PENG WANG

Prof. PENG WANG


Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

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

Representative Publications

View All Publications
9 December 2019 in IEEE Transactions on Industry Applications
Verified
Journal Article
Closed access
ZZZhenyuan Zhang
Zhenyuan Zhang
HTHaoyue Tang
Haoyue Tang
PWPENG WANG
PENG WANG
Qi HuangQi Huang
Qi Huang
WLWei-Jen Lee
Wei-Jen Lee
With more distributed energy resources penetrated into the residential community, nanogrid based peer-to-peer (P2P) energy market has rapidly emerged over recent years. Due to the complexities on the decision-making process of each market participant, an efficient, fair and beneficial oriented bidding strategy is thus necessary. In this article, a two-stage bidding strategy for P2P trading of nanogrid is proposed. To overcome the limitations of traditional methods, in the first stage, a supply-demand relationship considered two-step price predictor, which aims to promote the usage of local renewable energy, is formulated to provide the guidance on transaction adjustment. In the second stage, trading preference based simultaneous game-theoretic approach is fully introduced, which can optimize the market equilibrium and then increase the social welfare of the P2P market. Additionally, to mitigate the possible failure of price matching, value-at-risk is implemented through the trading process as a risk hedging tool. To verify the effectiveness of the proposed strategy, usages of local renewable energy, economic benefits and success rates of transaction is compared against the traditional method for various cases.

26 April 2022 in IEEE Transactions on Industry Applications
Verified
Journal Article
Closed access
ZZZhenyuan Zhang
Zhenyuan Zhang
PZPengfei Zhao
Pengfei Zhao
PWPENG WANG
PENG WANG
WLWei-Jen Lee
Wei-Jen Lee
In recent years, a large variety of short-term load forecasting methodologies have been proposed, but a common drawback of them is that require a large amount of historical data to train the model. However, in reality, the cases in limited historical data would be more frequently occurred, for example, a new house or substation is built. To tackle this issue, a combining forecasting method of deep residual neural network (ResNet)-based transfer learning for residential loads is proposed in this article. To improve the accuracy and reliability of transfer learning, a novel deep ResNet with dual skip connections (ResNet-DSC) is proposed as the base model. Then with sparse training data, a Bayesian probability weighted averaging (approach is proposed, to address the model combination parameters estimation problem. In addition, probability density function forecasting is also delivered by the method. For demonstration, a series of experiments, with actual residential load data, was performed. With the comparison with other transfer learning and nontransfer learning approaches, the effectiveness and improvement of the proposed method have been validated.

4 February 2021 in IEEE Transactions on Industry Applications
Verified
Journal Article
Closed access
PWPENG WANG
PENG WANG
ZZZhenyuan Zhang
Zhenyuan Zhang
Qi HuangQi Huang
Qi Huang
XTXiaotian Tang
Xiaotian Tang
WLWei-Jen Lee
Wei-Jen Lee
Accurate equivalent model can be efficient to analyze the dynamic properties of active distribution network (ADN) as well as assess their impacts on stabilities of interconnected power system. However, due to the stochastic nature of renewable resources and time-varying configurations of load conditions, traditional ADN equivalent model may not be robust enough to different operation conditions. To overcome the limitations, this article proposed a robustness-improved method for dynamic equivalent modeling of ADN. To sketch out the most representative operation conditions of ADN, two-step clustering method with fisher discriminant analysis are used for grouping of operation conditions featured characteristic data sets. With the key parameter based identification technique applied, the multiple solution issue in parameter identification process could be effectively avoided. To further enhance the robustness of equivalent model, long short-term memory neural network is adopted to generalize the identified parameters. The performance of proposed modeling method is comprehensively evaluated by an actual ADN based verification cases.

