on the drive command sequence, the target working equipment is precisely controlled to obtain the real-time adjustment height of the target working equipment.

[0012] In a preferred embodiment, the coaxial tracking of the target work equipment and the work object to obtain the transient splash image stream of the work object includes:

[0013] Visual focus is locked on the contact area between the front end of the actuator of the target working equipment and the working object to obtain a high-definition focused image of the contact area;

[0014] Based on the high-definition focused image, impact phase analysis is performed on the target operating equipment to obtain the timing action parameters of the target operating equipment.

[0015] Based on the timing action parameters, the high-definition focused image is synchronously acquired and triggered to obtain the transient change image sequence of the contact field;

[0016] Dynamic deformation tracking is performed on the splash area of ​​the transient image sequence to obtain the instantaneous displacement field of the splash area;

[0017] Based on the instantaneous displacement field, the transient change image sequence is corrected point by point to obtain the transient splash image stream of the working object.

[0018] In a preferred embodiment, the step of performing dynamic trend inversion on the transient splash image stream to obtain the adaptive attenuation factor of the transient splash image stream includes:

[0019] Trajectory tracking is performed on discrete splash particles in the transient splash image stream to obtain the motion trajectory map of the discrete splash particles;

[0020] Based on the motion trajectory diagram, the fracture intensity of the discrete splash particles is analyzed at the breakup nodes to obtain the fragmentation energy distribution of the discrete splash particles;

[0021] Based on the fragmentation energy distribution, the attenuation of the discrete splash particles is extrapolated to obtain the energy transfer attenuation rate of the discrete splash particles;

[0022] Based on the energy transfer attenuation rate, the overall attenuation trend of the transient splash image stream is reconstructed by attenuation curve to obtain the adaptive attenuation factor of the transient splash image stream.

[0023] In a preferred embodiment, the step of performing attenuation deduction on the discrete splash particles based on the fragmentation energy distribution to obtain the energy transfer attenuation rate of the discrete splash particles includes:

[0024] Topological sorting of the particle nodes in the fragmented energy distribution yields a directed tree for the energy transfer of the discrete splash particles.

[0025] Energy allocation quantization is performed on adjacent level particles of the energy transfer directed tree to obtain the inter-level energy allocation coefficients of the discrete splash particles;

[0026] The path energy dissipation factor of the discrete splash particles is obtained by parametrically fusing the interstage energy allocation coefficient with the spatial displacement vector of the discrete splash particles.

[0027] Based on the path energy dissipation factor, the energy transfer attenuation rate of the discrete splash particles is calculated, wherein the formula for calculating the energy transfer attenuation rate is:

[0028] ;

[0029] In the formula, The energy transfer attenuation rate is... Let be the index number of the terminal particle in the directed energy transfer tree. The total number of all terminal particles in the energy transfer directed tree. For the energy transfer directed tree, the first The measured kinetic energy of each end particle The initial input energy is used to transfer energy to the root node of the directed tree. The energy transfer directed tree is from the root node to the first node. The index of intermediate nodes along the path of each terminal particle. The energy transfer directed tree is from the root node to the first node. All intermediate nodes along the path of the terminal particle. For the energy transfer directed tree, the first Interstage energy allocation coefficients of intermediate nodes The preset spatial dissipation constant, For the energy transfer directed tree, the first The displacement vector magnitude between an intermediate node and its direct successor node. It is a natural constant.

[0030] In a preferred embodiment, the step of analyzing the deformation potential of the current point of action of the target working equipment based on the adaptive attenuation factor to obtain the estimated deformation of the current point of action includes:

[0031] Spatial domain mapping is performed on the adaptive attenuation factor to obtain the bearing capacity influence region of the adaptive attenuation factor;

[0032] Based on the bearing capacity influence area, the material density at the current point of action in the target operating equipment is evaluated for compression response to obtain the compression stiffness at the current point of action.

[0033] Based on the compressive stiffness, load response tracking is performed on the weak area at the current point of application to obtain the plastic flow triggering threshold at the current point of application.

[0034] Based on the plastic flow triggering threshold, the settlement trend of the current point of action is converged to obtain the estimated deformation of the current point of action.

[0035] In a preferred embodiment, the step of discretizing the estimated deformation with the reference elevation of the work object, and based on the comparison result, mapping the subsequent working height of the target work equipment to obtain the feedforward compensation coefficient of the target work equipment, includes:

[0036] Based on the distribution points of the estimated deformation, the spatial coordinates of the reference surface elevation of the work object are indexed point by point to obtain the pairing data of the estimated deformation and the reference surface elevation at the same point.

[0037] The elevation values ​​of the paired data pairs at the same location are compared numerically to obtain the original set of deviation values ​​for the paired data pairs at the same location.

[0038] The positive deviation values ​​of the original deviation set are marked as raised deviations, and the negative deviation values ​​of the original deviation set are marked as concave deviations.

[0039] By performing partitioned topological mapping on the bulging deviation and the concave deviation, a polarization deviation distribution map of the estimated deformation is obtained.

[0040] Based on the polarization deviation distribution map, the subsequent working height of the target working equipment is inversely compensated to obtain the original compensation height sequence of the target working equipment;

[0041] The original high-compensation program sequence is subjected to adjacent-point amplitude limiting smoothing to obtain the smoothed high-compensation program sequence of the original high-compensation program.

[0042] The feedforward compensation coefficients of the target operating equipment are obtained by normalizing the dimensions of the smooth compensation high-order sequence.

[0043] In a preferred embodiment, the step of performing partitioned topological mapping on the raised deviation and the recessed deviation to obtain the polarization deviation distribution map of the estimated deformation includes:

[0044] Clustering and hierarchical merging of the raised particles of the raised deviation are performed to obtain the topological skeleton of the main ridge of the raised deviation.

[0045] By fitting the directional trend of the concave particles of the concave deviation, the topological skeleton of the concave valley bottom line of the concave deviation is obtained.

[0046] Based on the topological framework of the main ridge of the uplift and the topological framework of the valley bottom, the boundaries of the spheres of influence of the uplift-type deviation and the depression-type deviation are divided into equipotential regions to obtain the Voronoi partition map of the uplift-type deviation and the depression-type deviation.

[0047] Calculate the polarization intensity of spatial points in the Volonuova partition map, wherein the formula for calculating the polarization intensity is:

[0048] ;

[0049] In the formula, Spatial point polarization intensity, For the spatial point, The sign function of the Volonouva partition map. It is a natural constant. The preset nonlinear attenuation coefficient, The maximum value of the average distance from the spatial point to all deviation particles within the corresponding partition in the Volonouva partition map. The average Euclidean distance from the spatial point to all deviation particles within the partition to which it belongs in the Volonouva partition map;

[0050] Based on the polarization intensity, the Volonuowa partition map is filled point by point to obtain the polarization deviation distribution map of the estimated deformation.

[0051] In a preferred embodiment, the step of data coupling encoding the feedforward compensation coefficient with the current execution elevation of the target operating equipment to obtain the drive command sequence of the target operating equipment includes:

[0052] The feedforward compensation coefficients are non-uniformly divided to obtain the variable periodic compensation amount of the feedforward compensation coefficients;

[0053] Displacement mapping is performed on the current execution elevation of the target operating equipment to obtain the desired pose set of the current execution elevation;

[0054] Based on the variable period compensation amount, the desired pose set is superimposed and corrected periodically to obtain the temporal fusion desired pose of the feedforward compensation coefficient and the current execution elevation.

[0055] The joint space trajectory of the target working device is obtained by inverse kinematic reconstruction of the temporal fusion desired pose.

