Press Load Monitoring for Die-Specific Eccentric Abnormality Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for detecting abnormalities in press machines are limited to stages where peak loads occur, failing to detect issues in stages with varying load timings, and cannot effectively manage die abnormalities or product accuracy defects within allowable eccentric load ranges.
Innovation Solution
A press machine and method that include a detection unit for load values, a storage unit for associating load values with die identification, a calculation unit for generating eccentric load distribution data, and a determination unit for detecting abnormalities based on eccentric loads at predetermined press angles, using machine learning to establish normal ranges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the conventional method calculates eccentric amount only when total load value indicates peak value, then the press machine protection function is provided, but abnormality detection in stages without peak load is not possible
Solution Approach 1:
The press cycle is divided into multiple stages, and eccentric load calculation is performed for each stage independently based on load values at each stage. This segmentation allows abnormality detection in each stage rather than only at the peak load stage, resolving the contradiction between protection reliability and detection coverage.
Solution Approach 2:
The system dynamically determines whether to calculate eccentric load based on whether the current stage is a peak load stage or a non-peak load stage. This dynamic approach enables flexible abnormality detection across all stages while maintaining protection function at peak stages, resolving the contradiction between reliability and versatility.
2Reliability
If the conventional method only detects abnormalities exceeding allowable eccentric load diagram, then press protection is achieved, but die abnormality and product accuracy defects within allowable range cannot be detected
Solution Approach 1:
The system adds a new detection dimension by comparing eccentric load against learned normal range in addition to the conventional allowable eccentric load diagram. This creates a two-layer detection system: the allowable diagram provides press protection while the learned normal range enables die abnormality detection, resolving the contradiction between protection and precision.
Solution Approach 2:
The system performs preliminary learning of normal eccentric load range during normal operation before actual abnormality detection. This preliminary action establishes a baseline for comparison, enabling detection of die abnormalities within the allowable range while maintaining press protection functionality.
3Adaptability or versatility
If load values are collected and processed for each stage, then abnormality detection in multiple processes becomes possible, but the system complexity increases
Solution Approach 1:
The same load detection unit and processing logic are used across all stages and press types. The system universally applies eccentric load calculation and comparison against learned normal range, enabling multi-process detection without proportionally increasing complexity. This universal approach resolves the contradiction between versatility and complexity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A press machine (1) includes: a detection unit (102) that detects a load value during press working on a workpiece material; a storage unit (103) that stores load values for one cycle of the press machine (1) detected by the detection unit (102) in association with identification information of a die (20, 21) attached to the press machine (1) during detection of the load values; a calculation unit (104) that calculates eccentric loads for predetermined press angles based on the stored load values for one cycle of the press machine (1) and generates eccentric load distribution data for each die (20, 21); a determination unit (105) that determines an abnormality based on the eccentric loads for the predetermined press angles obtained based on the load values for one cycle of the press machine (1) detected by the detection unit (102) and the eccentric load distribution data corresponding to a die (20, 21) attached to the press machine (1) during the detection of the load values; and a notification unit (106) that notifies of the abnormality based on a determination result of the determination unit (102).