Shengli Liao

Dr. Shengli Liao


Research Focus & Metrics

Academic Career Length

14Years (2013-Present)

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

Representative Publications

View All Publications
17 August 2015 in Water
Verified
Journal Article
Open access
PDF file
Chuntian ChengChuntian Cheng
Chuntian Cheng
Zhong-Kai FengZhong-Kai Feng
Zhong-Kai Feng
WNWen-Jing Niu
Wen-Jing Niu
Shengli LiaoShengli Liao
Shengli Liao
Reservoir monthly inflow is rather important for the security of long-term reservoir operation and water resource management. The main goal of the present research is to develop forecasting models for the reservoir monthly inflow. In this paper, artificial neural networks (ANN) and support vector machine (SVM) are two basic heuristic forecasting methods, and genetic algorithm (GA) is employed to choose the parameters of the SVM. When forecasting the monthly inflow data series, both approaches are inclined to acquire relatively poor performances. Thus, based on the thought of refined prediction by model combination, a hybrid forecasting method involving a two-stage process is proposed to improve the forecast accuracy. In the hybrid method, the ANN and SVM are, first, respectively implemented to forecast the reservoir monthly inflow data. Then, the processed predictive values of both ANN and SVM are selected as the input variables of a newly-built ANN model for refined forecasting. Three models, ANN, SVM, and the hybrid method, are developed for the monthly inflow forecasting in Xinfengjiang reservoir with 71-year discharges from 1944 to 2014. The comparison of results reveal that three models have satisfactory performances in the Xinfengjiang reservoir monthly inflow prediction, and the hybrid method performs better than ANN and SVM in terms of five statistical indicators. Thus, the hybrid method is an efficient tool for the long-term operation and dispatching of Xinfengjiang reservoir.

2 October 2020 in IEEE Transactions on Power Systems
Verified
Journal Article
Closed access
ZZZhipeng Zhao
Zhipeng Zhao
Chuntian ChengChuntian Cheng
Chuntian Cheng
Shengli LiaoShengli Liao
Shengli Liao
YLYapeng Li
Yapeng Li
QLQuan Lu
Quan Lu
The irregular forbidden zone (FZ) is a common phenomenon with giant hydropower plants developed in the past two decades in China. Irregular shapes of FZs significantly challenge hydro unit commitment (HUC). This paper proposes a novel MILP based framework for HUC considering irregular FZ related constraints including the FZ constraint, effects of linearization errors in both the net head and the output, and the FZ crossing constraint. In the framework, the FZ constraint is handled by the optimal convex partitioning algorithm and the common structure-based formulation method. Inspired by the planar translating robot placement problem, linearization errors are considered by the Minkowski sum method. To handle the FZ crossing constraint, we then propose a graph theory-based approximate formulation method. The framework is integrated into a HUC model with an objective of peak shaving. The model is then tested with a batch of real-world instances of a cascade hydropower system formed by ten giant units with highly irregular FZs. The results show our framework can effectively consider the irregular FZ related constraints. The major advantage of our framework is its ability to handle the irregular shapes of FZs without any tedious and error-prone manual processing.

29 August 2022 in Energies
Verified
Journal Article
Open access
PDF file
Shengli LiaoShengli Liao
Shengli Liao
XTXudong Tian
Xudong Tian
Benxi LiuBenxi Liu
Benxi Liu
TLTian Liu
Tian Liu
HSHuaying Su
Huaying Su
BZBinbin Zhou
Binbin Zhou
With the expansion of wind power grid integration, the challenges of sharp fluctuations and high uncertainty in preparing the power grid day-ahead plan and short-term dispatching are magnified. These challenges can be overcome through accurate short-term wind power process prediction based on mining historical operation data and taking full advantage of meteorological forecast information. In this paper, adopting the ERA5 reanalysis dataset as input, a short-term wind power prediction framework is proposed, combining light gradient boosting machine (LightGBM), mutual information coefficient (MIC) and nonparametric regression. Primarily, the reanalysis data of ERA5 provide more meteorological information for the framework, which can help improve the model input features. Furthermore, MIC can identify effective feature subsets from massive feature sets that significantly affect the output, enabling concise understanding of the output. Moreover, LightGBM is a prediction method with a stronger ability of goodness-of-fit, which can fully mine the effective information of wind power historical operation data to improve the prediction accuracy. Eventually, nonparametric regression expands the process prediction to interval prediction, which significantly improves the utility of the prediction results. To quantitatively analyze the prediction results, five evaluation criteria are used, namely, the Pearson correlation coefficient (CORR), the root mean square error (RMSE), the mean absolute error (MAE), the index of agreement (IA) and Kling–Gupta efficiency (KGE). Compared with support vector regression (SVR), random forest (RF) and extreme gradient boosting (XGBoost) models, the present framework can make full use of meteorological information and effectively improve the prediction accuracy, and the generated output prediction interval can also be used to promote the safe operation of power systems.

11 August 2023 in Renewable Energy
Verified
Journal Article
Closed access
Benxi LiuBenxi Liu
Benxi Liu
TLTengyuan Liu
Tengyuan Liu
Shengli LiaoShengli Liao
Shengli Liao
JLJia Lu
Jia Lu
CCChuntian Cheng
Chuntian Cheng


6 March 2020 in IEEE Access
Verified
Journal Article
Open access
Benxi LiuBenxi Liu
Benxi Liu
JLJay R. Lund
Jay R. Lund
Shengli LiaoShengli Liao
Shengli Liao
XJXiaoyu Jin
Xiaoyu Jin
LLLingjun Liu
Lingjun Liu
CCChuntian Cheng
Chuntian Cheng
To meet the rapid growth of electricity demand and reduce carbon intensity, China is developing renewable energies rapidly including hydropower, wind and solar power. Due to the geographical mismatch of energy sources and demands in China, many long-distance and large-scale UHVDC and HVDC transmission projects have been built to transmit electric power from the western renewable bases to eastern coastal load centers. Some provincial power sources serve both local demands and deliver power to multiple regional power grids via HVDC transmission lines. As large capacity HVDC power transmission projects have great impacts on receiving-end power grids. Thus, the local exporting power grid should consider both local demands and energy importing area demands. A mixed-integer linear programming day-ahead peak shaving model to minimize the peak-valley difference in residual load after renewable generation of multiple power grids is developed. The model uses chance constraints to compensate for forecast errors of wind and solar power with hydropower, and introduces maximum daily power regulation times and stair-like power curve constraints of HVDC tie lines to avoid frequent HVDC power change and ensure power grid safety. The case studies in Yunnan province, which has large scale hydro, wind and solar power sources and delivers power to multiple regional power grids via HVDC transmission lines, shows the proposed model can shave peaks from multiple power grids effectively, hydropower can compensate for wind and solar forecast error and obtain satisfying results for multiple power grids, and that HVDC constraints can avoid their frequent power change and ensure the power grid safety.