Robot Controller Error Correction for Mixed Teaching Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot controllers struggle to accurately position industrial robots due to inherent errors, which are not properly corrected by existing techniques, especially in the marginal regions of the operating range or when dealing with mixed teaching data from off-line and manual teaching methods.
Innovation Solution
A robot controller that includes input means for angular displacements, forward conversion means to convert these into position-attitude data, correcting means to account for inherent errors, and inverse conversion means to provide corrected angular displacements, allowing for accurate positioning and handling of mixed teaching data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If existing error correction techniques are used to correct positional deviations, then manufacturing precision is improved, but the robot cannot accurately position itself in marginal regions of the operating range or when using mixed teaching data
Solution Approach 1:
The patent applies dynamics by making the error correction adaptive rather than static. The controller dynamically selects between different correction methods (first error correction for positional data, second error correction for angular displacement data) based on the type of teaching data being used. This dynamic adaptation allows the system to maintain high positioning accuracy across the entire operating range, including marginal regions where static correction methods fail.
Solution Approach 2:
The patent changes the correction parameter based on the teaching data type. When positional data is used, the system applies correction based on positional errors; when angular displacement data is used, it applies correction based on angular errors. This parameter change allows the system to handle mixed teaching data and marginal operating regions effectively, resolving the contradiction between manufacturing precision and positioning reliability.
2Productivity
If off-line teaching method is used to specify angular displacements, then productivity is improved, but the robot cannot uniquely determine its attitude due to multiple possible angular displacements
Solution Approach 1:
The patent implements feedback by using the measured actual position and attitude of the robot to correct the taught angular displacements. The controller compares the actual robot state with the desired state and uses this feedback information to calculate correction amounts. This feedback mechanism allows the system to maintain high productivity from off-line teaching while resolving the attitude determination ambiguity through iterative correction.
Solution Approach 2:
The patent replaces the mechanical assumption of unique pose correspondence with a computational correction system. Instead of relying on the mechanical property that each pose corresponds to unique angular displacements, the system uses computational error correction based on actual measurements to determine the correct attitude, enabling unique determination even when multiple angular displacements are theoretically possible.
3Manufacturing precision
If manual teaching method is used to ensure accurate positioning, then manufacturing precision is improved, but productivity decreases due to time-consuming teaching process
Solution Approach 1:
The patent applies preliminary action by performing error correction calculations in advance during the off-line teaching phase. The system pre-calculates correction amounts based on the robot model and teaching data, so that when the robot executes the taught motions, the corrections are already prepared. This preliminary action allows the system to achieve manual-teaching-level accuracy without the time-consuming manual adjustment process, thereby maintaining high productivity.
Data Source
AI summary
An input unit enters angular displacements by which drive shafts of a robot arm are to be turned as teaching data into a control unit. The control unit converts the input angular displacements into position-attitude data, namely, converted commands, indicating a position of the free end of the robot arm and an attitude of the robot in a rectangular coordinate system through forward conversion. The control unit corrects the position-attitude data on the basis of inherent errors in the robot to provide corrected position-attitude data. The control unit converts the corrected position-attitude data into corrected angular displacements through inverse conversion and gives the corrected angular displacements to an actuator included in the robot. The inherent errors in the robot include mechanismic errors resulting from machining errors and assembling errors, installation errors and errors in the origins of axes.