Conveyor Segment Control for Self-Learning Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing conveyor systems lack efficient methods for identifying data patterns in real-time operations, relying on predefined reference values for error detection, which limits their adaptability and effectiveness.
Innovation Solution
A method that utilizes self-learning systems, such as neural networks, to identify conveyor segment-specific and conveyed material-specific data patterns without prior training, allowing for continuous data pattern generation and anomaly detection based on continuity criteria, enabling dynamic analysis of large data volumes generated during normal operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If predefined reference values are used for error detection, then the system structure remains simple, but the adaptability and effectiveness of anomaly detection deteriorates
Solution Approach 1:
The system performs self-learning by automatically generating reference values from operational data without requiring external training or manual configuration. The conveyor system itself serves as the training data source, enabling the algorithm to adapt to specific operational patterns autonomously
Solution Approach 2:
The system collects and stores operational data in advance during normal operation, building a database of reference patterns before anomaly detection is needed. This preliminary data accumulation enables rapid adaptation when anomalies occur
2Measurement precision
If large volumes of operational data are collected and analyzed, then the precision of anomaly detection improves, but the processing time and computational resources increase
Solution Approach 1:
The system extracts only the most relevant features and patterns from the large volume of operational data, focusing analysis on key parameters that indicate anomalies. This selective extraction maintains detection precision while reducing processing requirements
Solution Approach 2:
The system applies analysis selectively to specific conveyor segments or time periods when anomalies are suspected, rather than continuously analyzing all data. This partial application reduces processing time while maintaining detection effectiveness
3Productivity
If conveyor segments are operated with high productivity, then the output increases, but the difficulty of detecting and measuring anomalies increases due to reduced continuity of patterns
Solution Approach 1:
The system continuously monitors operational parameters and provides feedback when deviations from learned patterns are detected. This real-time feedback enables anomaly detection even during high-speed operation by comparing current state against reference patterns
Solution Approach 2:
The system combines multiple data sources and parameters (motor current, speed, position, sensor readings) into a composite analysis model. This multi-parameter approach detects anomalies more reliably even when individual parameters vary during high-productivity operation
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
The invention relates to a method for operating a conveyor assembly (1), in particular comprising controlling and/or monitoring the conveyor assembly. The conveyor assembly (1) comprises a plurality of conveyor segments (2a..e), and each conveyor segment (2) is designed to convey a material (9) to be conveyed in a conveying direction (F); the conveyor segments (2) are arranged one after the other such that the material (9) to be conveyed is transferred from an upstream conveyor segment (2a..2d) to a downstream conveyor segment (2b..e), and each conveyor segment (2) has a conveyor segment drive (3M) which is designed to provide a driving force, in particular isolated for this conveyor segment, in order to convey the material (9) to be conveyed on this conveyor segment; a zone controller (11) for controlling the conveyor segment drive (3M) is assigned to each conveyor segment (2). The method comprises the following method steps: detecting conveyor segment data (R) which are generated during the operation of a conveyor segment (2); and collecting conveyor segment data (R) from a plurality of conveyor segments.