研究目的
To propose and demonstrate experimentally a quantitative evaluation method of beam structure damage based on environmental random noise excitation and Laser Doppler Effect.
研究成果
The proposed method can diagnose the defects of the beam with high accuracy and high specificity, showing obvious advantages in the quantitative diagnosis of large structural beam defects compared with the traditional modal diagnosis method. It avoids the influence of the sensor on the measurement results and can be used for defect quantitative diagnosis of the special thin beam structure.
研究不足
The method's accuracy may be affected by environmental noise during the measurement of vibration using a non-contact laser. Additionally, the convolutional neural network requires large amounts of data during training, which may be impractical to provide enough input samples for specific working conditions.
1：Experimental Design and Method Selection:
The method involves obtaining vibration time-domain signals under random excitation of beam structure using Laser Doppler principle, followed by defect quantitative identification based on fast Fourier, continuous wavelet transform, and convolutional neural network.
2：Sample Selection and Data Sources:
Random vibration signals of steel beams with artificial defects are measured using Laser Doppler method.
3：List of Experimental Equipment and Materials:
Laser Doppler vibrator, random exciter, excitation amplifier, computer, simply supported beam.
4：Experimental Procedures and Operational Workflow:
The Laser Doppler Vibrometer scans the vibration velocity signal of each point on the beam point by point and saves it. The sampling frequency is set to 128 K.
5：Data Analysis Methods:
The vibration signal is subjected to fast Fourier transform, continuous wavelet transform, and then analyzed using a convolutional neural network for quantitative evaluation of beam defects.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