[0056] Based on the joint space trajectory, pulse width encoding is performed on the target working equipment to obtain the drive command sequence of the target working equipment.

[0057] In a preferred embodiment, the precise control of the target working equipment based on the drive command sequence to obtain the real-time adjustment height of the target working equipment includes:

[0058] The drive instruction sequence is segmented into time-axis segments to obtain single-cycle action primitives of the drive instruction sequence;

[0059] Based on the single-cycle action primitive, the target working equipment is driven and triggered to obtain the instantaneous output torque parameter of the target working equipment;

[0060] Based on the instantaneous output torque parameter, the actuator of the target working equipment is displaced and tractioned to obtain the real-time spatial pose of the actuator;

[0061] The vertical displacement of the real-time spatial pose is measured to obtain the real-time adjustment height of the target working equipment.

[0062] To address the aforementioned problems, the present invention also provides a real-time height adjustment system for a leveling machine based on industrial vision, the system comprising:

[0063] A coaxial tracking module is used to perform coaxial tracking of the target working equipment and the working object to obtain the transient splash image stream of the working object;

[0064] The dynamic trend inversion module is used to perform dynamic trend inversion on the transient splash image stream to obtain the adaptive attenuation factor of the transient splash image stream.

[0065] The deformation potential analysis module is used to analyze the deformation potential of the current action point of the target working equipment based on the adaptive attenuation factor, and obtain the estimated deformation of the current action point.

[0066] The deviation mapping compensation module is used to perform discrete comparison between the estimated deformation and the reference surface elevation of the work object, and based on the comparison result, to perform deviation mapping on the subsequent working height of the target work equipment to obtain the feedforward compensation coefficient of the target work equipment.

[0067] The data coupling encoding module is used to perform data coupling encoding on the feedforward compensation coefficient and the current execution elevation of the target working equipment to obtain the drive instruction sequence of the target working equipment;

[0068] A precision control module is used to precisely control the target working equipment based on the drive command sequence, so as to obtain the real-time adjustment height of the target working equipment.

[0069] Compared with the prior art, the present invention has the following beneficial effects:

[0070] 1. This invention relies on industrial vision technology to achieve coaxial tracking of the working equipment and the working object, accurately capture transient splash image streams and complete dynamic trend inversion. Through adaptive attenuation factor, it analyzes the deformation potential of the working point, and can quantitatively obtain accurate estimated deformation. It realizes refined analysis of dynamic technical parameters in the leveling operation, improves the accuracy of deformation prediction of the working object, and provides a scientific and practical technical basis for the height adjustment of the leveling machine, making the preliminary analysis of height adjustment more targeted.

[0071] 2. This invention achieves precise deviation mapping of working height through discrete comparison and polarization deviation analysis. The generated feedforward compensation coefficient can be efficiently coupled and encoded with the current operating elevation of the equipment, transforming it into a drive command sequence adapted to the operation of the equipment. This enables precise driving and height control of the leveling machine, completing real-time adjustment of the working height. This effectively improves the efficiency and accuracy of real-time height adjustment of the leveling machine, ensuring a high degree of match between the equipment's height execution actions and the work compensation requirements, and guaranteeing the overall effect of leveling work height control. Attached Figure Description

[0072] Figure 1 This is a flowchart illustrating a method for real-time height adjustment of a leveling machine based on industrial vision, according to an embodiment of the present invention.

[0073] Figure 2 A functional block diagram of a real-time height adjustment system for a leveling machine based on industrial vision is provided in an embodiment of the present invention.

[0074] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0075] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0076] This application provides a method for real-time height adjustment of a leveling machine based on industrial vision. The executing entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the method for real-time height adjustment of a leveling machine based on industrial vision can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0077] Reference Figure 1 The diagram shown is a flowchart illustrating a real-time height adjustment method for a leveling machine based on industrial vision, according to an embodiment of the present invention. In this embodiment, the real-time height adjustment method for a leveling machine based on industrial vision includes:

[0078] S01. Perform coaxial tracking on the target work equipment and the work object to obtain the transient splash image stream of the work object;

[0079] In this embodiment of the invention, the step of coaxially tracking the target working equipment and the working object to obtain the transient splash image stream of the working object includes:

[0080] Visual focus is locked on the contact area between the front end of the actuator of the target working equipment and the working object to obtain a high-definition focused image of the contact area;

[0081] Based on the high-definition focused image, impact phase analysis is performed on the target operating equipment to obtain the timing action parameters of the target operating equipment.

[0082] Based on the timing action parameters, the high-definition focused image is synchronously acquired and triggered to obtain the transient change image sequence of the contact field;

[0083] Dynamic deformation tracking is performed on the splash area of ​​the transient image sequence to obtain the instantaneous displacement field of the splash area;

[0084] Based on the instantaneous displacement field, the transient change image sequence is corrected point by point to obtain the transient splash image stream of the working object.

[0085] By using the lens focusing mechanism of the industrial vision acquisition equipment, the field of view is precisely aimed at the area where the front end of the target work equipment's actuator comes into contact with the work object. The lens focal length is continuously adjusted until all visual details within the contact area are clearly presented, completing the visual focus locking operation on the contact area. At this point, the contact area image captured and recorded by the industrial vision acquisition equipment is the high-definition focused image of the contact area.

[0086] Based on the visual information such as the contact state and positional relationship between the actuator of the target work equipment and the work object presented in the high-definition focused image, the complete action stages of the actuator of the target work equipment acting on the work object are sorted out, the visual feature identifiers of the start and end of each action stage are identified, the impact action phase of the actuator of the target work equipment is divided according to these identifiers, and then the relevant information such as the action occurrence time, action duration, and action stage sequence corresponding to each impact action phase is extracted. The integrated information is the timing action parameter of the target work equipment.

[0087] The extracted timing parameters of the target equipment are used as the trigger for the industrial vision acquisition device. According to the occurrence time and duration of each action stage in the timing parameters, the industrial vision acquisition device is controlled to start image acquisition at the corresponding time node. During the duration of each action stage, the images of the contact area are captured continuously. The acquisition process is completely synchronized with the action rhythm of the target equipment. The multiple frames of contact area images continuously acquired during this process are arranged in the order of acquisition time. The resulting image combination is the transient change image sequence of the contact area.

[0088] First, in each frame of the transient image sequence, the area where splashing occurs within the contact domain of the work object is defined as the splash area based on visual characteristics. Then, the changes in the position, shape, range, and other visual characteristics of the splash area in different frames are compared frame by frame. The position movement trajectory of each visual feature point in the splash area is tracked in the continuous frame images. The position movement information of all feature points at each acquisition time is integrated to form an information set that reflects the real-time displacement state of each position point in the splash area, which is the instantaneous displacement field of the splash area.

[0089] Based on the real-time displacement information of each feature point recorded in the instantaneous displacement field of the splash area, the position of each frame of the transient change image sequence is corrected pixel by pixel. The display position of the corresponding pixel in the image is adjusted according to the actual displacement state of each point, so that the visual presentation of the splash area in each frame of the image matches the actual displacement state. After completing the point-by-point correction operation of all frames of the image, all the corrected frames of the image are continuously and dynamically stitched together in time order. The resulting continuous dynamic image is the transient splash image stream of the working object.

[0090] The beneficial effects include: precise focusing and lock-on of industrial vision acquisition equipment enables the clear capture of complete visual information of the contact area, providing a clear and comprehensive image foundation for subsequent phase analysis. Impact phase analysis based on high-definition focused images allows for the accurate extraction of equipment timing parameters. Synchronous image acquisition based on this ensures that the contact area image acquisition rhythm perfectly matches the equipment's operational actions, guaranteeing the timeliness and continuity of the transient image sequence acquisition. Precise delineation and dynamic tracking of the splash area completely reconstructs the displacement state of the splash area. The resulting instantaneous displacement field serves as a precise basis for image correction. Point-by-point correction ensures a high degree of match between the image presentation and the actual displacement state. The resulting transient splash image stream possesses high realism and completeness, providing accurate and practical visual data support for subsequent visual analysis related to leveling machine height adjustment.

