Sensor Signal Harmonic Removal Using Discounted Averaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing signal processing techniques, such as time synchronous averaging, are inefficient or ineffective in removing periodic noise from sensor signals from rotating machinery components like magnetostrictive sensors due to varying runout signatures under different operating conditions.
Innovation Solution
Implementing a discounted averaging process in conjunction with phase reference information to efficiently remove periodic noise from sensor signals, allowing for real-time or continuous noise reduction with minimal computational overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If time synchronous averaging is used to remove periodic noise from sensor signals, then noise reduction can be achieved, but the method becomes inefficient and ineffective when runout signatures vary under different operating conditions
Solution Approach 1:
The patent applies dynamics by making the averaging process adaptive rather than static. The discounted averaging algorithm dynamically adjusts the weight of historical data based on a discount factor, allowing the system to adapt to varying runout signatures under different operating conditions. This resolves the contradiction by enabling the noise reduction method to remain effective (improving measurement precision) while maintaining processing efficiency through incremental updates rather than full re-processing (improving productivity).
Solution Approach 2:
The patent changes the parameter of data weighting in the averaging process by introducing a discount factor that exponentially decreases the weight of older measurements. This parameter change allows the system to adapt to varying operating conditions by emphasizing recent data while still benefiting from historical trends, thereby maintaining both noise reduction effectiveness and processing efficiency under changing conditions.
2Measurement precision
If traditional averaging methods are applied to remove periodic noise, then some noise reduction is achieved, but computational overhead increases and real-time processing becomes difficult
Solution Approach 1:
The patent applies preliminary action by pre-defining the discount factor and the structure of the discounted averaging algorithm before actual signal processing begins. This preliminary setup allows the system to perform lightweight incremental updates during real-time operation, avoiding the need for computationally intensive full-data re-processing. The noise reduction capability is maintained through this pre-configured adaptive structure while computational overhead is significantly reduced.
Solution Approach 2:
The patent uses partial action by applying discounted averaging that processes only the necessary portion of historical data with exponentially decreasing weights, rather than processing all historical data equally. This partial processing approach maintains noise reduction capability by focusing on relevant recent data while avoiding the excessive computational overhead of processing entire historical datasets, enabling real-time processing.
3Ease of manufacture
If conventional noise removal techniques are used, then processing can be performed, but noteworthy fluctuations indicating mechanical issues may be obscured
Solution Approach 1:
The patent applies dynamics by using an adaptive discounted averaging process that dynamically adjusts to varying operating conditions, preventing the obscuring of noteworthy fluctuations. Unlike static conventional methods that may mask transient mechanical issues, the dynamic discounting mechanism allows significant fluctuations to stand out against the adaptively filtered background, maintaining both processing capability and information integrity for mechanical issue detection.
Solution Approach 2:
The patent incorporates feedback by continuously updating the averaged signal with new measurements weighted by the discount factor, creating a feedback loop that adapts to changing conditions. This feedback mechanism ensures that noteworthy fluctuations indicating mechanical issues are not obscured, as the system continuously adjusts to the current operating state while maintaining the ability to detect anomalies.
Data Source
AI summary
A method of removing noise data from a sensor signal can include receiving, by at least one data processor, from a sensor configured to detect a variable physical property of a target object, a sensor signal corresponding to the detected variable physical property, receiving, by the at least one data processor, from a phase reference generator, phase reference information representative of a noise feature of the sensor signal, removing, by the at least one data processor, using a discounted averaging process and the phase reference information, periodic noise from the sensor signal to produce a noise-reduced signal, and providing, by the at least one data processor, the noise-reduced signal.