Fault Detection via Command Signal Standard Deviation Ratio

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing fault detection methods in control systems face challenges in accurately distinguishing between inherent errors and faults due to cumulative error sources like gain/offset errors, phase shift, latency, and random noise, especially in high-frequency systems where the average command value is lower than the inherent error in the wraparound signal.

Innovation Solution

A fault detection method that calculates the standard deviation of both the command value and the wraparound signal and determines their ratio to detect faults, neutralizing the effects of inherent errors and noise, allowing for accurate fault detection even when the command value approaches the error level in the wraparound circuit.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If error threshold limit is set to account for inherent errors in wraparound signal, then false fault detection is minimized, but fault detection capability is reduced due to large error band

Engineering Contradiction:
Improvefault detection reliabilityVSAvoidfault detection precision
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The patent changes the parameter used for fault detection from absolute error comparison to ratio-based standard deviation comparison. By monitoring the ratio of standard deviations between command signal and wraparound signal, the system can detect faults without being constrained by fixed error thresholds, thus resolving the contradiction between reliability and precision in fault detection

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent uses the wraparound signal as a copy of the command signal to establish a baseline for comparison. By comparing the standard deviation ratio between the original command signal and its wraparound copy, the system can identify faults while accounting for inherent errors, thereby improving both reliability and precision simultaneously

Inventive Principle:
Principle #26Copying

2Measurement precision

If error threshold limit is set low to maximize fault detection capability, then fault detection sensitivity is improved, but probability of false fault detection increases

Engineering Contradiction:
Improvefault detection sensitivityVSAvoidfalse fault detection rate
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent transitions from using fixed error thresholds to using dynamic standard deviation ratios. This parameter change allows the system to adapt to varying operating conditions and inherent error levels, enabling high sensitivity fault detection without increasing false alarm rates, as the baseline automatically adjusts to normal system behavior

Inventive Principle:
Principle #35Parameter changes

3Speed

If average command value is low in high frequency valve operation, then system responsiveness is improved, but fault detection becomes unreliable as command value approaches error level

Engineering Contradiction:
Improvevalve response speedVSAvoidfault detection accuracy
Core Design Contradiction:
SpeedVSMeasurement precision

Solution Approach 1:

The patent changes the detection parameter from absolute command value comparison to standard deviation ratio analysis. This allows fault detection to remain accurate even when command values are low, as the method detects changes in signal variability rather than relying on absolute value thresholds, thus maintaining both fast response and detection accuracy

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP2343612B1Fault detection in a system under control
Publication Date: 2013.06.19 HAMILTON SUNDSTRAND CORP
  • EP2343612B1 patent drawingFigure 1
  • EP2343612B1 patent drawingFigure 2
  • EP2343612B1 patent drawingFigure 3

AI summary

A method (300) for fault detection in a system under control includes monitoring a command value and a wraparound signal (301) by a fault detection module (203); determining a standard deviation of the command value (302) by the fault detection module; determining a standard deviation of the wraparound signal (303) by the fault detection module; determining a ratio of the standard deviation of the command value to the standard deviation of the wraparound signal (304) by the fault detection module; and determining the presence or absence of a fault in the system under control by the fault detection module based on the ratio (305) by the fault detection module. A fault detection module (203) and a computer program product comprising a computer readable storage medium containing computer code that, when executed by a computer, implements a method (300) for fault detection are also provided.