Robot Abnormality Position Estimation Using Servo Waveform Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot maintenance systems cannot accurately determine the position of abnormalities, leading to uncertainty in which parts need replacement, causing inefficiencies and increased maintenance time.
Innovation Solution
A method that detects abnormalities in robots using servo motors by analyzing current values and position deviations, calculating representative values over cycles and days, and estimating the abnormality's position through waveform analysis, allowing for precise identification of maintenance needs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If only current value and deviation amount monitoring is performed, then abnormality detection capability is improved, but position identification capability deteriorates
Solution Approach 1:
The patent segments the monitoring process into two distinct phases: an abnormality detection phase that monitors current values and deviation amounts, and a position identification phase that analyzes waveform patterns after abnormality detection. This segmentation allows each phase to focus on its specific function without interference, resolving the contradiction between detection capability and position identification capability.
Solution Approach 2:
The patent performs preliminary waveform data collection and pattern analysis by storing multiple types of waveform data (current, position, speed) before abnormality occurs. When abnormality is detected, the system compares the actual waveform against pre-stored patterns to identify the abnormal position, eliminating the need for real-time complex analysis during critical moments.
2Measurement precision
If comprehensive waveform analysis is performed for all parameters, then position estimation accuracy is improved, but system complexity deteriorates
Solution Approach 1:
The patent applies local quality by selecting and analyzing only the specific waveform parameters and patterns that are relevant to particular types of abnormalities. Different abnormal positions exhibit distinct waveform characteristics in specific parameters, so the system focuses analysis on those relevant parameters rather than uniformly analyzing all parameters, thereby maintaining high accuracy while reducing complexity.
Solution Approach 2:
The patent changes the analysis parameters dynamically based on the detected abnormality type and position. The system selects from multiple types of waveform data (current, position, speed) and their various characteristics (amplitude, frequency, phase) depending on what is most relevant for identifying the specific abnormal position, optimizing the analysis process for each situation.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided is a method for estimating a position where an abnormality has occurred. The method for estimating a position where an abnormality has occurred in a robot includes an abnormality detection step of detecting occurrence of an abnormality, and a position detection step of detecting a position where the abnormality has occurred when the occurrence of the abnormality is detected in the abnormality detection step.