Search

CN-121989450-A - 3D printing platform leveling method and system based on multipoint tactile feedback

CN121989450ACN 121989450 ACN121989450 ACN 121989450ACN-121989450-A

Abstract

The invention discloses a 3D printing platform leveling method and system based on multipoint tactile feedback, which realize fine scanning and deformation analysis of a printing platform by initializing a tactile sensing network and establishing a coordinate calibration reference, and ensure the accuracy of a coordinate system by detecting each sensing unit to output an electric signal by using a distributed piezoresistive sensing grid, shielding abnormal nodes, determining a coordinate origin, setting a background reference value and the like. And then, performing first-round full-coverage low-speed scanning based on the calibrated coordinate reference, traversing the divided grid points, recording readings of the height encoders of the points and resistance change data, and generating a preliminary deformation distribution map. The method not only improves the automation degree of leveling work and reduces human intervention, but also effectively solves the problem of local unevenness by carrying out comprehensive and fine scanning analysis on the whole printing surface.

Inventors

  • TAN YONGLIANG
  • Hong Yingsheng
  • HE GUIHUA
  • WU JIE
  • ZHOU XUEFENG

Assignees

  • 深圳市智能派科技有限公司

Dates

Publication Date
20260508
Application Date
20260312

Claims (10)

  1. 1. A 3D printing platform leveling method based on multi-point tactile feedback, comprising: initializing a touch sensing network, establishing a coordinate calibration standard, obtaining an idle electric signal as a background standard value, performing first-round full-coverage low-speed scanning based on the calibrated coordinate standard, traversing divided grid points, recording readings of a height encoder of each point and resistance change data, and generating a preliminary deformation distribution map; Identifying an abnormal response area according to the preliminary deformation distribution diagram, planning an enhanced detection path, executing local scanning in the abnormal response area, fusing first-round full-coverage scanning data and local scanning data, and constructing a comprehensive surface topology model by combining a height field and a resistance response value; dividing functional areas based on the comprehensive surface topology model, quantifying compensation requirements of all areas, generating a three-dimensional adjustment guide map, mapping the three-dimensional adjustment guide map to adjustable supporting nodes, calculating adjustment quantity of all nodes, and generating a driving sequence; And carrying out support structure adjustment in stages according to the driving sequence, monitoring area response change after each stage of adjustment, dynamically correcting subsequent adjustment quantity, executing final state verification scanning according to the finally adjusted surface state, generating a leveling report based on a verification result, and judging whether leveling quality meets the standard or not.
  2. 2. The method for leveling a 3D printing platform based on multi-point haptic feedback according to claim 1, wherein initializing a haptic sensation network and establishing a coordinate calibration reference comprises: Starting a distributed piezoresistive sensing grid, detecting nodes with the shielding response value lower than the lower limit of the threshold value or exceeding the upper limit of the threshold value, wherein the nodes are used for outputting electric signals by all sensing units; based on the induction grid of the shielded abnormal node, driving the three-axis linkage device to trigger each axis limit switch, repeatedly approaching the limiter three times, taking an average value, and determining a coordinate origin; planning a standard test point array by taking a coordinate origin as a center, moving a printing head to a preset coordinate, and collecting data after judging that the fluctuation of the acceleration sensor reading is smaller than a set tolerance within five continuous seconds; and setting a background reference value based on the acquired no-load electric signals for calculating the net change amount in the first-round full-coverage low-speed scanning.
  3. 3. The 3D printing platform leveling method based on multi-point tactile feedback according to claim 1, wherein the performing the first-round full-coverage low-speed scan comprises: Dividing a printing platform into M multiplied by N equidistant sampling units, and setting the distance between adjacent points as the smallest dimension of the platform divided by a resolution coefficient; Sequentially moving the printing head to each grid point, switching to a stepping mode when approaching to the target position, and judging the touch time based on the fact that the resistance change rate of the central area unit exceeds a preset slope threshold value; Based on the determined touch time, reading the resistance increment of the adjacent sensing unit, and calculating a weighted average value by adopting a weight attenuated along with the distance; And mapping the weighted average value of each grid point to a two-dimensional plane, and generating a continuous preliminary deformation distribution map by using a bilinear interpolation method.
  4. 4. The 3D printing platform leveling method based on multi-point tactile feedback according to claim 1, wherein the identifying the abnormal response area comprises: defining abnormal response as point positions exceeding the range of the standard deviation of the global mean plus-minus twice of the primary deformation distribution map; counting abnormal point space aggregation conditions, and marking a sub-region with the abnormal point proportion exceeding 30% as a first-level attention region; forming an extended detection domain based on the marked primary region of interest expansion boundary, and repartitioning a sub-grid with the spacing of one third of the original grid in the extended detection domain; when the sub grid points perform high-density local scanning, the traveling speed is reduced to one fifth of the original speed.
