Merged Reference Fingerprint for Digital Twin Prognostics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Digital twins face challenges in accurately monitoring high-frequency waveforms from assets like vibration, acoustics, and electromagnetic interference due to bandwidth constraints, leading to inefficiencies in prognostic anomaly detection and increased false or missed alarms.
Innovation Solution
The implementation of a merged reference fingerprint system that aligns and merges time series signals from a golden system to create a 3D fingerprint, which is then used to train a machine learning model for improved prognostic anomaly detection, allowing for more accurate and rapid analysis of field twins, even with high-frequency waveforms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensor telemetry is sampled at high frequency to capture detailed asset operation data, then measurement precision is improved, but bandwidth constraints of cloud services cause data loss and increased false/missed alarms
Solution Approach 1:
The patent segments the high-frequency sensor telemetry into multiple frequency bands using Fourier transforms. By dividing the signal spectrum into different bands and selectively processing only the most relevant bands, the system maintains measurement precision for critical frequencies while reducing overall data transmission volume to fit within cloud service bandwidth constraints.
Solution Approach 2:
The patent extracts and removes redundant or less important frequency components from the sensor telemetry before transmission. By identifying and eliminating unnecessary data elements (low-information frequency bands), the system reduces bandwidth requirements while preserving the essential operational characteristics needed for accurate digital twin comparison.
2Measurement precision
If multiple reference fingerprints are merged to improve signal-to-noise ratio, then prognostic accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent performs preliminary alignment of multiple reference fingerprints using cross-correlation techniques before the actual merging process. By pre-synchronizing the temporal and phase characteristics of individual fingerprints, the system reduces the computational burden during the merging phase and improves the effectiveness of noise cancellation through constructive interference of aligned signals.
3Measurement precision
If high-frequency waveforms are transmitted to cloud environments for analysis, then measurement precision is improved, but bandwidth constraints increase transmission costs and processing time
Solution Approach 1:
The patent extracts only the essential frequency band information from high-frequency waveforms before transmission to the cloud. By removing redundant high-frequency components that contain minimal prognostic value, the system maintains sufficient analysis accuracy while dramatically reducing data transmission volumes and associated processing time in cloud environments.
Data Source
AI summary
Systems, methods, and other embodiments associated with a merged-surface 3D fingerprint technique for improved prognostics for assets are described. In one embodiment, a method includes generating a set of time series signals from sensor readings of a reference device while the reference device is operated through multiple individual iterations of an exercise profile. The reference device operates in a known undegraded state. The method then separates the set of time series signals into segments that correspond to the individual iterations of the exercise profile. The method then aligns and merges the segments to generate a merged reference fingerprint. The method then trains a machine learning model to detect anomalous departures from the known undegraded state based on the merged reference fingerprint.