[0091] S02. Perform dynamic trend inversion on the transient splash image stream to obtain the adaptive attenuation factor of the transient splash image stream;

[0092] In this embodiment of the invention, the step of performing dynamic trend inversion on the transient splash image stream to obtain the adaptive attenuation factor of the transient splash image stream includes:

[0093] Trajectory tracking is performed on discrete splash particles in the transient splash image stream to obtain the motion trajectory map of the discrete splash particles;

[0094] Based on the motion trajectory diagram, the fracture intensity of the discrete splash particles is analyzed at the breakup nodes to obtain the fragmentation energy distribution of the discrete splash particles;

[0095] Based on the fragmentation energy distribution, the attenuation of the discrete splash particles is extrapolated to obtain the energy transfer attenuation rate of the discrete splash particles;

[0096] Based on the energy transfer attenuation rate, the overall attenuation trend of the transient splash image stream is reconstructed by attenuation curve to obtain the adaptive attenuation factor of the transient splash image stream.

[0097] The process of performing attenuation deduction on the discrete splash particles based on the fragmentation energy distribution to obtain the energy transfer attenuation rate of the discrete splash particles includes:

[0098] Topological sorting of the particle nodes in the fragmented energy distribution yields a directed tree for the energy transfer of the discrete splash particles.

[0099] Energy allocation quantization is performed on adjacent level particles of the energy transfer directed tree to obtain the inter-level energy allocation coefficients of the discrete splash particles;

[0100] The path energy dissipation factor of the discrete splash particles is obtained by parametrically fusing the interstage energy allocation coefficient with the spatial displacement vector of the discrete splash particles.

[0101] Based on the path energy dissipation factor, the energy transfer attenuation rate of the discrete splash particles is calculated, wherein the formula for calculating the energy transfer attenuation rate is:

[0102] ;

[0103] In the formula, The energy transfer attenuation rate is... Let be the index number of the terminal particle in the directed energy transfer tree. The total number of all terminal particles in the energy transfer directed tree. For the energy transfer directed tree, the first The measured kinetic energy of each end particle The initial input energy is used to transfer energy to the root node of the directed tree. The energy transfer directed tree is from the root node to the first node. The index of intermediate nodes along the path of each terminal particle. The energy transfer directed tree is from the root node to the first node. All intermediate nodes along the path of the terminal particle. For the energy transfer directed tree, the first Interstage energy allocation coefficients of intermediate nodes The preset spatial dissipation constant, For the energy transfer directed tree, the first The displacement vector magnitude between an intermediate node and its direct successor node. It is a natural constant.

[0104] In each frame of the transient splash image stream, the individual discrete splash particles are identified one by one based on visual features. The pixel position of each particle in the corresponding frame is marked. The pixel positions of the same particle in different frames are continuously connected according to the time sequence of the image stream. The complete movement path of each discrete splash particle in the image stream is sorted out. The movement paths of all discrete splash particles are integrated and drawn on the same visual coordinate plane. The resulting graphic is the motion trajectory diagram of the discrete splash particles.

[0105] In the motion trajectory diagram of discrete splash particles, the locations where morphological splitting and volume fragmentation occur on the trajectory of each particle are identified as fragmentation nodes. By analyzing the degree of visual morphological change of particles at the fragmentation nodes, the fracture intensity of each fragmentation node is determined hierarchically and numerically. Combining the fragmentation node location, fracture intensity calibration results, and energy correlation characteristics of each discrete splash particle during motion, the energy distribution and transfer state of all discrete splash particles at different fragmentation nodes is sorted out. The overall distribution characteristics formed by integrating these state information are the fragmentation energy distribution of discrete splash particles.

[0106] In the fragmentation energy distribution of discrete splash particles, each discrete splash particle and the sub-particles formed at each fragmentation node are regarded as independent particle nodes. Based on the order of particle splitting and the direction of energy transfer in the fragmentation energy distribution, all particle nodes are arranged in an orderly manner, and the energy transfer direction relationship from the initial particle to each split sub-particle is sorted out. The hierarchical structure constructed according to this direction relationship and the arrangement order is the directed energy transfer tree of discrete splash particles.

[0107] For each level of the directed tree of energy transfer of discrete splash particles, the proportional relationship of energy transfer from the particle node to the next level of sub-particle nodes is analyzed. The specific numerical value of the energy allocation ratio from each intermediate particle node to the subsequent sub-particles is determined. This value is the inter-level energy allocation coefficient of the intermediate node corresponding to the discrete splash particle. The inter-level energy allocation coefficients of all intermediate nodes are integrated to form a complete set of inter-level energy allocation coefficients.

[0108] All values ​​in the set of interstage energy allocation coefficients of discrete splash particles are extracted. At the same time, the spatial displacement vectors formed by the spatial position changes between each particle node in the directed energy transfer tree are sorted out. The interstage energy allocation coefficient of each intermediate node is combined with the spatial displacement vector between the corresponding nodes. The influence of the characteristics of both on energy dissipation is comprehensively considered to form a characteristic index that can reflect the energy dissipation state in each energy transfer path. This index is the path energy dissipation factor of discrete splash particles.

[0109] By combining the path energy dissipation factor of discrete splash particles, the energy dissipation state in all energy transfer paths is sorted out. By combining the energy dissipation of all paths, the degree of energy transfer loss of discrete splash particles in the overall motion and breakup process is specifically determined numerically. This value is the energy transfer attenuation rate of discrete splash particles.

[0110] By integrating the energy transfer attenuation rate of discrete splash particles, the energy transfer attenuation law of all discrete splash particles in the transient splash image stream is analyzed. Based on this law, a curve that reflects the overall energy attenuation trend of the transient splash image stream is plotted, and the attenuation curve of the overall attenuation trend of the transient splash image stream is reconstructed. Based on the reconstructed attenuation curve, parameters that can characterize the core features of the curve are extracted. These parameters are the adaptive attenuation factors of the transient splash image stream.

[0111] The energy transfer attenuation rate is extracted in subsequent steps. The index number of the terminal particles is obtained by sequentially marking all terminal particles in the directed energy transfer tree. The total number of all terminal particles is obtained by counting the number of all terminal particles in the directed energy transfer tree. The measured kinetic energy of the terminal particle was obtained by visually tracking its motion state and combining it with its motion characteristics. The initial input energy of the root node was obtained by analyzing the motion state of the root node particle before breakage. The intermediate node index was obtained by indexing the particles from the root node to the... The path of each terminal particle is obtained by sequentially marking all intermediate nodes along its path, from the root node to the... The complete energy transfer path from the root node to the terminal particle is obtained by sorting out all intermediate nodes along the path of the terminal particle in the directed energy transfer tree. The inter-stage energy allocation coefficient of each intermediate node is obtained by analyzing the energy transfer ratio between the intermediate node and its direct successor node. The preset spatial dissipation constant is a fixed value set in advance. The displacement vector magnitude between an intermediate node and its direct successor node is obtained by measuring the straight-line distance between the intermediate node and the direct successor node, and the natural constant is a fixed known value.

