Xiaosheng Peng

Prof. Dr. Xiaosheng Peng

State Key Laboratory of Advanced Electromagnetic Engineering and Technology, School of Electrical and Electronic Engineering, Huazhong University of Science and Technology, 430074 Wuhan, China

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

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

Representative Publications

View All Publications
19 March 2019 in IEEE Transactions on Power Delivery
Verified
Journal Article
Closed access
XPXiaosheng Peng
Xiaosheng Peng
FYFan Yang
Fan Yang
GWGanjun Wang
Ganjun Wang
YWYijiang Wu
Yijiang Wu
LLLee Li
Lee Li
WSWah Hoon Siew
Wah Hoon Siew
It is a great challenge to differentiate Partial Discharge (PD) induced by different types of insulation defects in high voltage cables. Some types of PD signals have very similar characteristics and are specifically difficult to be differentiate, even for the most experienced specialists. To overcome the challenge, a Convolutional Neural Network (CNN) based deep learning methodology for PD pattern recognition is presented in this paper. Firstly, PD testing for five types of artificial defects in Ethylene-Propylene-Rubber (EPR) cables was carried out in the High Voltage (HV) laboratory to generate signals containing PD data. Secondly, 3500 sets of PD transient pulses were extracted and then 33 kinds of PD features were established. The third stage applies a CNN to the data: typical CNN architecture and the key factors which affect the CNN based pattern recognition accuracy, are described. Factors discussed include the number of the network layers, convolutional kernel size, activation function, and pooling method. The paper presents a flowchart of the CNN based PD pattern recognition method and an evaluation with 3500 sets of PD samples. Finally, the CNN based pattern recognition results are shown and the proposed method is compared with two more traditional analysis methods, i.e. Support Vector Machine (SVM) and Back Propagation Neural Network (BPNN). The results show that the proposed CNN method has higher pattern recognition accuracy than SVM and BPNN, and that the novel method is especially effective for PD type recognition in cases of signals of high similarity, which is applicable for industrial applications.

22 May 2019 in IEEE Transactions on Power Delivery
Verified
Journal Article
Closed access
XPXiaosheng Peng
Xiaosheng Peng
JLJin-Shu Li
Jin-Shu Li
GWGanjun Wang
Ganjun Wang
YWYijiang Wu
Yijiang Wu
LLLee Li
Lee Li
WSWah Hoon Siew
Wah Hoon Siew
Optimal selection of features of Partial Discharge (PD) signals recorded from defects in High Voltage (HV) cables will contribute not only to the improvement of PD pattern recognition accuracy and efficiency but also to PD parameter visualization in HV cable condition monitoring and diagnostics. This paper presents a novel Random Forest (RF)-based feature selection algorithm for PD pattern recognition of HV cables. Firstly, the experimental data acquisition and feature extraction processes are introduced. PD signals were captured from faults created in a cable to obtain the raw PD data, then a set of 3500 transient PD pulses and a set of 3500 typical interference pulses were extracted, based on which 34 PD features were extracted for further processing. Furthermore, 119 two-dimensional features and 1082 three-dimensional features were generated. The paper then discusses the basic principle of the RF algorithm. Finally, RF-based feature selection was implemented to determine the optimal features for PD pattern recognition. The results were obtained and evaluated with the Back Propagation Neural Network (BPNN) and Support Vector Machine (SVM). Results show that the proposed RF-based method is effective for PD feature selection of HV cables with the potential for application to additional HV power apparatus.

15 March 2023 in IEEE Transactions on Power Systems
Verified
Journal Article
Closed access
ZYZimin Yang
Zimin Yang
XPXiaosheng Peng
Xiaosheng Peng
JSJiajiong Song
Jiajiong Song
RDRuiqin Duan
Ruiqin Duan
YJYan Jiang
Yan Jiang
SLShuangquan Liu
Shuangquan Liu
Sufficiently accurate short-term wind power prediction is important for the grid dispatch of the power system. To improve the accuracy by selecting suitable model for each piece of wind processes, this paper presents a short-term wind power prediction method based on multi-parameters similarity wind process matching and weighed-voting-based deep learning model selection. First, a novel multi-parameters similarity wind process matching method is presented to match each forecast target sample with groups of highly similar historical wind processes, in which each 96h-time-scale sample is divided into multiple wind processes by a tumbling time window, and a combinational similarity matching algorithm that consider four similarity indexes is proposed to judge the similarity among wind processes. Second, a weighed-voting-based deep learning model selection method, in which the matched highly similar historical wind processes are introduced to vote the optimal candidate deep learning model, is proposed to select the optimal model from LSTM, BLSTM, CNN, CNN-LSTM, CNN-BLSTM, and SDAE for each forecast target wind process. Case studies are presented to verify the effectiveness and superiority of the proposed method. Based on this new method, the 24h-day-ahead and 96h-short-term prediction RMSE can be decreased by 0.69% to 1.7% and 1.15% to 2.2% respectively compared to single deep learning model, which demonstrates the effectiveness of the proposed method.