  5. 5. The 3D printing platform leveling method based on multipoint tactile feedback according to claim 1, wherein the constructing the integrated surface topology model comprises: introducing a multi-resolution pyramid structure, performing multiple downsampling on the preliminary deformation distribution map, and performing upsampling on data in the extended detection domain; On each scale level, replacing a corresponding region in the preliminary deformation distribution map by data, and pushing downwards layer by layer from the coarsest level; performing Laplacian pyramid reconstruction from bottom to top, and overlapping detail residual errors of all levels to the finest level; Coupling the reconstructed deformation distribution with the synchronously updated height field, and attaching a confidence label to form a comprehensive surface topology model comprising height, resistance response and data reliability.
  6. 6. The 3D printing platform leveling method based on multi-point tactile feedback according to claim 1, wherein the dividing the functional area comprises: Setting a dynamic threshold value based on a resistance response mean value and a standard deviation in the comprehensive surface topology model, and distinguishing a low contact area, an overvoltage area and a moderate area; combining the type of the high-field data verification area, confirming the sinking recess for the low contact area and the bulge protrusion for the overvoltage area; Calculating a local curvature identification transition zone, and introducing a softening weight in a high curvature area; and summarizing the requirements of each region, and generating a three-dimensional adjustment guide map containing the height variation and the execution weight according to the original confidence level label, the region type and the local gradient stability.
  7. 7. The 3D printing platform leveling method based on multi-point tactile feedback according to claim 1, wherein the mapping the three-dimensional adjustment guide map to the adjustable support node comprises: Determining a supporting node set and the current position of the supporting node set, and acquiring material parameters and geometric dimensions of a platform; based on a thin plate bending theory, constructing an influence function of single-point adjustment, and representing the displacement influence of single-node adjustment on each point of the surface; Establishing a linear superposition model, regarding the total displacement response as a linear combination of independent actions of each supporting node, and solving the optimal adjustment quantity of each node through a weighted least square method; and generating a driving instruction sequence, checking whether the adjusted position exceeds the mechanical travel limit, and reassigning the adjustment quantity exceeding the limit to the adjacent node according to the weight proportion of the influence function.
  8. 8. The 3D printing platform leveling method based on multi-point tactile feedback of claim 1, wherein the staged implementation of support structure adjustment comprises: dividing the total adjustment quantity of each node into 3 to 5 equal sub-step sizes, and executing first-round adjustment according to the proportion; Based on the support state after the first round adjustment, selecting a first-level attention area and an area right above the support node to form a key monitoring set, and executing quick rechecking; comparing the monitored data of the primary round with the monitored data of the secondary round, calculating a normalized response change rate, and detecting an abnormal region deviating from the historical average sensitivity by more than 40%; And updating the residual adjustment quantity based on the abnormal detection result, slowing down the advancing speed of the associated node for the area which is close to the target response, and properly accelerating the adjustment rhythm for the area which has a larger gap.
  9. 9. The 3D printing platform leveling method based on multi-point tactile feedback according to claim 1, wherein the performing a final verification scan comprises: moving the printing head along the same grid path as the first-round full coverage scanning, maintaining pressure stability for 500 milliseconds at each sampling point, and collecting resistance readings every 50 milliseconds; removing the head-tail transition value, taking the median of the middle eight readings as a final value, and generating a final deformation graph; Carrying out pixel level difference on the final deformation map and the deformation map before leveling to generate a change map; and automatically judging the leveling quality based on three indexes of standard deviation reduction proportion, qualification rate and maximum residual error, and re-planning a supplementary adjustment scheme according to a new abnormal region when the leveling quality does not reach the standard, and circularly optimizing for at most three times.
  10. 10. A 3D printing platform leveling system based on multi-point haptic feedback for implementing the method of any one of claims 1-9, comprising: the touch sensing module is used for initializing a touch sensing network, establishing a coordinate calibration standard, obtaining an idle electric signal as a background standard value, executing first-round full-coverage low-speed scanning based on the calibrated coordinate standard, traversing divided grid points, recording readings of height encoders of all the points and resistance change data, and generating a preliminary deformation distribution map; The model building module is used for identifying an abnormal response area according to the preliminary deformation distribution diagram, planning an enhanced detection path, executing local scanning in the abnormal response area, fusing first-round full-coverage scanning data and local scanning data, and building a comprehensive surface topology model by combining a height field and a resistance response value; The motion control module is used for dividing functional areas based on the comprehensive surface topology model, quantifying compensation requirements of all areas, generating a three-dimensional adjustment guide map, mapping the three-dimensional adjustment guide map to adjustable supporting nodes, calculating adjustment quantity of all the nodes and generating a driving sequence; The verification module is used for carrying out support structure adjustment in stages according to the driving sequence, monitoring area response change after each stage of adjustment, dynamically correcting subsequent adjustment quantity, executing final state verification scanning according to the final adjusted surface state, generating a leveling report based on the verification result, and judging whether the leveling quality meets the standard.