[0112] This calculation method is used to quantify the degree of energy dissipation of discrete splash particles in the directed energy transfer tree from the root node to the terminal particles. By comparing the total measured kinetic energy of all terminal particles with the total energy that should theoretically be transferred to the terminal particles, the loss ratio in the energy transfer process is obtained, thereby reflecting the energy transfer efficiency of the particle breakup process in the transient splash image stream, and providing a quantitative basis for reconstructing the overall attenuation trend of the transient splash image stream.

[0113] When the ratio of the measured total kinetic energy of all end particles to the theoretical total energy transferred decreases, the energy transfer attenuation rate increases, indicating an increase in dissipation during the energy transfer process. Conversely, when the ratio of the measured total kinetic energy of all end particles to the theoretical total energy transferred increases, the energy transfer attenuation rate decreases, indicating a decrease in dissipation during the energy transfer process.

[0114] The beneficial effects include: frame-by-frame identification and trajectory matching of discrete splash particles in transient splash image streams, enabling precise mapping of the complete particle movement path and plotting of motion trajectory diagrams, clearly presenting the overall motion state of the particles. Precise identification of breakage nodes and calibration of fracture intensity allow for a complete understanding of the energy distribution and transfer state of particles during the breakage process; the resulting fragmentation energy distribution accurately reflects the energy change characteristics of the particles. By constructing a directed energy transfer tree through hierarchical analysis, combined with layer-by-layer analysis of energy allocation ratios and spatial displacement characteristics, the dissipation process of particle energy transfer can be precisely deconstructed, and the obtained energy transfer attenuation rate objectively reflects the overall energy loss degree of the particles. Based on the attenuation rate, the reconstructed attenuation curve and extracted adaptive attenuation factor accurately characterize the overall attenuation trend of the transient splash image stream, providing accurate and realistic feature basis for subsequent deformation potential analysis related to leveling machine height adjustment.

[0115] By clearly defining the specific methods for obtaining each parameter, the calculation process of the energy transfer attenuation rate can be ensured to have a clear and practical basis for implementation. The accurate extraction of each parameter guarantees the reliability of the attenuation rate value. This calculation formula can accurately quantify the dissipation degree of discrete splash particles throughout the entire energy transfer process. By comparing the measured kinetic energy with the theoretical transferred energy, it clearly reflects the energy transfer efficiency, providing quantitative technical indicators for subsequent analysis of the attenuation characteristics of transient splash image streams. The energy transfer attenuation rate obtained based on this formula can accurately characterize the trend of energy dissipation changes, accurately supporting the analysis of the overall attenuation trend of transient splash image streams. It provides a scientific and practical quantitative reference for the deformation potential analysis related to the height adjustment of the leveling machine, ensuring the accuracy and effectiveness of subsequent height adjustment decisions.

[0116] S03. Based on the adaptive attenuation factor, perform deformation potential analysis on the current action point of the target working equipment to obtain the estimated deformation amount of the current action point.

[0117] In this embodiment of the invention, the step of analyzing the deformation potential of the current point of action of the target working equipment based on the adaptive attenuation factor to obtain the estimated deformation of the current point of action includes:

[0118] Spatial domain mapping is performed on the adaptive attenuation factor to obtain the bearing capacity influence region of the adaptive attenuation factor;

[0119] Based on the bearing capacity influence area, the material density at the current point of action in the target operating equipment is evaluated for compression response to obtain the compression stiffness at the current point of action.

[0120] Based on the compressive stiffness, load response tracking is performed on the weak area at the current point of application to obtain the plastic flow triggering threshold at the current point of application.

[0121] Based on the plastic flow triggering threshold, the settlement trend of the current point of action is converged to obtain the estimated deformation of the current point of action.

[0122] By establishing a correspondence between the characteristics of the adaptive attenuation factor and the spatial coordinates of the work object, the attenuation characteristics reflected by the adaptive attenuation factor are projected onto the actual work space of the work object, and the specific spatial range in which the factor can affect the load-bearing capacity of the work object is delineated. This range accurately covers the current point of action of the target work equipment and the surrounding related areas. Through this spatial projection method, the spatial domain mapping of the adaptive attenuation factor is completed, and the final delineated specific spatial range is the load-bearing capacity influence area of ​​the adaptive attenuation factor.

[0123] Using the bearing capacity influence area of ​​the adaptive attenuation factor as a fixed analysis range, layered contact detection is carried out on the material at the current action point of the target operating equipment. The morphological changes of different material layers at this point under constant pressure are recorded. Based on the specific manifestations of the morphological changes of each material layer, the material density of each layer is determined and a comprehensive evaluation is performed to obtain the overall characteristics of the material density at the current action point. Then, combined with the overall characteristics of the material density and the deformation response law of the material under compression, the ability of the material at the current action point to resist compression deformation is specifically quantitatively characterized. The result of this quantitative characterization is the compressive stiffness of the current action point.

[0124] Based on the material's resistance to deformation characterized by the compressive stiffness at the current application point, the specific area with relatively weak resistance to deformation at that point is identified as the weak area at the current application point. Gradual, progressively increasing loads are applied to this weak area, and the deformation state of the weak area under different load values ​​is continuously recorded. The entire transformation process from elastic deformation to plastic flow in the weak area is fully tracked, and the specific load value that can directly trigger plastic flow in the weak area is determined. This specific load value is the plastic flow trigger threshold at the current application point.

[0125] Using the plastic flow trigger threshold at the current point of application as the core criterion, the settlement development trend at the current point of application under the operating load of the target equipment is analyzed. Combining the continuous action characteristics of the operating load and the material characteristics at the current point of application, the change trend of the settlement at this point with the operating load time is sorted out. By gradually summarizing and determining the final change characteristics of the settlement trend, the specific value of the final settlement deformation at this point under the continuous action of the operating load is determined. This specific value is the estimated deformation at the current point of application.

[0126] The beneficial effects are that by mapping the adaptive attenuation factor to the spatial domain, the attenuation characteristics can be accurately correlated with the actual working space of the work object. The defined load-bearing capacity influence area provides a precise and fixed range for subsequent material analysis, making the subsequent analysis work more targeted. Layered contact testing based on this area can comprehensively grasp the material's density characteristics. The resulting compressive stiffness can accurately quantify the material's ability to resist compressive deformation, providing a reliable basis for identifying weak areas. Applying gradient loads to weak areas and tracking the load response can accurately determine the plastic flow trigger threshold and clarify the key load values ​​for deformation transformation. Combining this threshold with settlement trend analysis and convergence, the obtained estimated deformation can accurately quantify the deformation values ​​at the work points, providing a precise deformation basis that fits the actual working conditions for the deviation mapping of subsequent height adjustments of the leveling machine.

[0127] S04. Discretely compare the estimated deformation with the reference elevation of the work object, and based on the comparison results, perform deviation mapping on the subsequent working height of the target work equipment to obtain the feedforward compensation coefficient of the target work equipment.

[0128] In this embodiment of the invention, the step of discretizing the estimated deformation and the reference elevation of the work object, and based on the comparison result, performing deviation mapping on the subsequent working height of the target work equipment to obtain the feedforward compensation coefficient of the target work equipment includes:

[0129] Based on the distribution points of the estimated deformation, the spatial coordinates of the reference surface elevation of the work object are indexed point by point to obtain the pairing data of the estimated deformation and the reference surface elevation at the same point.

[0130] The elevation values ​​of the paired data pairs at the same location are compared numerically to obtain the original set of deviation values ​​for the paired data pairs at the same location.

[0131] The positive deviation values ​​of the original deviation set are marked as raised deviations, and the negative deviation values ​​of the original deviation set are marked as concave deviations.

[0132] By performing partitioned topological mapping on the bulging deviation and the concave deviation, a polarization deviation distribution map of the estimated deformation is obtained.

