Causal Structure Learning for Faster Fault Cause Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In autonomous control loop schemes, relying solely on observable data for fault handling leads to prolonged time in identifying the cause of failures, as it depends on maintenance personnel's manual investigation.
Innovation Solution
An information processing device that collects and combines observable data at predetermined intervals, uses a generator and discriminator to update a causal structure matrix, and outputs the causal structure between data points, enabling faster fault cause identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If only observable data is displayed for fault handling, then the system maintains simplicity in the monitoring approach, but it takes a long time to identify the cause of failure
Solution Approach 1:
The system performs preliminary actions by pre-learning and storing causal relationships between observable data items before actual fault occurrence. The causal structure matrix is constructed in advance through repeated learning with generator and discriminator, so that when a failure occurs, the pre-established causal relationships can be immediately queried to identify the cause, eliminating the need for time-consuming manual analysis during actual fault handling
Solution Approach 2:
The patent introduces a causal structure matrix as an intermediary between raw observable data and fault diagnosis. This matrix serves as a mediator that encodes causal relationships learned through GAN training, allowing the system to quickly infer failure causes by querying pre-established causal patterns rather than directly analyzing raw data during incidents
2Productivity
If manual investigation by maintenance personnel is used, then the system requires minimal processing infrastructure, but fault handling efficiency is low
Solution Approach 1:
The system implements self-service by automatically learning causal relationships from operational data and autonomously identifying failure causes without requiring manual investigation. The generator and discriminator work together to self-train the causal structure matrix, and the system automatically queries this matrix to diagnose failures, replacing manual maintenance personnel analysis with automated self-diagnosis capabilities
Solution Approach 2:
The patent transforms the parameter representation by converting raw observable data into a structured causal relationship framework. Through GAN-based learning, the system changes the data parameters from simple observable values to a causal structure matrix that encodes probabilistic causal relationships, enabling efficient automated query and diagnosis
Data Source
AI summary
An information processing device includes: a working unit configured to collect and combine various observable data acquired from a managed target at predetermined time intervals; a processing unit configured to input the combined observable data and update a causal structure matrix by repeatedly learning with a generator that generates pseudo-generated data using the causal structure matrix and a discriminator that identifies whether the pseudo-generated data is false or not, wherein the causal structure matrix represents a causal structure between the combined observable data; and an output unit configured to output the causal structure between the observable data based on the causal structure matrix.