Neural Network Sheet Bending Angle Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for measuring the bending angle of sheets in press brake machines are limited in accuracy and efficiency, particularly in dynamically changing bending operations, as they rely on distance measurements and optical detection of light patterns which can be prone to distortion and require extensive data processing.
Innovation Solution
The apparatus employs a structured light source and video camera system connected to a processing unit with a neural network that projects and records a light pattern on the sheet, converts the image into a point cloud, samples key points, and uses a trained neural network to accurately determine the bending angle, allowing for real-time adjustment and high precision measurement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If distance measurements and optical detection of light patterns are used to measure bending angles, then measurement capability is provided, but measurement precision deteriorates due to distortion and data processing complexity
Solution Approach 1:
The patent replaces traditional mechanical distance measurement systems with a laser line projection and video capture system. Instead of using complex distance sensors and optical pattern analysis, the invention projects a laser line onto the sheet and uses video camera coordinates to directly determine bending angle, substituting mechanical measurement with optical-field measurement that requires simpler processing.
2Productivity
If traditional light pattern projection and distance measurement methods are used, then bending angle detection is enabled, but measurement speed decreases due to extensive data processing requirements
Solution Approach 1:
The patent extracts only the essential information needed for bending angle measurement by projecting a laser line and capturing video coordinates, eliminating the need for extensive distance measurements and complex pattern analysis. This extraction approach reduces data processing time while maintaining measurement capability.
3Reliability
If complex optical detection systems are used to measure bending angles dynamically, then measurement capability is provided, but reliability deteriorates due to distortion susceptibility
Solution Approach 1:
Instead of analyzing how light patterns deform on the sheet surface (which is susceptible to distortion), the patent inverts the approach by projecting the laser line and using the video camera's coordinate system to directly calculate bending angle. This inversion makes the measurement reference frame independent of sheet surface distortion, improving reliability in dynamic bending conditions.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This solution provides a highly accurate and efficient method for measuring bending angles by reducing data processing complexity and eliminating the need for extensive databases, enabling precise and autonomous reconstruction of bending angles, thus improving the accuracy and speed of the measurement process.
Implementation Method 1
Each sensor (4) comprises a laser (42) adapted to project a light line onto a surface (31, 32) of the sheet (3)
Implementation Method 2
and a video camera (41) adapted to record an image obtained on the surface (31, 32) by the light line projection during the operation of bending sheet (3)
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An apparatus (2) for measuring a bending angle of a sheet (3) is described, comprising a processing unit (10) and at least one sensor (4) comprising a light source (42) which projects a light pattern (52) on at least one side (31, 32) of the sheet (3), and recording means (41) adapted to record the image of the projection of said light pattern on a side of the sheet. The processing unit (10) is adapted to control the recording means for recording the image in at least one time instant (Tregl; Tregl, Treg2... Tregn) during the operation of bending the sheet (3); the control unit is capable of transforming the recorded image (56) into a point cloud (57) and comprises a neural network (115) adapted to associated a bending angle value (60, 200) with the point cloud.