30 January 2020 in IEEE Transactions on Industry Applications
Verified
Journal Article
Closed access
PWPENG WANG
PENG WANG
ZZZhenyuan Zhang
Zhenyuan Zhang
Qi HuangQi Huang
Qi Huang
WLWei-Jen Lee
Wei-Jen Lee
The uncertainty of power system is intensified by the integration of large-scale renewable energy resources such as Wind Farm (WF). Considering the impact of system uncertainties, the probabilistic stability analysis methods have been used in the stability assessments of power system with the WF integration. However, in traditional probabilistic analysis methods, WF is normally considered as PQ bus or aggregated as one single-machine equivalent model regardless of wake effect, which might reduce the accuracy of probabilistic stability studies. In this paper, a dynamic equivalent modeling method of WF for probabilistic stability assessments is proposed. The wake effect is considered in the modeling process and a practical four-machine clustering method is used in Wind Turbines (WTs) clustering. Besides, the Fisher Discriminant Analysis (FDA) is adopt to combine the similar WTs clustering results reasonably. Then, the WF is aggregated to a multi-machine equivalent model by capacity weighted method. Also, the established WF probabilistic model can be directly used in time-domain simulation based on Monte Carlo simulation and FDA. Finally, the efficiency of the proposed method is verified in an actual WF in China.

23 August 2023 in IEEE Transactions on Industry Applications
Verified
Journal Article
Closed access
PWPENG WANG
PENG WANG
ZZZhenyuan Zhang
Zhenyuan Zhang
CCChenxu Chen
Chenxu Chen
Qi HuangQi Huang
Qi Huang
NDNingyi DAI
Ningyi DAI
WLWei-Jen Lee
Wei-Jen Lee
With the high penetration of large-scale wind farms (WFs), accurate grid-connected WF models are crucial to be developed for studying the stability of power grids. Due to the large scale of the detailed model, a grid-connected WF is usually simplified into several aggregated wind turbine generators (WTGs) with the dynamic equivalent technique. However, identifying large amounts of parameters in an aggregated WTG may result in a multi-solution issue that reduces the accuracy and robustness of the developed models. Moreover, several aggregated WTG models which need to be identified simultaneously further complicate the issue. Therefore, to solve this issue, a dynamic equivalent modeling approach that could accurately estimate the full parameter of the equivalent WF model is developed. First, a WTG clustering approach is designed to simplify the WF model into several aggregated WTGs. Then, the external system of the aggregated WTG is simplified with the hybrid dynamic simulation (HDS) technique, by which the parameters of each aggregated WTG could be identified independently. Also, the WTG parameters are grouped with the trajectory sensitivity and identified using the well-designed multistage parameter identification approach. The simulation results in the modified IEEE 39-bus system demonstrated the effectiveness of the proposed method.

2 August 2021 in IEEE Geoscience and Remote Sensing Letters
Verified
Journal Article
Closed access
ZLZhiqiang Lan
Zhiqiang Lan
PGPeng Gao
Peng Gao
PWPENG WANG
PENG WANG
YWyaojun wang
yaojun wang
JLJiandong Liang
Jiandong Liang
GHGuangmin Hu
Guangmin Hu
It is crucial to retain as many signal components as possible, while noise is eliminated for wavelet packet noise reduction algorithms. Conventional thresholding functions take coefficients smaller than a given threshold as the noise component and get them removed in many ways. In this letter, considering the potential probability of being the signal-related component of coefficients whose value is smaller than the given threshold and the nonnegligible fact that microseismic (MS) event is sparse compared with the noise, we proposed a new wavelet packet-based denoising method via fuzzy partition. First, instead of a hard partition from a given threshold, wavelet packet coefficients get reduced by a fuzzy partition to retain more potential signal elements and suppress the noise. Second, we employ fuzzy c-means (FCM) clustering to identify the interval period of the MS event further to remove the residual noise and additional resonance in processed time series. We tested our method on synthetic datasets and real-field data from an MS monitoring experiment in a coal mine in Sichuan Basin, China. We utilize Pearson correlation coefficient and root-mean-square error between ideal signal and denoised data as performance indicators in synthetic tests, while sample entropy and kurtosis of denoised data are involved in the real field dataset. Test results from synthetic datasets and the real field dataset demonstrate that the proposed noise reduction method is superior to traditional hard-, soft-, and garrote-thresholding and is more applicable and effective in MS data processing.