[0133] Based on the polarization deviation distribution map, the subsequent working height of the target working equipment is inversely compensated to obtain the original compensation height sequence of the target working equipment;

[0134] The original high-compensation program sequence is subjected to adjacent-point amplitude limiting smoothing to obtain the smoothed high-compensation program sequence of the original high-compensation program.

[0135] The feedforward compensation coefficients of the target operating equipment are obtained by normalizing the dimensions of the smooth compensation high-order sequence.

[0136] The step of performing partitioned topological mapping on the raised deviation and the recessed deviation to obtain the polarization deviation distribution map of the estimated deformation includes:

[0137] Clustering and hierarchical merging are performed on the raised particles of the raised deviation to obtain the topological skeleton of the main ridge of the raised deviation; the directional trend fitting is performed on the concave particles of the concave deviation to obtain the topological skeleton of the valley bottom of the concave deviation; based on the topological skeleton of the main ridge and the valley bottom, the boundaries of the spheres of influence of the raised deviation and the concave deviation are divided into equipotential regions to obtain the Voronoi partition map of the raised deviation and the concave deviation.

[0138] Calculate the polarization intensity of spatial points in the Volonuova partition map, wherein the formula for calculating the polarization intensity is:

[0139] ;

[0140] In the formula, Spatial point polarization intensity, For the spatial point, The sign function of the Volonouva partition map. It is a natural constant. The preset nonlinear attenuation coefficient, The maximum value of the average distance from the spatial point to all deviation particles within the corresponding partition in the Volonouva partition map. The average Euclidean distance from the spatial point to all deviation particles within the partition to which it belongs in the Volonouva partition map;

[0141] Based on the polarization intensity, the Volonuowa partition map is filled point by point to obtain the polarization deviation distribution map of the estimated deformation.

[0142] The spatial coordinate information of all distribution points of the estimated deformation within the space of the work object is sorted out. This coordinate information is used as the basis for retrieval. In the spatial coordinate system of the reference surface elevation of the work object, the reference surface elevation coordinates corresponding to each distribution point are matched and retrieved one by one. The reference surface elevation value corresponding to each distribution point of the estimated deformation is extracted. The estimated deformation value of each point and the corresponding retrieved reference surface elevation value are matched one by one. Each set of corresponding data is the pair of data of the same point of the estimated deformation and the reference surface elevation.

[0143] For each pair of data at the same location, the estimated deformation value and the reference elevation value are calculated to make a difference. The difference result for each data pair is recorded. After all the pairs of data at the same location have completed the difference calculation, all the difference results are integrated and collected according to the corresponding spatial point order. The resulting set of difference results is the original deviation set of the pairs of data at the same location.

[0144] For each deviation value in the original deviation set, determine the sign of the value. Identify deviation values ​​with positive signs and add a unique raised deviation label to these values. Identify deviation values ​​with negative signs and add a unique recessed deviation label to these values. This process ensures that all deviation values ​​in the original deviation set are labeled with their corresponding types.

[0145] All points marked as raised deviations are extracted as raised particles. The spatial correlation and clustering density between raised particles are analyzed. The raised particles are hierarchically divided and merged according to different densities. The line structure that can reflect the overall spatial distribution and core area of ​​the raised deviation is sorted out. This structure is the topological skeleton of the raised main ridge of the raised deviation.

[0146] All points marked as concave deviations are extracted as concave particles. The spatial coordinate distribution characteristics of each concave particle are analyzed. Based on these characteristics, the overall spatial extension direction of the concave particles is fitted with a trend, and a line structure that can reflect the overall spatial distribution direction and core area of ​​the concave deviation is constructed. This structure is the topological skeleton of the concave valley bottom line of the concave deviation.

[0147] Using the topological framework of the main ridge and the bottom line of the valley as the core reference, the spatial influence range of the ridge-type deviation and the depression-type deviation is defined in the spatial coordinate system of the work object. According to the principle of equipotentiality, the spatial region boundaries of the two types of deviations are divided. The divided spatial regions are correlated with the deviation types and drawn into a diagram. This diagram is the Voronoi zoning map of the ridge-type deviation and the depression-type deviation.

[0148] Within each spatial region defined by the Volonovo partition map, spatial points within the region are selected one by one. The spatial positional relationship between each spatial point and all deviation particles within its region is analyzed. Combining this positional relationship with the characteristics of the deviation type, the characteristic values ​​of each spatial point are calibrated. These calibrated characteristic values ​​are the polarization intensity of the corresponding spatial point in the Volonovo partition map.

[0149] The polarization intensity of each spatial point in the Volonovo partition map is converted into a corresponding visual filling feature. All spatial points in the Volonovo partition map are then visually filled point by point according to this visual filling feature. The complete graphic formed after filling is the polarization deviation distribution map of the estimated deformation.

[0150] Based on the deviation type and polarization intensity characteristics of each spatial point in the polarization deviation distribution map, a corresponding working height compensation scheme is formulated for each spatial point. The subsequent working height compensation value for each point is determined according to the compensation scheme. The compensation values ​​of all spatial points are integrated and arranged according to their spatial distribution order. The resulting numerical sequence is the original compensation height sequence of the target working equipment.

[0151] The compensation values ​​of adjacent spatial points in the original compensation height program sequence are sorted out, and a fixed limit for the range of numerical variation is set. The range of variation of compensation values ​​between adjacent points is checked. If the range of variation exceeds the limit, the compensation value is adjusted and corrected according to the limit. If it does not exceed the limit, the original value remains unchanged. After this process is completed for all adjacent points in the original compensation height program sequence, the corrected value sequence is the smooth compensation height program sequence of the original compensation height program.

[0152] Analyze the dimensional characteristics of all compensation values ​​in the smooth compensation high-order sequence, determine a unified dimensional standard, transform and adjust each compensation value in the sequence according to the unified dimensional standard to eliminate the numerical differences caused by different dimensions, integrate all the transformed and adjusted values, and the normalized value that can characterize the overall compensation characteristics is the feedforward compensation coefficient of the target operating equipment.

[0153] The spatial point is obtained by selecting specific spatial coordinates in the Voronoi zoning map. The sign function is obtained by determining the deviation type of the zoning to which the spatial point belongs. A raised deviation corresponds to a positive sign characteristic, and a concave deviation corresponds to a negative sign characteristic. The natural constant is a fixed known value. The nonlinear attenuation coefficient is a pre-set fixed value. The average Euclidean distance from the spatial point to all deviation particles in its zoning is obtained by averaging the straight-line distances between the spatial point and each deviation particle in the zoning. The maximum value of the average distance from the spatial point to all deviation particles in its zoning is obtained by statistically analyzing the average Euclidean distances of all spatial points in the zoning and taking the maximum value among them. The polarization intensity is the final value characterizing the degree of polarization of the spatial point deviation.

[0154] This calculation method is used to quantify the degree of deviation polarization of each spatial point in the Voronois partition map. By combining the sign characteristics of the deviation type and the average distance characteristics from the spatial point to the deviation particle, it reflects the strength and direction of the influence of the deviation of the partition to which the point belongs. This provides a precise quantitative basis for the point-by-point visual filling of the polarization deviation distribution map, making the spatial distribution characteristics of the deviation of the estimated deformation more clearly identifiable.

[0155] When the average Euclidean distance from a spatial point to the corresponding deviation particle decreases, the absolute value of the polarization intensity increases, indicating that the point is more strongly affected by the deviation of its corresponding zone. When the average Euclidean distance from a spatial point to the corresponding deviation particle increases, the absolute value of the polarization intensity decreases, indicating that the point is less affected by the deviation of its corresponding zone. The sign of the polarization intensity determines the direction of the polarization intensity, corresponding to the type of deviation, either raised or concave.

