跳转到主要内容

Research & Achievements

Highlights of my academic research and intellectual property

Academic Papers

Rainfall Trends and Multi-Scale Variability in the Receiving Areas of the Zhedong Water Diversion Project

Water Resources and Power, 2025

To reveal rainfall variation patterns in the receiving areas of the Zhedong Water Diversion Project and optimize project operations, this study analyzes long-term daily rainfall data from 1961 to 2022 across 15 typical sub-regions affected by the project. Using methods including the Mann-Kendall trend test, Sen's slope estimator, and multi-scale sliding-window analysis, the study systematically examines the spatiotemporal distribution, long-term trends, multi-scale variability, and inter-regional correlation of rainfall. Findings show that all sub-regions exhibit a significant upward trend in annual rainfall (p<0.05), and the Hurst index analysis (H>0.5) indicates that this upward trend is persistent. However, the magnitude of increase varies significantly in space (coastal > river basin > hilly), and inter-annual fluctuations have intensified after 2010. Rainfall characteristics show significant scale dependence: short timescales (3 months) exhibit high variability and strong spatial heterogeneity, while longer timescales (12 months) are more stable with greater regional synergy. Inter-regional rainfall correlation generally strengthens as the analysis timescale increases, although uncertainty remains in the trend magnitudes estimated by Sen's slope. This study provides a refined characterization of long-term and multi-scale rainfall variations in the study area, offering important scientific evidence for adaptive scheduling of the Zhedong Water Diversion Project and optimal regional water resources management.

Dynamic Water Demand Forecasting in Zhejiang Province Based on Major Driving Factor Selection and Deep Learning Algorithms

Journal of Hydraulic Engineering, 2024

This study collected water demand data for various water-use industries in Zhejiang Province from 2000 to 2020. A Spearman rank correlation-based major driving factor selection method was used to identify the principal factors affecting industry water demand. An improved Long Short-Term Memory (LSTM) neural network model was then constructed to perform dynamic rolling forecasts of water demand for each industry. The forecasting results of the improved LSTM model were compared with those of a traditional univariate LSTM model, a convolutional neural network model, and a support vector regression model. Results show that the LSTM model improved with major driving factor selection can perform real-time dynamic rolling forecasts of annual water demand for each industry, with higher accuracy than the other three models.

Non-Uniform Sediment Transport in River Networks Considering Lateral Sediment Outflow

Journal of Hydraulic Engineering, 2016

Diverting water into a river network can rapidly improve water quality but also causes sediment deposition. As flow advances along main channels, it carries sediment that disperses into connected tributaries. Therefore, lateral sediment outflow must be considered when computing sediment transport in river networks. Based on year-round real-time water level-discharge and daily sediment field measurements at multiple stations in the western Hangzhou river network, the authors established a logarithmic non-uniform sediment-carrying capacity formula applicable to river networks. By incorporating lateral outflow, a method for computing non-uniform sediment transport in river networks that accounts for time-varying lateral sediment outflow was derived for the first time, and the one-dimensional river-network sediment transport differential equation and discrete computational scheme were improved accordingly. Validation against measured data shows good agreement between computed and observed values. Sediment transport calculations for the western Hangzhou river network show that lateral sediment outflow accounts for approximately 11% of total deposition, indicating that ignoring lateral sediment outflow in river-network models would introduce significant errors.

A Depth-Averaged 2D Physically-Based Model for Flow and Sediment Transport in Open Channels

ASCE Environmental and Water Resources Congress, 2019

This paper presents a novel depth-averaged two-dimensional model for simulating flow and sediment transport in open channels. The model incorporates physically-based parameters and has been validated using field measurements from various river systems. The research was conducted during my joint Ph.D. training at Clarkson University under the supervision of Prof. Weiming Wu.

Invention Patents

A Depth-Adjustable Water Intake and Aeration Device

发明年份: 2022

This utility model discloses a depth-adjustable water intake and aeration device, including an energy module, a water pump, an outflow pipeline, a floating platform, a telescopic structure, and a three-way water intake pipeline. The outflow pipeline and the three-way water intake pipeline are connected to the outlet and inlet of the water pump respectively. The energy module is mounted on the floating platform, which is integrated with the outflow pipeline as a single unit. The floating platform is a cubic platform cast from lightweight materials, and the outflow pipeline is enveloped by it. The device has a simple structure that effectively promotes circulation and mixing of water bodies at different depths. It can float on the water surface and is equipped with an energy module, ensuring the recyclability of the device to a certain extent and effectively improving the effectiveness of water intake and aeration.

A Multi-Model-Fusion-Based System and Method for Rainfall Data Analysis and Visualization

发明年份: 2025

This invention discloses a multi-model-fusion-based system and method for rainfall data analysis and visualization, comprising the following steps: S1, data preparation and preprocessing — collect and integrate rainfall observation data, then perform quality control and standardization; S2, multi-model analysis — apply multiple algorithms including the Mann-Kendall test, Sen's slope estimator, and sliding-window analysis to reveal trends, magnitudes, and multi-scale characteristics in rainfall data; S3, inter-regional correlation analysis — compute rainfall correlations across regions and multiple scales, and identify regions with similar patterns; S4, visualization and interactive display — present analysis results intuitively through diverse charts, with an interactive interface supporting exploration. By integrating various analysis models and visualization techniques, the invention solves the existing technology's shortcomings in comprehensiveness, multi-scale analysis, and visualization variety in rainfall data analysis. It provides more comprehensive, in-depth, and multi-dimensional insights into rainfall characteristics, enhancing the intuitiveness of result presentation and the efficiency of analysis.

Software Copyrights

ZSDI Water Resources Optimal Scheduling Model Software

A water resources scheduling system based on multi-objective optimization algorithms, supporting real-time decision-making and scenario evaluation to improve water resources utilization efficiency.

ZSDI Water Resources Carrying Capacity Model Software

A water resources carrying capacity assessment system integrating multi-dimensional evaluation indicators, supporting dynamic monitoring and early warning to provide a scientific basis for regional water resources management.

ZSDI Excel-to-MIKE Intelligent Data Conversion Software

A professional intelligent data format conversion tool with batch processing, substantially improving data preparation efficiency for water conservancy models.

ZSDI Cascade Reservoir Group Multi-Objective Joint Scheduling Model V1.0

A multi-reservoir, multi-period joint scheduling system developed in Python and PyQt, using a water-balance iterative algorithm to support multi-objective optimal scheduling of cascade reservoir groups.

ZSDI Water Efficiency Assessment and Analysis System

A water efficiency assessment tool for industrial agglomeration zones based on AHP+CRITIC combined weighting and TOPSIS comprehensive ranking, supporting a three-tier park → pipeline → enterprise assessment. Streamlit web version and Tauri desktop version deployed in sync.

ZSDI Water Environment Functional Zone Pollutant Assimilation Capacity Calculation System

A pollutant assimilation capacity calculation tool for water environment functional zones based on the river channel decay formula, supporting multi-scenario comparison, tributary segmentation, and monthly calculation. Excel import/export, with Streamlit web version and Tauri desktop version deployed in sync.

Honors and Awards