GIS Mechanical Fault Diagnosis Using Wavelet Entropy and BP Neural Nets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing GIS mechanical fault diagnosis methods fail to effectively detect low-frequency mechanical signals and accurately diagnose faults in operating mechanisms due to the lack of comprehensive data collection and integration of mathematical tools, leading to low effectiveness and reliability.
Innovation Solution
Collect vibration signals from various excitation sources, perform wavelet packet-feature entropy vector extraction, and input the extracted vectors into a pre-trained BP neural network for GIS mechanical fault identification, using formulas to determine abnormal signals and employing wavelet soft threshold denoising and Daubechies wavelet series for decomposition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional mechanical state signal detection methods are used, then the detection is simple and fast, but low-frequency mechanical signals cannot be effectively detected and fault detection accuracy is low
Solution Approach 1:
The patent segments the vibration signal analysis into multiple frequency bands using wavelet packet decomposition. The signal is divided into different resolution levels, allowing separate analysis of low-frequency and high-frequency components. This segmentation enables effective detection of low-frequency mechanical signals that were previously missed by traditional single-band detection methods.
Solution Approach 2:
The patent transforms the vibration signal from the time domain to the frequency domain through wavelet packet transformation. This dimensional change allows the system to analyze signal characteristics in the frequency dimension, particularly enhancing the visibility and detectability of low-frequency components that are imperceptible in the time domain using traditional methods.
2Reliability
If comprehensive vibration signal collection from multiple excitation sources is performed, then fault diagnosis comprehensiveness is improved, but data processing complexity increases
Solution Approach 1:
The patent merges vibration signals from multiple excitation sources (electromagnetic force, operating mechanism, etc.) into a unified analysis framework. By combining these diverse signal sources and processing them through the same wavelet packet-feature entropy-BP neural network system, the method achieves comprehensive fault diagnosis while managing data processing complexity through standardized procedures.
Solution Approach 2:
The patent introduces wavelet packet-feature entropy as an intermediary that bridges the gap between raw multi-source vibration signals and the BP neural network classifier. This intermediary transforms complex multi-dimensional signal data into standardized feature vectors, simplifying the processing burden on the neural network while preserving essential fault information from all excitation sources.
3Measurement precision
If wavelet packet-feature entropy vector extraction is performed on abnormal vibration signals, then fault identification accuracy is improved, but computational time increases
Solution Approach 1:
The patent performs preliminary wavelet packet decomposition and feature entropy calculation on vibration signals during the normal operation phase. By pre-processing and extracting feature vectors in advance, the system reduces the computational burden during actual fault diagnosis, achieving high identification accuracy without excessive real-time computational delays.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The method comprehensively diagnoses GIS mechanical faults by integrating vibration signals from multiple sources, enhancing the accuracy and reliability of fault detection.
Implementation Method 1
performing wavelet packet decomposition by using db10 of Daubechies wavelet series as wavelet base function
Implementation Method 2
extracting the envelope of the vibration signal excited by the operating mechanism to obtain the envelope area by using Hilbert method
Implementation Method 3
extracting the vibration energy of the vibration signal excited by the electromagnetic force at the set frequency by using the FFT method
Implementation Method 4
inputting the extracted wavelet packet-feature entropy vectors into the pre-trained BP neural network for GIS mechanical fault identification
Data Source
AI summary
A GIS mechanical fault diagnosis method and the device are disclosed. The method includes: collecting vibration signals to be measured of various excitation sources of GIS in mechanical operation; performing wavelet packet-feature entropy vector extraction on the vibration signals to be measured, when it is determined that the vibration signals to be measured are abnormal according to standard vibration signals in the normal state; inputting the extracted wavelet packet-feature entropy vectors into the pre-trained BP neural network for GIS mechanical fault identification, and outputting the corresponding fault. The disclosure integrates the vibration signals under the action of various excitation sources, extracts the feature entropy vectors according to the entropy theory, and constructs and trains a BP neural network that can classify and recognize various GIS mechanical faults, so as to perform comprehensive and effective GIS mechanical faults diagnose.