一种新的智能故障诊断方法用于基于多维关注的变速箱,拒绝卷积
Wei Liu1, Zeqiao Zhang1, Zhiwei Ye1
1School of Computer Science, Hubei University of Technology, Wuhan, 430068, China.
Scientific reports
|October 21, 2024
概括
一个新的多维融合残留注意网络 (MFRANet) 通过抑制噪音和从工业振动数据中提取多尺度特征,有效地诊断变速箱故障. 这种深度学习方法提高了旋转机械故障诊断的可靠性和准确性.
科学领域:
- 人工智能的人工智能
- 机械工程 机械工程
- 信号处理 信号处理
背景情况:
- 卷积神经网络 (CNN) 在工业旋转机械故障诊断中与杂的多尺度振动数据作斗争.
- 有效的噪声抑制和多尺度特征提取对于准确的智能故障诊断至关重要.
- 现有的方法在处理现实世界的工业振动信号的复杂性方面面临挑战.
研究的目的:
- 提出一种新的深度神经网络框架,即多维融合残留注意网络 (MFRANet),用于变速箱故障诊断.
- 解决CNN在处理噪音和多尺度工业振动数据方面的局限性.
- 为了提高旋转机械故障诊断的准确性和稳定性.
主要方法:
- 开发了多维融合残留注意网络 (MFRANet),采用了多尺度深可分离卷积模块.
- 集成了一个剩余通道注意模块,用于特征地图无声化和权重.
- 采用外部注意模块来捕捉在多尺度特征中隐含的相关性.
主要成果:
- 与基准和最先进的方法相比,MFRANet在变速箱故障数据集上表现出优异的诊断性能.
- 拟议的方法在各种噪声级别中表现出强大的抗噪能力,表明提高了可靠性.
- 实验评估证实了MFRANet在准确提取多尺度断层特征方面的有效性.
结论:
- MFRANet为旋转机械的变速箱故障诊断提供了一个创新的,高效的解决方案.
- 该研究强调了先进的深度学习架构在解决实际工业挑战方面的潜力.
- 拟议的框架为要求苛刻的工业环境中智能故障诊断提供了一种可靠的方法.
相关概念视频
Transmission Shafts: Problem Solving
218
Designing a solid shaft that transmits power from a motor to a machine tool involves a series of calculations to ensure the shaft can withstand the stresses applied by bending moments and torques. First, calculate the torque exerted on the gear, considering the power transmitted by the shaft and its rotational speed. Following this, compute the tangential forces acting on the gears, which directly relate to the torque and the gear radius.
Next, use bending moment diagrams for the shaft to...
Next, use bending moment diagrams for the shaft to...
218
Multi-input and Multi-variable systems
100
Cruise control systems in cars are designed as multi-input systems to maintain a driver's desired speed while compensating for external disturbances such as changes in terrain. The block diagram for a cruise control system typically includes two main inputs: the desired speed set by the driver and any external disturbances, such as the incline of the road. By adjusting the engine throttle, the system maintains the vehicle's speed as close to the desired value as possible.
In the absence...
In the absence...
100
Time-Domain Interpretation of PD Control
84
Proportional-Derivative (PD) control is a widely used control method in various engineering systems to enhance stability and performance. In a system with only proportional control, common issues include high maximum overshoot and oscillation, observed in both the error signal and its rate of change. This behavior can be divided into three distinct phases: initial overshoot, subsequent undershoot, and gradual stabilization.
Consider the example of control of motor torque. Initially, a positive...
Consider the example of control of motor torque. Initially, a positive...
84
Relative Motion Analysis using Rotating Axes-Problem Solving
389
Consider a crane whose telescopic boom rotates with an angular velocity of 0.04 rad/s and angular acceleration of 0.02 rad/s2. Along with the rotation, the boom also extends linearly with a uniform speed of 5 m/s. The extension of the boom is measured at point D, which is measured with respect to the fixed point C on the other end of the boom. For the given instant, the distance between points C and D is 60 meters.
Here, in order to determine the magnitude of velocity and acceleration for point...
Here, in order to determine the magnitude of velocity and acceleration for point...
389
Design of Transmission Shafts - Stress Analysis
316
Designing a transmission shaft requires a thorough understanding of the stresses induced by bending moments and torques, especially in systems where power is transferred through gears. These forces create force-couple systems at the centers of the shaft's cross-sections, leading to both transverse and torsional loading. Although shearing stresses from transverse loads are typically smaller than those from torques and are often overlooked, the significant normal stresses from these loads...
316


