Real-Time Goertzel Algorithm for Rotating Machine Condition Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Condition Monitoring (CM) systems for rotating machines require high memory and processing power due to the need for large amounts of data to calculate spectral features, making them costly and inefficient, especially for real-time analysis.
Innovation Solution
The implementation of a computer-implemented method using real-time Goertzel Algorithm (GA) to compute spectral features from accumulator variables and supplemental variables, reducing the need for large memory and processing power by processing each data sample individually and applying a combination of real-time algorithms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If classical Fourier Transform methods are used to calculate spectral features, then measurement precision is improved, but device complexity and memory requirements increase
Solution Approach 1:
The patent extracts and applies only the necessary spectral analysis components through the Goertzel algorithm, which computes specific frequency bins from the DFT without calculating the entire spectrum. This selective extraction reduces computational complexity while maintaining the precision needed for bearing fault detection through envelope spectrum analysis.
Solution Approach 2:
The patent changes the computational approach from full FFT to the Goertzel algorithm, which is optimized for calculating individual frequency components. This parameter change in the mathematical method reduces the computational burden from O(N log N) to O(N) for specific frequency bins, thereby reducing device complexity while preserving measurement precision for critical frequencies.
2Measurement precision
If large amounts of data are collected for spectral analysis, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The patent implements preliminary action by continuously updating accumulator variables as data arrives, rather than waiting to collect large datasets before analysis. The Goertzel algorithm processes incoming vibration data in real-time, maintaining running sums and frequency components, which enables immediate detection of bearing faults without time-consuming batch processing.
Solution Approach 2:
The patent ensures continuity of useful action by implementing a continuous real-time processing pipeline where vibration data is constantly fed into the Goertzel algorithm, which continuously updates spectral features. This continuous action maintains measurement precision through ongoing analysis while eliminating idle time between data collection and analysis, achieving true real-time monitoring.
3Productivity
If high processing power is allocated for real-time spectral analysis, then productivity is improved, but device complexity and cost increase
Solution Approach 1:
The patent employs computationally lightweight algorithms (Goertzel algorithm with accumulator variables) that can be implemented on low-cost microcontrollers or embedded systems. Rather than requiring expensive high-performance computing hardware, the solution uses efficient mathematical methods that run quickly on inexpensive processors, making real-time bearing monitoring accessible in cost-sensitive applications.
Solution Approach 2:
The patent replaces complex mechanical signal processing systems with efficient mathematical algorithms. The Goertzel algorithm provides a mathematical substitution for hardware-based spectral analysis, achieving the same productivity (real-time analysis) with significantly reduced hardware complexity and cost by using software-based frequency domain computation.
Data Source
AI summary
A computer-implemented method and system for Condition Monitoring (CM) for rotating machines. The method and system include continuously receiving samples of the envelope of physical quantity data such as speed, vibration, or current, updating in real-time accumulator variables, computing in real-time spectral features based on the accumulator variables and supplemental variables, and determining a condition based on the real-time spectral features. The spectral features, exemplary as amplitudes at specific frequencies, are computed in real-time by a Goertzel Algorithm. The totality of the accumulator variables is sufficient to determine the condition of the rotating machine and the supplemental variables are temporarily needed for computing the spectral features. The one or more supplemental variables, such as memory addresses, are not based on the received samples of the input data.