Signal Processing Phase Difference Index for Vibration Diagnosis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In vibration diagnosis of rotating devices, existing methods require preprocessing with signal conditioners like PLL, tracking filters, or low-pass filters to obtain rotation pulse signals, which is inefficient and not directly applicable for calculating the state of the device using phase information.
Innovation Solution
A signal processing method that acquires time waveforms from sensors, generates frequency spectra, and calculates the difference between phases of signal components with the same frequency as an index to indicate the state of the object, eliminating the need for rotation pulse signals and additional filters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If preprocessing with signal conditioners (PLL, tracking filter, low-pass filter) is performed to obtain rotation pulse signals, then the rotation pulse signal can be extracted, but the device complexity and processing time increase
Solution Approach 1:
The patent extracts only the necessary vibration waveform component that is synchronized with the rotation pulse signal from the original vibration signal, using a tracking filter that follows the rotational frequency. This extracts the essential information while eliminating the need for complex preprocessing with multiple signal conditioners like PLL and low-pass filters.
Solution Approach 2:
The vibration diagnosis device can perform multiple functions: it can diagnose both rotating devices (by tracking rotational frequency) and non-rotating devices (by analyzing frequency spectra directly). The same core processing unit handles both cases, eliminating the need for separate rotation pulse signal extraction pathways and reducing overall device complexity.
2Measurement precision
If preprocessing with signal conditioners is performed to obtain rotation pulse signals, then the vibration waveform component can be extracted, but the processing time and operational complexity increase
Solution Approach 1:
The system performs preliminary frequency analysis to identify the rotational frequency component before extracting the synchronized vibration waveform. By pre-identifying the tracking frequency, the system avoids iterative searching and complex PLL operations, reducing processing time while maintaining extraction accuracy.
Solution Approach 2:
The patent replaces complex mechanical signal conditioning systems (multiple filters and PLL circuits) with a simplified digital signal processing approach using tracking filters and frequency spectrum analysis. This substitution reduces both processing time and operational complexity while maintaining the ability to extract synchronized vibration components.
3Measurement precision
If rotation pulse signals are required for vibration diagnosis, then synchronized vibration components can be obtained, but the adaptability to different device types is reduced
Solution Approach 1:
The vibration diagnosis device is designed to handle both rotating and non-rotating devices using the same core processing architecture. For rotating devices, it tracks the rotational frequency to extract synchronized components. For non-rotating devices, it performs general frequency spectrum analysis. This universal approach maintains measurement precision for synchronized components while extending adaptability to different device types.
Solution Approach 2:
The system dynamically adapts its processing mode based on the input signal characteristics. When rotational components are detected, it activates frequency tracking mode to extract synchronized vibration waveform components. When no rotational components are present, it switches to general spectral analysis mode. This dynamic adaptation maintains precision for rotating devices while enabling versatility for non-rotating devices.
Data Source
AI summary
A signal processing method includes a time waveform acquisition step of acquiring, from an i-th sensor, an i-th time waveform related to an i-th physical quantity generated by an external force, a velocity, or a displacement having at least a periodic variation acting on an object for each integer i of 1 or more and N or less with N being a predetermined integer of 1 or more, a frequency spectrum generation step of generating an i-th frequency spectrum for each integer i based on the i-th time waveform, and a first state index calculation step of calculating, for each integer i, a difference between a phase of a first signal component corresponding to a first peak included in a first frequency spectrum and a phase of a second signal component that corresponds to a second peak included in the i-th frequency spectrum and has a frequency that is a rational multiple of a frequency of the first signal component as an index indicating a state of the object.