[0156] The beneficial effects include: precise matching of estimated deformation with datum elevation through point-by-point indexing of spatial coordinates, ensuring that deviation comparisons are based on the same points and guaranteeing the accuracy of the original deviation set; clear distinction between bulge and depression deviations by labeling deviation values, providing direction for subsequent deviation analysis; precise extraction of core deviation distribution characteristics through a topological framework constructed by hierarchical merging and trend fitting; clear definition of the spatial range of the two types of deviations by the equipotential partitioning of the partition map; and a polarization deviation distribution map obtained by point-by-point filling of polarization intensity, which intuitively and comprehensively presents the spatial distribution and characteristics of the deviations. The compensation sequence generated by reverse compensation, after amplitude limiting and smoothing, makes the compensation values ​​more stable and reasonable. The feedforward compensation coefficient obtained by dimension normalization can uniformly represent the overall compensation requirements, providing a precise and suitable quantitative basis for subsequent height adjustments of the leveling machine.

[0157] By clearly defining the specific methods for obtaining each parameter, the calculation process of polarization intensity can be ensured to have a clear and practical basis. The accurate extraction of each parameter guarantees the reliability of the polarization intensity values, providing a stable quantitative foundation for subsequent deviation analysis. This calculation method can accurately quantify the degree of deviation polarization at each spatial point in the Volonoa zoning map. Combining the sign characteristics of the deviation type with the distance characteristics from the spatial point to the deviation particle, it clearly reflects the strength and direction of the influence of the deviation on the point's region. This provides a precise quantitative basis for the point-by-point visual filling of the polarization deviation distribution map, making the spatial distribution characteristics of the predicted deformation deviation more clearly identifiable. Based on the polarization intensity change trend obtained from this calculation, the degree of influence of deviation on spatial points can be accurately reflected, making the spatial gradient change of the deviation influence more intuitive. This provides a more quantitative reference that closely matches the actual deviation distribution for subsequent reverse compensation of operational height, ensuring the pertinence and effectiveness of the compensation scheme.

[0158] S05. Data coupling encoding is performed on the feedforward compensation coefficient and the current execution elevation of the target working equipment to obtain the drive command sequence of the target working equipment;

[0159] In this embodiment of the invention, the step of data coupling encoding of the feedforward compensation coefficient and the current execution elevation of the target operating equipment to obtain the drive instruction sequence of the target operating equipment includes:

[0160] The feedforward compensation coefficients are non-uniformly divided to obtain the variable periodic compensation amount of the feedforward compensation coefficients;

[0161] Displacement mapping is performed on the current execution elevation of the target operating equipment to obtain the desired pose set of the current execution elevation;

[0162] Based on the variable period compensation amount, the desired pose set is superimposed and corrected periodically to obtain the temporal fusion desired pose of the feedforward compensation coefficient and the current execution elevation.

[0163] The joint space trajectory of the target working device is obtained by inverse kinematic reconstruction of the temporal fusion desired pose.

[0164] Based on the joint space trajectory, pulse width encoding is performed on the target working equipment to obtain the drive command sequence of the target working equipment.

[0165] Based on the operational cycle characteristics of the target equipment and the differences in height compensation requirements at different operational stages of the leveling machine, the feedforward compensation coefficient is segmented and decomposed. Matching compensation values ​​are set for different operational cycle segments. The time span and compensation value of each cycle segment are determined according to the actual height adjustment requirements. The compensation values ​​of each cycle segment after division can fully reflect the compensation characteristics of the original feedforward compensation coefficient. The set of compensation values ​​corresponding to each cycle segment obtained after decomposition is the variable cycle compensation amount of the feedforward compensation coefficient.

[0166] The spatial position information of the actuator corresponding to the current operating elevation of the target equipment is sorted out. Combined with the position and posture specifications of the leveling machine operation, the spatial position information of the current operating elevation is transformed into the ideal spatial position and posture information of the actuator corresponding to each preset operation node in the work space. Each preset operation node corresponds to a unique ideal spatial position and posture. The ideal spatial positions and postures of all preset operation nodes are integrated according to the order of operation. The resulting set of position and posture information is the expected position and posture set of the current operating elevation.

[0167] The compensation values ​​of each period segment in the variable period compensation quantity are matched one by one with the preset operation nodes in the desired pose set according to the operation time period. Based on the ideal spatial pose of each preset operation node, the compensation values ​​of the corresponding period segment are superimposed on the elevation dimension of the node's pose to complete the elevation dimension correction and adjustment of the ideal spatial pose of the node. After completing the pose correction of all preset operation nodes in sequence, all the corrected spatial poses are connected continuously according to the operation time sequence. The complete pose sequence formed is the temporal fusion of the feedforward compensation coefficient and the current execution elevation of the desired pose.

[0168] The end effector pose requirements of the device actuator corresponding to each corrected spatial pose in the expected pose of the time-series fusion are analyzed. Combining the connection relationship and range of motion of each joint of the target working device, the spatial position that each joint of the device needs to reach under each end effector pose is derived. The spatial positions of each joint corresponding to all end effector poses are sorted out according to the working time sequence. The change path of the spatial position of each joint with the working time sequence is continuously constructed. The continuous position change path of each joint is the joint spatial trajectory of the target working device.

[0169] The position change path and speed requirements of each joint in the joint space trajectory are analyzed. Each position change of each joint is converted into the pulse width and duration of the corresponding drive signal. Different pulse widths and durations correspond to different motion amplitudes and movement distances of the equipment drive components. The pulse widths and durations of the drive signals corresponding to each joint are ordered according to the operation sequence. The corresponding drive signal trigger command is matched for each pulse width and duration. All trigger commands are integrated and arranged according to the order of joint movement and coordination relationship. The resulting command set is the drive command sequence of the target operation equipment.

[0170] The beneficial effects include: non-uniformly dividing the feedforward compensation coefficient by combining the equipment's operating cycle and compensation requirements; the resulting variable-cycle compensation amount can accurately match the height adjustment needs of the leveling machine at different operating stages, making the compensation more targeted; obtaining the desired pose set by displacement mapping of the current execution elevation clarifies the ideal pose of each operating node of the equipment, providing a clear benchmark for subsequent pose correction; the temporal fusion of the desired pose formed by periodic superposition correction achieves accurate temporal fusion of the feedforward compensation coefficient and the current execution elevation, ensuring that the pose requirements match the actual compensation needs; the joint space trajectory obtained by reverse reconstruction transforms the end-effector pose requirements into continuous motion paths of each moving joint, completing the accurate conversion from pose to joint motion; and the drive command sequence formed by pulse width encoding based on the trajectory transforms the joint motion path into drive commands that the equipment can directly execute, achieving efficient coupling between compensation data and equipment drive, providing a highly adaptable execution basis for the precise control of the leveling machine's height.

[0171] S06. Based on the drive command sequence, the target working equipment is precisely controlled to obtain the real-time adjustment height of the target working equipment;

[0172] In this embodiment of the invention, the precise control of the target working equipment based on the drive command sequence to obtain the real-time adjustment height of the target working equipment includes:

[0173] The drive instruction sequence is segmented into time-axis segments to obtain single-cycle action primitives of the drive instruction sequence;

[0174] Based on the single-cycle action primitive, the target working equipment is driven and triggered to obtain the instantaneous output torque parameter of the target working equipment;

[0175] Based on the instantaneous output torque parameter, the actuator of the target working equipment is displaced and tractioned to obtain the real-time spatial pose of the actuator;

[0176] The vertical displacement of the real-time spatial pose is measured to obtain the real-time adjustment height of the target working equipment.