24 August 2021 in IEEE Transactions on Industry Applications
Verified
Journal Article
Closed access
XPXiaosheng Peng
Xiaosheng Peng
YLYinhuan Li
Yinhuan Li
LDLie Dong
Lie Dong
KCKai Cheng
Kai Cheng
HWHongyu Wang
Hongyu Wang
WLWenze Li
Wenze Li
Wind power prediction (WPP) has an important impact on the security and reliability operation of power grid after a large amount of wind power integration into the system. There are two main challenges in WPP: 1) Both numerical weather prediction (NWP) and wind power contain abundant frequency information. If these data are directly input to the prediction model, the connection between different frequency bands is difficult to be mined. 2) Wind power has strong randomness and volatility, so the nonlinear relationship between input and output is difficult to be reflected by traditional prediction models. To overcome the challenge, a novel short-term WPP model based on wavelet feature arrangement and convolutional neural networks (CNN) is proposed in the article. First, wavelet transform is applied to split the original NWP data and historical power data into multiple sets of different frequency components. Then, the features of different frequencies are arranged in various ways, named feature arrangement (FA), which are input into the CNN model for WPP, and finally, the prediction results are obtained. Two case studies demonstrated the effectiveness of the proposed novel WT-FA-CNN deep learning model for short-term WPP.

4 May 2021 in IEEE Access
Verified
Journal Article
Open access
ZZZhendong Zhang
Zhendong Zhang
CWChao Wang
Chao Wang
XPXiaosheng Peng
Xiaosheng Peng
HQHui Qin
Hui Qin
HLHao Lv
Hao Lv
HWHongyu Wang
Hongyu Wang
Solar radiation intensity is intermittent and uncertain under the influence of meteorological conditions. Clustering them and obtaining high-precision and reliable probabilistic forecasting results play a vital role in the planning and management of solar power. In this study, a novel K-means time series clustering (K-MTSC) algorithm is first proposed to cluster solar radiation intensity and compared with astronomy method and K-means. Then, different feature inputs for different categories of solar radiation intensity are screened. Afterwards, the different kernel functions of Gaussian process regression (GPR) are compared and optimal kernel function is selected in terms of deterministic forecasting and probabilistic forecasting for different categories. Finally, the case study in Tibet province, China are performed to verify the validity and practicability of this research model and method. In this experiment, the average accuracy of GPR is 44% higher than that of Artificial Neural Network ANN, and 17% higher than that of Support Vector Regression. The experiments show that (1) the clustering results obtained by the K-MTSC algorithm have a larger inter-group distance and a smaller intra-group distance, and at the same time, it will not destroy the continuity of the time series. (2) The probability forecast results obtained by GPR are reliable and high-accuracy.

26 May 2021 in Energies
Verified
Journal Article
Open access
PDF file
TCTao Cai
Tao Cai
LJlu junjie
lu junjie
JLJianxun Lang
Jianxun Lang
XPXiaosheng Peng
Xiaosheng Peng
KCKai Cheng
Kai Cheng
SDShanxu Duan
Shanxu Duan
In this paper, a hybrid model that considers both accuracy and efficiency is proposed to predict photovoltaic (PV) power generation. To achieve this, improved forward feature selection is applied to obtain the optimal feature set, which aims to remove redundant information and obtain related features, resulting in a significant improvement in forecasting accuracy and efficiency. The prediction error is irregularly distributed. Thus, a bias compensation–long short-term memory (BC–LSTM) network is proposed to minimize the prediction error. The experimental results show that the new feature selection method can improve the prediction accuracy by 0.6% and the calculation efficiency by 20% compared to using feature importance identification based on LightGBM. The BC–LSTM network can improve accuracy by 0.3% using about twice the time compared with the LSTM network, and the hybrid model can further improve prediction accuracy and efficiency based on the BC–LSTM network.