Zehua Du

Dr. Zehua Du


Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
20 May 2021 in IEEE Access
Verified
Journal Article
Open access
DJDongning Jia
Dongning Jia
YLYunxin Li
Yunxin Li
ZDZehua Du
Zehua Du
JXJiali XU
Jiali XU
BYBo Yin
Bo Yin
Non-intrusive load monitoring (NILM) is crucial because it helps monitor the operating status of electrical appliances online; detailed power consumption data regarding the appliances can then be obtained. However, the identification of resistive appliances that have similar features in a power grid is still a major problem. In this study, the reconstructed image of a voltage-current (VI) trajectory is used as input data for a convolutional neural network (CNN) to classify the appliances, particularly resistive appliances. Two dataset PLAID and IDOUC are introduced to verify the performance of the proposed method. According to the results, the excellent performance of the reconstructed VI image method for the identification of the household appliances with similar waveform is validated by comparing it with the other two methods.

8 July 2020 in IEEE Access
Verified
Journal Article
Open access
XDXifeng Dong
Xifeng Dong
BYBo Yin
Bo Yin
YCYanping Cong
Yanping Cong
ZDZehua Du
Zehua Du
XHXianqing Huang
Xianqing Huang
In recent years, with the construction of intelligent cities, the importance of environmental sound classification (ESC) research has become increasingly prominent. However, due to the non-stationary nature of environment sound and the strong interference of ambient noise, the recognition accuracy of ESC is not high enough. Even with deep learning methods, it is difficult to fully extract features from models with a single input. Aiming to improve the recognition accuracy of ESC, this paper proposes a two-stream convolutional neural network (CNN) based on raw audio CNN (RACNN) and logmel CNN (LMCNN). In this method, a pre-emphasis module is first constructed to deal with raw audio signal. The processed audio data and logmel data are imported into RACNN and LMCNN, respectively to obtain both of time and frequency features of audio. In addition, a random-padding method is proposed to patch shorter data sequences. In such a way, the available data for experiment are greatly increased. Finally, the effectiveness of the methods has been verified based on UrbanSound8K dataset in experimental part.

24 December 2020 in IEEE Access
Verified
Journal Article
Open access
KLKexin Li
Kexin Li
BYBo Yin
Bo Yin
ZDZehua Du
Zehua Du
YSYufei Sun
Yufei Sun
Nonintrusive load monitoring (NILM) plays a key role in the real-time electricity consumption monitoring of household appliances. However, it is difficult to realize high precision load identification by using a single waveform feature. Therefore, this article proposes a two-stream convolutional neural network based on current time-frequency feature fusion for nonintrusive load identification. First, a time series image coding method for current time-frequency multi-feature fusion is proposed. The method can extract the time domain and frequency domain features of the current timing signal effectively. Then, we present a two-stream neural network combining the gated recurrent unit (GRU) and a two-dimensional convolutional neural network (2D-CNN) to improve the load identification performance. Finally, the experimental results on the PLAID and IDOUC datasets show that the proposed model outperforms the state-of-the-art methods.

3 November 2020 in Sensors
Verified
Journal Article
Open access
PDF file
RLRuixue Li
Ruixue Li
BYBo Yin
Bo Yin
YCYanping Cong
Yanping Cong
ZDZehua Du
Zehua Du
Soil nutrient prediction based on near-infrared spectroscopy has become the main research direction for rapid acquisition of soil information. The development of deep learning has greatly improved the prediction accuracy of traditional modeling methods. In view of the low efficiency and low accuracy of current soil prediction models, this paper proposes a soil multi-attribute intelligent prediction method based on convolutional neural networks, by constructing a dual-stream convolutional neural network model Multi_CNN that combines one-dimensional convolution and two-dimensional convolution, the intelligent prediction of soil multi-attribute is realized. The model extracts the characteristics of soil attributes from spectral sequences and spectrograms respectively, and multiple attributes can be predicted simultaneously by feature fusion. The model is based on two different-scale soil near-infrared spectroscopy data sets for multi-attribute prediction. The experimental results show that the RP2 of the three attributes of Total Carbon, Total Nitrogen, and Alkaline Nitrogen on the small dataset are 0.94, 0.95, 0.87, respectively, and the RP2 of the attributes of Organic Carbon, Nitrogen, and Clay on the LUCAS dataset are, respectively, 0.95, 0.91, 0.83, And compared with traditional regression models and new prediction methods commonly used in soil nutrient prediction, the multi-task model proposed in this paper is more accurate.