[0177] The timing characteristics of the drive instruction sequence are analyzed. Using the single operation cycle of the target work equipment as a fixed division basis, the range of the drive instruction sequence is defined segment by segment on the timing axis. The time interval and specific action content corresponding to each segment are clarified. After the segment is defined, the independent instruction set corresponding to a single work cycle is completely extracted. The extracted independent instruction set is the single-cycle action primitive of the drive instruction sequence.

[0178] The instruction information in the single-cycle action primitive is converted into a recognizable trigger signal for the drive component of the target working equipment. According to the order of instructions in the primitive and the action execution requirements, the trigger signal is sent to the drive component one by one, triggering the drive component to output torque according to the preset action rhythm. The real-time value of the output torque of the drive component is continuously and accurately collected by a dedicated torque detection component, and the specific torque value corresponding to each trigger moment is recorded. The set of torque values ​​arranged in sequence according to the trigger moment is the instantaneous output torque parameter of the target working equipment.

[0179] Each torque value in the instantaneous output torque parameter is matched one by one with the displacement traction rules of the target working equipment actuator. The drive component is controlled to output the corresponding torque according to the matched traction requirements. Through the transmission of torque, the actuator's various moving joints complete the corresponding displacement movement. The spatial position and attitude state of the actuator are captured and recorded in real time during the entire movement process using a spatial pose detection component. The spatial position coordinates and attitude characteristics of the actuator at each moment of movement are stored. The overall set of these spatial positions and attitude characteristics arranged in order according to the moment of movement is the real-time spatial pose of the actuator.

[0180] From all the recorded information of real-time spatial pose, the vertical displacement coordinate information of the actuator is accurately extracted. The change process of the vertical displacement coordinate as the actuator moves is completely sorted out in chronological order. The vertical displacement coordinate value corresponding to the actuator when it finally reaches a stable state after the target working equipment completes the height adjustment action is determined. The final vertical displacement coordinate value is the real-time adjustment height of the target working equipment.

[0181] The beneficial effects include segmenting the drive command sequence according to the time axis of a single work cycle, resulting in single-cycle action primitives that can be broken down into commands and clarify the action requirements of each cycle, making drive execution more phased and targeted, and laying a solid foundation for precise drive triggering. Converting primitives into trigger signals and acquiring torque allows the instantaneous output torque parameters to accurately match the torque requirements of the equipment's actions, ensuring a high degree of fit between the output torque of the drive components and the work actions. By using torque to pull the actuator and capturing and recording its pose, the obtained real-time spatial pose can completely present the pose changes of the actuator during its movement, providing comprehensive and accurate pose information for height measurement. Extracting the vertical displacement from the real-time spatial pose and determining the stable value, the obtained real-time adjusted height can accurately reflect the final result of the equipment's height adjustment, achieving precise real-time control of the leveler's height, ensuring that height adjustment perfectly matches the work compensation requirements, and improving the execution accuracy and actual effect of the leveler's height adjustment.

[0182] like Figure 2 The diagram shown is a functional block diagram of a real-time height adjustment system for a leveling machine based on industrial vision, provided in an embodiment of the present invention.

[0183] The real-time height adjustment system 10 for a leveling machine based on industrial vision described in this invention can be installed in an electronic device. Depending on the functions implemented, the real-time height adjustment system 10 for a leveling machine based on industrial vision may include a coaxial tracking module 11, a dynamic trend inversion module 12, a deformation potential analysis module 13, a deviation mapping compensation module 14, a data coupling encoding module 15, and a precision control module 16. The module described in this invention can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and are stored in the memory of the electronic device.

[0184] In this embodiment, the functions of each module / unit are as follows:

[0185] The coaxial tracking module 11 is used to perform coaxial tracking of the target working equipment and the working object to obtain the transient splash image stream of the working object;

[0186] The dynamic trend inversion module 12 is used to perform dynamic trend inversion on the transient splash image stream to obtain the adaptive attenuation factor of the transient splash image stream.

[0187] The deformation potential analysis module 13 is used to analyze the deformation potential of the current action point of the target working equipment based on the adaptive attenuation factor, and obtain the estimated deformation of the current action point.

[0188] The deviation mapping compensation module 14 is used to perform discrete comparison between the estimated deformation and the reference surface elevation of the work object, and based on the comparison result, to perform deviation mapping on the subsequent working height of the target work equipment to obtain the feedforward compensation coefficient of the target work equipment.

[0189] The data coupling encoding module 15 is used to perform data coupling encoding on the feedforward compensation coefficient and the current execution elevation of the target working equipment to obtain the drive instruction sequence of the target working equipment.

[0190] The precision control module 16 is used to precisely control the target working equipment based on the drive command sequence, so as to obtain the real-time adjustment height of the target working equipment.

[0191] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0192] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0193] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0194] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0195] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0196] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for real-time height adjustment of a leveling machine based on industrial vision, characterized in that, The method includes: S01. Perform coaxial tracking on the target work equipment and the work object to obtain the transient splash image stream of the work object; S02. Perform dynamic trend inversion on the transient splash image stream to obtain the adaptive attenuation factor of the transient splash image stream; S03. Based on the adaptive attenuation factor, perform deformation potential analysis on the current action point of the target working equipment to obtain the estimated deformation amount of the current action point. S04. Discretely compare the estimated deformation with the reference elevation of the work object, and based on the comparison results, perform deviation mapping on the subsequent working height of the target work equipment to obtain the feedforward compensation coefficient of the target work equipment. S05. Data coupling encoding is performed on the feedforward compensation coefficient and the current execution elevation of the target working equipment to obtain the drive command sequence of the target working equipment; S06. Based on the drive command sequence, the target working equipment is precisely controlled to obtain the real-time adjustment height of the target working equipment.

2. The method for real-time height adjustment of a leveling machine based on industrial vision as described in claim 1, characterized in that, The method of coaxially tracking the target equipment and the work object to obtain the transient splash image stream of the work object includes: Visual focus is locked on the contact area between the front end of the actuator of the target working equipment and the working object to obtain a high-definition focused image of the contact area; Based on the high-definition focused image, impact phase analysis is performed on the target operating equipment to obtain the timing action parameters of the target operating equipment. Based on the timing action parameters, the high-definition focused image is synchronously acquired and triggered to obtain the transient change image sequence of the contact field; Dynamic deformation tracking is performed on the splash area of ​​the transient image sequence to obtain the instantaneous displacement field of the splash area; Based on the instantaneous displacement field, the transient change image sequence is corrected point by point to obtain the transient splash image stream of the working object.

3. The method for real-time height adjustment of a leveling machine based on industrial vision as described in claim 1, characterized in that, The step of performing dynamic trend inversion on the transient splash image stream to obtain the adaptive attenuation factor of the transient splash image stream includes: Trajectory tracking is performed on discrete splash particles in the transient splash image stream to obtain the motion trajectory map of the discrete splash particles; Based on the motion trajectory diagram, the fracture intensity of the discrete splash particles is analyzed at the breakup nodes to obtain the fragmentation energy distribution of the discrete splash particles; Based on the fragmentation energy distribution, the attenuation of the discrete splash particles is extrapolated to obtain the energy transfer attenuation rate of the discrete splash particles; Based on the energy transfer attenuation rate, the overall attenuation trend of the transient splash image stream is reconstructed by attenuation curve to obtain the adaptive attenuation factor of the transient splash image stream.

