Effective vibration recognition can improve the performance of vibration control and structural damage detection and is in high demand for signal processing and advanced classification.Signal-processing methods can ex...Effective vibration recognition can improve the performance of vibration control and structural damage detection and is in high demand for signal processing and advanced classification.Signal-processing methods can extract the potent time-frequency-domain characteristics of signals;however,the performance of conventional characteristics-based classification needs to be improved.Widely used deep learning algorithms(e.g.,convolutional neural networks(CNNs))can conduct classification by extracting high-dimensional data features,with outstanding performance.Hence,combining the advantages of signal processing and deep-learning algorithms can significantly enhance vibration recognition performance.A novel vibration recognition method based on signal processing and deep neural networks is proposed herein.First,environmental vibration signals are collected;then,signal processing is conducted to obtain the coefficient matrices of the time-frequency-domain characteristics using three typical algorithms:the wavelet transform,Hilbert-Huang transform,and Mel frequency cepstral coefficient extraction method.Subsequently,CNNs,long short-term memory(LSTM)networks,and combined deep CNN-LSTM networks are trained for vibration recognition,according to the time-frequencydomain characteristics.Finally,the performance of the trained deep neural networks is evaluated and validated.The results confirm the effectiveness of the proposed vibration recognition method combining signal preprocessing and deep learning.展开更多
Considering the dynamic influence of the roll vibration on the lubricant film thickness in the rolling deformation area,nonlinear dynamic rolling forces related to film thickness in the vertical and horizontal directi...Considering the dynamic influence of the roll vibration on the lubricant film thickness in the rolling deformation area,nonlinear dynamic rolling forces related to film thickness in the vertical and horizontal directions were obtained based on the Karman balance theory.Based on these dynamic rolling forces and the mechanical vibration of the rolling mill,a vertical-horizontal coupling nonlinear vibration dynamic model was established.The amplitude-frequency equation of the main resonance was derived by using the multiple-scale method.At last,the parameters of the 1780 rolling mill were used for numerical simulation,and the time-domain response curves of the system’s vibration displacement and lubricating film thickness under the steady and unsteady conditions were analyzed.The influences of parameters such as interface contact ratio,nonlinear parameters and external disturbances on the primary resonance frequency characteristics were obtained,which provided a theoretical reference for the suppression of rolling mill vibration.展开更多
In the inductively coupled data transmission system of the mooring buoy, the carrier signal frequency of the transmission channel is limited due to the inherent characteristics of the system, resulting in limited chan...In the inductively coupled data transmission system of the mooring buoy, the carrier signal frequency of the transmission channel is limited due to the inherent characteristics of the system, resulting in limited channel bandwidth. The limited channel bandwidth limits the increase in inductively coupled data transmission rate.In order to improve the inductively coupled data transmission rate of mooring buoy as much as possible without damaging the data transmission performance, a new method was proposed in this paper. The method is proposed to improve the data transmission rate by selecting the appropriate carrier signal frequencies based on the principle of maximizing the amplitude value of amplitude-frequency characteristic curve of the system. Research has been done according to this method as follows. Firstly, according to the inductively coupled transmission mooring buoy structure, the inductively coupled data transmission circuit model was established. The binary frequency shift keying(2FSK) digital signal modulation mode was selected. Through theoretical analysis, the relation between the carrier signal frequency and the data transmission performance, the relation between the carrier signal frequency and the 2FSK signal bandwidth were obtained. Secondly, the performance and the bandwidth of the signal transmission were studied for the inherent characteristics of the actual inductively coupled data transmission system. The amplitude-frequency characteristic of the system was analyzed by experiments. By selecting the appropriate carrier signal frequency parameters, an excellent data transmission performance was guaranteed and a large 2FSK signal bandwidth was obtained. Finally, an inductively coupled data transmission rate optimization experiment and a bit error rate analysis experiment were designed and carried out. The results show that the high-speed and reliable data transmission of the system was realized and the rate can reach 100 kbps.展开更多
