Robot Tracking Control with Reliability-Based Memory Deletion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In robot welding operations, memory capacity issues arise due to displacement of workpieces caused by heat strain, leading to welding defects, and existing solutions are complex to implement, potentially resulting in insufficient memory space for accumulating sensing information.
Innovation Solution
A robot control apparatus that automatically deletes information with low reliability from memory to prevent memory overflow, allowing for continuous accumulation of sensing data without manual adjustment of memory intervals, thereby maintaining tracking precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If information according to sensing result is accumulated in memory at high frequency, then tracking precision is improved, but memory capacity is insufficient
Solution Approach 1:
The patent changes the parameter of information reliability by calculating dispersion levels of sensing results. Information is selectively deleted based on whether its dispersion level exceeds a threshold, rather than deleting based on time or position alone. This allows the system to maintain high tracking precision by preserving low-dispersion (high-reliability) information while freeing memory capacity by removing high-dispersion (low-reliability) information.
2Quantity of substance
If manual adjustment of memory accumulation intervals is performed, then memory capacity sufficiency is improved, but operation complexity increases
Solution Approach 1:
The system performs self-service by automatically managing memory capacity through dispersion level calculation. The control device autonomously determines which information to delete based on calculated dispersion levels, eliminating the need for manual adjustment of memory accumulation intervals. This self-managing mechanism resolves the contradiction by providing automatic memory management that maintains sufficiency without operational complexity.
Solution Approach 2:
The patent implements feedback by continuously calculating dispersion levels of sensing results and using this information to dynamically delete inappropriate data from memory. This closed-loop feedback mechanism allows the system to automatically adapt memory content based on actual sensing quality, ensuring memory capacity sufficiency without manual intervention while maintaining tracking precision through preservation of high-reliability information.
3Quantity of substance
If information is deleted from memory to prevent overflow, then memory capacity is improved, but tracking precision may deteriorate
Solution Approach 1:
The patent applies local quality by treating different information differently based on their individual dispersion levels. Rather than uniformly deleting information or keeping all information, the system selectively deletes only those information whose dispersion levels exceed the threshold. This localized quality-based deletion ensures memory capacity is freed while preserving the precision-critical low-dispersion information, thus resolving the contradiction between memory capacity and tracking precision.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
In order to solve a problem that, in a memory in which information that is used in a tracking operation is to be temporarily stored, information cannot be accumulated in the memory when the accumulating intervals and the reading intervals do not match each other, a robot control apparatus 1 includes: a memory 13; an accepting unit 12 that accepts a sensing result of a laser sensor 5 configured to detect a shape of a working target 6 before an operation of a working tool 4 of a robot 2, and accumulates information according to the sensing result, in the memory 13; a memory management unit 14 that, in a case in which the memory 13 is running short of free space, deletes the information in the memory 13; and a control unit 15 that moves the working tool 4 based on teaching data, and corrects the movement of the working tool 4 based on the information according to the sensing result stored in the memory 13. Accordingly, it is possible to prevent a situation in which information according to a sensing result cannot be accumulated in the memory 13.