4. The method for real-time height adjustment of a leveling machine based on industrial vision as described in claim 3, characterized in that, The attenuation deduction of the discrete splash particles based on the fragmentation energy distribution to obtain the energy transfer attenuation rate of the discrete splash particles includes: Topological sorting of the particle nodes in the fragmented energy distribution yields a directed tree for the energy transfer of the discrete splash particles. Energy allocation quantization is performed on adjacent level particles of the energy transfer directed tree to obtain the inter-level energy allocation coefficients of the discrete splash particles; The path energy dissipation factor of the discrete splash particles is obtained by parametrically fusing the interstage energy allocation coefficient with the spatial displacement vector of the discrete splash particles. Based on the path energy dissipation factor, the energy transfer attenuation rate of the discrete splash particles is calculated, wherein the formula for calculating the energy transfer attenuation rate is: ； In the formula, The energy transfer attenuation rate is... Let be the index number of the terminal particle in the directed energy transfer tree. The total number of all terminal particles in the energy transfer directed tree. For the energy transfer directed tree, the first The measured kinetic energy of each end particle The initial input energy is used to transfer energy to the root node of the directed tree. The energy transfer directed tree is from the root node to the first node. The index of intermediate nodes along the path of each terminal particle. The energy transfer directed tree is from the root node to the first node. All intermediate nodes along the path of the terminal particle. For the energy transfer directed tree, the first Interstage energy allocation coefficients of intermediate nodes The preset spatial dissipation constant, For the energy transfer directed tree, the first The displacement vector magnitude between an intermediate node and its direct successor node. It is a natural constant.

5. The method for real-time height adjustment of a leveling machine based on industrial vision as described in claim 1, characterized in that, The step of analyzing the deformation potential of the current point of action of the target working equipment based on the adaptive attenuation factor to obtain the estimated deformation of the current point of action includes: Spatial domain mapping is performed on the adaptive attenuation factor to obtain the bearing capacity influence region of the adaptive attenuation factor; Based on the bearing capacity influence area, the material density at the current point of action in the target operating equipment is evaluated for compression response to obtain the compression stiffness at the current point of action. Based on the compressive stiffness, load response tracking is performed on the weak area at the current point of application to obtain the plastic flow triggering threshold at the current point of application. Based on the plastic flow triggering threshold, the settlement trend of the current point of action is converged to obtain the estimated deformation of the current point of action.

6. The method for real-time height adjustment of a leveling machine based on industrial vision as described in claim 1, characterized in that, The step of discretizing and comparing the estimated deformation with the reference elevation of the work object, and based on the comparison results, mapping the subsequent working height of the target work equipment to obtain the feedforward compensation coefficient of the target work equipment, includes: Based on the distribution points of the estimated deformation, the spatial coordinates of the reference surface elevation of the work object are indexed point by point to obtain the pairing data of the estimated deformation and the reference surface elevation at the same point. The elevation values ​​of the paired data pairs at the same location are compared numerically to obtain the original set of deviation values ​​for the paired data pairs at the same location. The positive deviation values ​​of the original deviation set are marked as raised deviations, and the negative deviation values ​​of the original deviation set are marked as concave deviations. By performing partitioned topological mapping on the bulging deviation and the concave deviation, a polarization deviation distribution map of the estimated deformation is obtained. Based on the polarization deviation distribution map, the subsequent working height of the target working equipment is inversely compensated to obtain the original compensation height sequence of the target working equipment; The original high-compensation program sequence is subjected to adjacent-point amplitude limiting smoothing to obtain the smoothed high-compensation program sequence of the original high-compensation program. The feedforward compensation coefficients of the target operating equipment are obtained by normalizing the dimensions of the smooth compensation high-order sequence.

7. The method for real-time height adjustment of a leveling machine based on industrial vision as described in claim 6, characterized in that, The step of performing partitioned topological mapping on the raised deviation and the recessed deviation to obtain the polarization deviation distribution map of the estimated deformation includes: Clustering and hierarchical merging of the raised particles of the raised deviation are performed to obtain the topological skeleton of the main ridge of the raised deviation. By fitting the directional trend of the concave particles of the concave deviation, the topological skeleton of the concave valley bottom line of the concave deviation is obtained. Based on the topological framework of the main ridge of the uplift and the topological framework of the valley bottom, the boundaries of the spheres of influence of the uplift-type deviation and the depression-type deviation are divided into equipotential regions to obtain the Voronoi partition map of the uplift-type deviation and the depression-type deviation. Calculate the polarization intensity of spatial points in the Volonuova partition map, wherein the formula for calculating the polarization intensity is: ； In the formula, Spatial point polarization intensity, For the spatial point, The sign function of the Volonouva partition map. It is a natural constant. The preset nonlinear attenuation coefficient, The maximum value of the average distance from the spatial point to all deviation particles within the corresponding partition in the Volonouva partition map. The average Euclidean distance from the spatial point to all deviation particles within the partition to which it belongs in the Volonouva partition map; Based on the polarization intensity, the Volonuowa partition map is filled point by point to obtain the polarization deviation distribution map of the estimated deformation.

8. The method for real-time height adjustment of a leveling machine based on industrial vision as described in claim 1, characterized in that, The step of data coupling encoding the feedforward compensation coefficient with the current execution elevation of the target operating equipment to obtain the drive command sequence of the target operating equipment includes: The feedforward compensation coefficients are non-uniformly divided to obtain the variable periodic compensation amount of the feedforward compensation coefficients; Displacement mapping is performed on the current execution elevation of the target operating equipment to obtain the desired pose set of the current execution elevation; Based on the variable period compensation amount, the desired pose set is superimposed and corrected periodically to obtain the temporal fusion desired pose of the feedforward compensation coefficient and the current execution elevation. The joint space trajectory of the target working device is obtained by inverse kinematic reconstruction of the temporal fusion desired pose. Based on the joint space trajectory, pulse width encoding is performed on the target working equipment to obtain the drive command sequence of the target working equipment.

9. The method for real-time height adjustment of a leveling machine based on industrial vision as described in claim 1, characterized in that, The precise control of the target working equipment based on the drive command sequence to obtain the real-time adjustment height of the target working equipment includes: The drive instruction sequence is segmented into time-axis segments to obtain single-cycle action primitives of the drive instruction sequence; Based on the single-cycle action primitive, the target working equipment is driven and triggered to obtain the instantaneous output torque parameter of the target working equipment; Based on the instantaneous output torque parameter, the actuator of the target working equipment is displaced and tractioned to obtain the real-time spatial pose of the actuator; The vertical displacement of the real-time spatial pose is measured to obtain the real-time adjustment height of the target working equipment.

10. A real-time height adjustment system for a leveling machine based on industrial vision, characterized in that, The system for implementing the real-time height adjustment method for a leveling machine based on industrial vision as described in claim 1 includes: A coaxial tracking module is used to perform coaxial tracking of the target working equipment and the working object to obtain the transient splash image stream of the working object; The dynamic trend inversion module is used to perform dynamic trend inversion on the transient splash image stream to obtain the adaptive attenuation factor of the transient splash image stream. The deformation potential analysis module is used to analyze the deformation potential of the current action point of the target working equipment based on the adaptive attenuation factor, and obtain the estimated deformation of the current action point. The deviation mapping compensation module is used to perform discrete comparison between the estimated deformation and the reference surface elevation of the work object, and based on the comparison result, to perform deviation mapping on the subsequent working height of the target work equipment to obtain the feedforward compensation coefficient of the target work equipment. The data coupling encoding module is used to perform data coupling encoding on the feedforward compensation coefficient and the current execution elevation of the target working equipment to obtain the drive instruction sequence of the target working equipment; A precision control module is used to precisely control the target working equipment based on the drive command sequence, so as to obtain the real-time adjustment height of the target working equipment.

Images