文摘Effective vibration recognition can improve the performance of vibration control and structural damage detection and is in high demand for signal processing and advanced classification.Signal-processing methods can extract the potent time-frequency-domain characteristics of signals;however,the performance of conventional characteristics-based classification needs to be improved.Widely used deep learning algorithms(e.g.,convolutional neural networks(CNNs))can conduct classification by extracting high-dimensional data features,with outstanding performance.Hence,combining the advantages of signal processing and deep-learning algorithms can significantly enhance vibration recognition performance.A novel vibration recognition method based on signal processing and deep neural networks is proposed herein.First,environmental vibration signals are collected;then,signal processing is conducted to obtain the coefficient matrices of the time-frequency-domain characteristics using three typical algorithms:the wavelet transform,Hilbert-Huang transform,and Mel frequency cepstral coefficient extraction method.Subsequently,CNNs,long short-term memory(LSTM)networks,and combined deep CNN-LSTM networks are trained for vibration recognition,according to the time-frequencydomain characteristics.Finally,the performance of the trained deep neural networks is evaluated and validated.The results confirm the effectiveness of the proposed vibration recognition method combining signal preprocessing and deep learning.
基金This research is supported by the National Natural Science Foundation of China(Grant Nos.61973262 and 51405068)the Natural Science Foundation of Hebei Province of China(Grant No.E2019203146).
文摘Considering the dynamic influence of the roll vibration on the lubricant film thickness in the rolling deformation area,nonlinear dynamic rolling forces related to film thickness in the vertical and horizontal directions were obtained based on the Karman balance theory.Based on these dynamic rolling forces and the mechanical vibration of the rolling mill,a vertical-horizontal coupling nonlinear vibration dynamic model was established.The amplitude-frequency equation of the main resonance was derived by using the multiple-scale method.At last,the parameters of the 1780 rolling mill were used for numerical simulation,and the time-domain response curves of the system’s vibration displacement and lubricating film thickness under the steady and unsteady conditions were analyzed.The influences of parameters such as interface contact ratio,nonlinear parameters and external disturbances on the primary resonance frequency characteristics were obtained,which provided a theoretical reference for the suppression of rolling mill vibration.
基金supported by the National Natural Science Foundation of China [Grant number 61733012]Qingdao Ocean Engineering and Technology Think Tank Joint Fund Project [Grant number 20190131-2]the Shandong Provincial Natural Science Fund Project [Grant number ZR2017MEE072]。
文摘In the inductively coupled data transmission system of the mooring buoy, the carrier signal frequency of the transmission channel is limited due to the inherent characteristics of the system, resulting in limited channel bandwidth. The limited channel bandwidth limits the increase in inductively coupled data transmission rate.In order to improve the inductively coupled data transmission rate of mooring buoy as much as possible without damaging the data transmission performance, a new method was proposed in this paper. The method is proposed to improve the data transmission rate by selecting the appropriate carrier signal frequencies based on the principle of maximizing the amplitude value of amplitude-frequency characteristic curve of the system. Research has been done according to this method as follows. Firstly, according to the inductively coupled transmission mooring buoy structure, the inductively coupled data transmission circuit model was established. The binary frequency shift keying(2FSK) digital signal modulation mode was selected. Through theoretical analysis, the relation between the carrier signal frequency and the data transmission performance, the relation between the carrier signal frequency and the 2FSK signal bandwidth were obtained. Secondly, the performance and the bandwidth of the signal transmission were studied for the inherent characteristics of the actual inductively coupled data transmission system. The amplitude-frequency characteristic of the system was analyzed by experiments. By selecting the appropriate carrier signal frequency parameters, an excellent data transmission performance was guaranteed and a large 2FSK signal bandwidth was obtained. Finally, an inductively coupled data transmission rate optimization experiment and a bit error rate analysis experiment were designed and carried out. The results show that the high-speed and reliable data transmission of the system was realized and the rate can reach 100 kbps.
