Model-Based Actuator Abnormality Detection From Control Deviations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems fail to detect unstable actuator states indicative of potential failures, as they only monitor for abnormalities at specific instances rather than providing early detection based on continuous system behavior.
Innovation Solution
An information processing device that calculates the degree of suitability between observed and predicted system information using a model, and calculates differences in manipulation amounts to detect anomalies before they lead to unstable actuator states.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If abnormality detection is based on single-point event monitoring (as in PTL 1 and PTL 2), then the detection method is simple, but unstable states indicating potential failures cannot be detected
Solution Approach 1:
The system performs preliminary actions by continuously calculating the degree of suitability between observation information and prediction information before actual failure occurs. This allows early detection of unstable states by monitoring trends over time rather than waiting for critical failure events, enabling proactive maintenance while keeping the detection mechanism relatively simple.
Solution Approach 2:
The patent implements continuous monitoring by repeatedly calculating the degree of suitability between observed and predicted system states over time. This continuous assessment captures unstable transient states that single-point monitoring would miss, improving reliability without requiring complex additional hardware beyond continuous data processing.
2Reliability
If continuous monitoring of system state is implemented to detect unstable states, then early detection of abnormalities is enabled, but computational load and processing complexity increase
Solution Approach 1:
The system uses feedback by comparing observation information with prediction information derived from system models. This feedback mechanism allows the system to identify deviations from expected behavior and detect unstable states. The feedback loop processes data efficiently by focusing computational resources on calculating the degree of suitability rather than full system simulation, balancing early detection capability with energy consumption.
3Measurement precision
If the degree of suitability calculation is performed continuously, then early detection precision is improved, but processing time and computational resources increase
Solution Approach 1:
The patent applies partial action by calculating the degree of suitability only when necessary - specifically when comparing observation information with prediction information from system models. Rather than performing full system analysis continuously, the method focuses computational effort on the specific calculation needed for abnormality detection, achieving sufficient precision while reducing overall processing time and resource consumption.
Data Source
AI summary
Provided is an information processing device, etc., that provides information which is the basis for quick detection of abnormalities that occur in a device. An information processing device calculates a degree of suitability between observation information and prediction information, the observation information observed for a system suffering an effect from an certain device, the prediction information predicted in accordance with a model for a state of the system; and calculates a difference between manipulation amount to the certain device and predictive manipulation amount predicted for the manipulation amount based on the model, the difference being a difference in case that the degree satisfies a predetermined calculation condition.