3D Path Deformation Fitting for Accurate Robot Workpiece Paths
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In robot applications, real work pieces often suffer from machining errors and deformations, leading to inaccuracies in path generation when mapping CAD models to point clouds, especially when small features are involved.
Innovation Solution
A method for deformation fitting that uses parametric expressions to model and optimize deformation variables, reducing computational effort while enhancing accuracy by iteratively adjusting parameters using an iterative gradient descent approach.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If traditional mapping operation is used to transform paths from CAD model base to world base, then the path generation process is simple, but the path accuracy deteriorates due to machining errors and deformations in real parts
Solution Approach 1:
The patent applies preliminary action by performing deformation fitting of the CAD model to the point cloud before path generation. The system pre-aligns and pre-deforms the CAD model to match the actual workpiece geometry, so that when paths are generated on the deformed model, they automatically account for machining errors and deformations. This preliminary alignment step ensures high path accuracy without requiring complex real-time adjustments during execution.
Solution Approach 2:
The patent utilizes parameter changes by transforming the rigid transformation parameters (rotation and translation) into deformation parameters that capture machining errors and geometric variations. The system represents the deformation as a combination of rigid transformation and non-rigid deformation, allowing the model to adapt to actual workpiece geometry by adjusting these parameters to minimize alignment errors between CAD and point cloud data.
2Measurement precision
If non-rigid registration is used to align CAD model with point cloud, then the alignment accuracy is improved, but the computational effort increases significantly
Solution Approach 1:
The patent applies segmentation by dividing the alignment process into two distinct stages: rigid transformation and non-rigid deformation. First, a rigid transformation (rotation and translation) is applied to roughly align the CAD model with the point cloud. Then, a non-rigid deformation is applied to refine the alignment. This segmentation allows the computationally intensive non-rigid registration to be applied only to the already roughly aligned models, significantly reducing computational effort while maintaining high alignment accuracy.
Solution Approach 2:
The patent applies preliminary action by performing rigid transformation before non-rigid registration. The rigid transformation pre-aligns the CAD model and point cloud, establishing a good initial correspondence between points. This preliminary alignment reduces the search space for the subsequent non-rigid registration, allowing it to converge faster and with less computational effort while achieving higher precision alignment.
3Manufacturing precision
If CAD model is directly mapped to point cloud without deformation fitting, then the process is computationally efficient, but the path accuracy deteriorates due to not accounting for machining errors
Solution Approach 1:
The patent applies preliminary action by performing deformation fitting as a preprocessing step before path generation. The CAD model is deformed to match the actual workpiece geometry by aligning it with the point cloud data. This preliminary deformation accounts for machining errors and geometric variations, ensuring that subsequent path generation is based on an accurate model without requiring time-consuming adjustments during path execution.
Solution Approach 2:
The patent segments the overall process into deformation fitting and path generation stages. The deformation fitting stage, which accounts for machining errors, is performed once as preprocessing. The path generation stage then operates on the already-fitted model, achieving high path accuracy without repeating the computationally intensive deformation calculations, thus minimizing computational time loss.
Data Source
AI summary
Embodiments of present disclosure provide a method, a device, a computer-readable storage medium and a computer program product for deformation for path generation. The method includes deforming a three-dimensional model of a path for an object based on a mapping of the three-dimensional model and a point cloud of the object. A plurality of parameters in the deformed three-dimensional model can be adjusted based on a comparison of the three-dimensional model and the point cloud of the object. The method includes generating an updated path for the object according to the adjusted three-dimensional model.