Abstract:Mechanical vibration signal denoising is a pivotal task in various industrial applications, including system health monitoring and failure prediction. This paper introduces a novel deep learning transformer-based architecture specifically tailored for denoising mechanical vibration signals. The model leverages a Multi-Head Attention layer with 8 heads, processing input sequences of length 128, embedded into a 64-dimensional space. The architecture also incorporates Feed-Forward Neural Networks, Layer Normalization, and Residual Connections, resulting in enhanced recognition and extraction of essential features. Through a training process guided by the Mean Squared Error loss function and optimized using the Adam optimizer, the model demonstrates remarkable effectiveness in filtering out noise while preserving critical information related to mechanical vibrations. The specific design and choice of parameters offer a robust method adaptable to the complex nature of mechanical systems, with promising applications in industrial monitoring and maintenance. This work lays the groundwork for future exploration and optimization in the field of mechanical signal analysis and presents a significant step towards advanced and intelligent mechanical system diagnostics.

What problem does this paper attempt to address?

The paper primarily addresses the issue of denoising mechanical vibration signals and proposes a deep learning method based on the Transformer architecture. Specifically, the problems addressed in the paper can be summarized as follows: ### Research Background and Problem Definition - **Importance of System Health Monitoring (SHM)**: With the increasing reliance on complex systems in fields such as industrial manufacturing, healthcare, and transportation, system health monitoring has become crucial. It can track system performance in real-time, detect and predict anomalies or failures promptly, thereby reducing downtime and optimizing system lifecycle and efficiency. - **Characteristics and Challenges of Mechanical Vibration Signals**: Vibration signals from mechanical systems typically contain a significant amount of noise interference, which may originate from the environment, the equipment itself, or other factors. Therefore, extracting the true vibration information from noise-polluted signals is a challenging task. ### Solution - **Proposed Solution**: The paper proposes a novel deep learning model based on the Transformer architecture, specifically designed for the task of denoising mechanical vibration signals. - **Model Structure**: - **Multi-Head Attention Mechanism**: Utilizes a multi-head attention layer with 8 attention heads, capable of simultaneously focusing on different parts of the signal sequence to capture complex patterns and relationships in mechanical vibrations. - **Feedforward Neural Network**: Comprises two hidden layers, each with 64 units, and employs the ReLU activation function to introduce non-linear transformations. - **Residual Connections**: Implemented between the attention layer and the feedforward network layers, aiding in the training of deeper models. - **Layer Normalization**: Used to stabilize the training process and ensure numerical stability. - **Training and Evaluation**: - **Loss Function and Optimizer**: Uses Mean Squared Error (MSE) as the loss function and employs the Adam optimizer for training. - **Training Setup**: During training, data is divided into batches of 32 samples, with a total of 50 complete training epochs, and 20% of the data is set aside as a validation set to evaluate model performance. ### Main Contributions - Proposed a Transformer architecture specifically designed for denoising mechanical vibration signals, capable of removing noise while preserving critical information. - Experimentally validated the effectiveness and robustness of the proposed method, demonstrating its potential applications in complex mechanical systems. - Provided new directions and technical foundations for future research in the field of mechanical signal analysis and denoising. In summary, this paper aims to address the challenges in denoising mechanical vibration signals by developing an advanced deep learning method, thereby bringing significant improvements to fields such as industrial monitoring, maintenance, and diagnostics.

Transformer-Based Denoising of Mechanical Vibration Signals

A hybrid transformer masked time-domain denoising network for vibration signals

LN-MRSCAE: A Novel Deep Learning Based Denoising Method for Mechanical Vibration Signals

Neural-transformer: A Brain-Inspired Lightweight Mechanical Fault Diagnosis Method under Noise

Intelligent Vibration Signal Denoising Method Based on Non-Local Fully Convolutional Neural Network for Rolling Bearings.

Failure Mode Classification for Rolling Element Bearings Using Time-Domain Transformer-Based Encoder

Dynamic Temporal Denoise Neural Network with Multi-Head Attention for Fault Diagnosis Under Noise Background

ME-TransNet: A Modal-Enhanced Transformer Denoising Network for Attenuating Low-Frequency Swell Noise

Fault Diagnosis of Rotating Machinery under Noisy Environment Conditions Based on a 1-D Convolutional Autoencoder and 1-D Convolutional Neural Network

Recursive signal denoising method for predictive maintenance of equipment by using deep learning based temporal masking

Bearing Fault Vibration Signal Denoising Based on Adaptive Denoising Autoencoder

An innovative approach to vibration signal denoising and fault diagnosis using attention-enriched joint learning

Method for Denoising the Vibration Signal of Rotating Machinery through VMD and MODWPT

Denoising method of machine tool vibration signal based on variational mode decomposition and Whale-Tabu optimization algorithm

Elevator Vibration Signal Denoising by Deep Residual U-Net

Noise Reduction Method for Wind Turbine Gearbox Vibration Signals Based on CVMD-DRDSAE

Improved signal processing for bearing fault diagnosis in noisy environments using signal denoising, time–frequency transform, and deep learning

Structural Vibration Signal Denoising Using Stacking Ensemble of Hybrid CNN-RNN

Denoising Method of Vibration Signal for CNC Machine Tool Multi-Component Feed System Based on Joint Analysis Method

Transformative Noise Reduction: Leveraging a Transformer-Based Deep Network for Medical Image Denoising

Vibration Signal-Based Deep Noisy Filtering Model for Online Transformer Diagnosis