Description

3D printing platform leveling method and system based on multipoint tactile feedback Technical Field The invention relates to the technical field of 3D printing, in particular to a 3D printing platform leveling method and system based on multipoint tactile feedback. Background In the field of 3D printing, leveling of the printing platform is one of the key steps to ensure a high quality printing result. In the prior art, the leveling of the printing platform is usually performed manually or semi-automatically, for example, by adjusting screws under the platform to achieve physical leveling. These methods rely on the experience and feel of the operator, are prone to human error, and take a long time. In addition, although some automatic leveling systems can reduce human intervention, the actual state of the whole printing surface cannot be comprehensively and accurately reflected only based on simple point-to-point measurement, so that the leveling precision is insufficient. The existing 3D printing platform leveling method is difficult to efficiently and accurately finish the leveling work of the whole printing surface, and particularly when a large-scale or complex-shaped printing task is processed, the problem of local unevenness in the leveling process is particularly remarkable, and the quality and consistency of a final printed piece are affected. Disclosure of Invention The invention aims to provide a 3D printing platform leveling method and system based on multipoint tactile feedback, which realize fine scanning and deformation analysis of a printing platform by initializing a tactile sensing network and establishing a coordinate calibration reference. In order to achieve the purpose, the technical scheme adopted by the invention is that the 3D printing platform leveling method based on the multipoint tactile feedback comprises the following steps: initializing a touch sensing network, establishing a coordinate calibration standard, obtaining an idle electric signal as a background standard value, performing first-round full-coverage low-speed scanning based on the calibrated coordinate standard, traversing divided grid points, recording readings of a height encoder of each point and resistance change data, and generating a preliminary deformation distribution map; Identifying an abnormal response area according to the preliminary deformation distribution diagram, planning an enhanced detection path, executing local scanning in the abnormal response area, fusing first-round full-coverage scanning data and local scanning data, and constructing a comprehensive surface topology model by combining a height field and a resistance response value; dividing functional areas based on the comprehensive surface topology model, quantifying compensation requirements of all areas, generating a three-dimensional adjustment guide map, mapping the three-dimensional adjustment guide map to adjustable supporting nodes, calculating adjustment quantity of all nodes, and generating a driving sequence; And carrying out support structure adjustment in stages according to the driving sequence, monitoring area response change after each stage of adjustment, dynamically correcting subsequent adjustment quantity, executing final state verification scanning according to the finally adjusted surface state, generating a leveling report based on a verification result, and judging whether leveling quality meets the standard or not. Preferably, initializing the haptic sensation network and establishing the coordinate calibration reference includes: Starting a distributed piezoresistive sensing grid, detecting nodes with the shielding response value lower than the lower limit of the threshold value or exceeding the upper limit of the threshold value, wherein the nodes are used for outputting electric signals by all sensing units; based on the induction grid of the shielded abnormal node, driving the three-axis linkage device to trigger each axis limit switch, repeatedly approaching the limiter three times, taking an average value, and determining a coordinate origin; planning a standard test point array by taking a coordinate origin as a center, moving a printing head to a preset coordinate, and collecting data after judging that the fluctuation of the acceleration sensor reading is smaller than a set tolerance within five continuous seconds; and setting a background reference value based on the acquired no-load electric signals for calculating the net change amount in the first-round full-coverage low-speed scanning. Preferably, the performing the first-round full-coverage low-speed scan includes: Dividing a printing platform into M multiplied by N equidistant sampling units, and setting the distance between adjacent points as the smallest dimension of the platform divided by a resolution coefficient; Sequentially moving the printing head to each grid point, switching to a stepping mode when approaching to the target position, and ju