9 July 2024 in IEEE Internet of Things Journal
Verified
Journal Article
Closed access
SLShuang Li
Shuang Li
SLShuai Liu
Shuai Liu
JWJingjing Wang
Jingjing Wang
SYShefeng Yan
Shefeng Yan
JLJiahao Liu
Jiahao Liu
ZDZehua Du
Zehua Du
With little or no human intervention, almost every object in Internet of Things (IoT) has ability to communicate, sense, and process information to make everything connected. Noise in complex environment has a great impact on machine-to-machine (M2M) interaction in IoT. Adaptive wavelet threshold function (AWTF)-based M2M Gaussian Noise Removal Method is proposed in this article. First, a bilateral enhanced wavelet threshold function is derived based on the adjustable zeroing window. Further, threshold is used to construct a bilateral enhanced wavelet threshold function, which can eliminate the oscillations in the existing wavelet threshold function. This ensures that there is no break in wavelet coefficients during the reconstruction process and that stable wavelet decomposition and reconstruction can be achieved. The signal-to-noise ratio (SNR) of a signal is estimated based on the variance of the noise-containing signal, and the zeroing window parameters are adjusted adaptively according to the SNR value to eliminate the noisy wavelet coefficients and improve the denoising performance. In addition, when the wavelet coefficients are reconstructed, the proposed algorithm can select a suitable threshold function according to a particular threshold value, which improves the robustness of the algorithm. “Doppler” and “Bumps” standard test signals are used as interactive signals to simulate the proposed algorithms. For “Doppler” signals, the SNR, root mean square error (RMSE), and noise suppression ratio (NSR) of AWTF increased by 8.48%, 42.59%, and 1.69%, respectively, compared with the GDES+ABC algorithm. For “Bumps” signal, the SNR, RMSE, and NSR of AWTF are improved by 11.67%, 24.46%, and 2.99%, respectively, compared with GDES+ABC algorithm. In addition, we also use different modulated signals to carry out real field experiments in Qingdao cruise ship home port, which proves the effectiveness of the proposed algorithm.

21 May 2021 in Neural Computing and Applications
Verified
Journal Article
Open access
PDF file
BYBo Yin
Bo Yin
ZLZhenhuan Li
Zhenhuan Li
JXJiali Xu
Jiali Xu
LLLin Li
Lin Li
XYXinghai Yang
Xinghai Yang
ZDZehua Du
Zehua Du
The construction of smart grid is an important part of improving the utilization rate of electric energy. As an important way for the construction of smart grid, non-intrusive load decomposition methods have been extensively studied. In this type of method, limited by transmission cost and network bandwidth, low-frequency data has been widely used in practical applications. However, the accuracy of device identification in this case faces challenges. Due to the relatively single characteristics of low-frequency data, it is difficult to express the operating status of complex electrical appliances, resulting in low decomposition performance. In this paper, a non-intrusive load is proposed based on household electrical habits by studying the relationship between household electricity consumption habits and load status decomposition method. The Gaussian mixture model and time information are used to model the probability distribution of the electrical appliance state. This probability distribution is then used as the observation probability distribution of the factor hidden Markov model. In such a way, the BH-FHMM model is proposed. Finally, load decomposition is carried out through the load decomposition process of the FHMM model. In order to verify the performance of the proposed method, an experimental comparison is conducted based on the REDD data set. According to the results, a significant improvement in equipment recognition accuracy is obtained.