Controller Feature Compression for Short-Cycle Abnormality Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current abnormal symptom diagnosis apparatuses in factory automation are limited by the difficulty in collecting input data in short cycles due to transmission through communication networks like LAN or WAN, which restricts their ability to diagnose rapid abnormalities.
Innovation Solution
A controller with a feature quantity generation unit, machine learning unit, abnormality detection unit, and data compression unit that enables the detection of abnormalities in control targets without returning responses, allowing for shorter-cycle monitoring and data transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multidimensional sensor data is transmitted through communication networks (LAN/WAN) for abnormal symptom diagnosis, then the system can diagnose abnormal symptoms, but the data collection cycle cannot be shortened to milliseconds or microseconds due to network transmission limitations
Solution Approach 1:
The patent segments the data processing function by separating the controller (which collects data at high speed) from the abnormal symptom diagnosis apparatus (which analyzes the data). The controller generates feature quantities from raw sensor data and transmits only these extracted features through the network, rather than transmitting all raw multidimensional sensor data. This segmentation allows the controller to operate at millisecond/microsecond cycles while the diagnosis apparatus receives processed information suitable for network transmission.
Solution Approach 2:
The patent extracts essential feature quantities from the raw sensor data at the controller level before transmission. The feature quantity generation unit identifies and extracts only the critical abnormality-indicating features from the multidimensional sensor data, removing redundant information. This extraction enables shorter data collection cycles because only the essential features need to be transmitted and analyzed, not the complete raw data set.
2Speed
If feature quantity generation and machine learning are performed to enable shorter-cycle monitoring, then the monitoring speed increases, but the device complexity increases due to additional processing units
Solution Approach 1:
The patent implements multi-functionality by integrating the feature quantity generation unit, machine learning unit, and abnormality detection unit within the existing controller architecture. Rather than adding separate dedicated hardware for each function, the controller is designed to perform multiple functions (data collection, feature extraction, machine learning, and abnormality detection) using a unified processing platform. This approach enables shorter monitoring cycles while avoiding the complexity of multiple separate devices.
Data Source
AI summary
A controller includes a feature quantity generation unit that generates, from data associated with a control target, a feature quantity appropriate for detecting an abnormality in the control target, a machine learning unit that performs machine learning using the feature quantity, an abnormality detection unit that detects the abnormality based on an abnormality detection parameter determined from a learning result of the machine learning, and the feature quantity, an instruction unit that instructs the abnormality detection unit to detect the abnormality, and a data compression unit that compresses data about the feature quantity and provides the compressed data to the machine learning unit and the abnormality detection unit. The instruction unit transmits a request for detecting the abnormality to the abnormality detection unit. The abnormality detection unit detects the abnormality without returning a response to the request.