Xu Fang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Xu Fang has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Sen Peng
School of Environmental Science and Engineering, Tianjin University, Tianjin 300350, China
3 shared publications - Qing Wu
School of Environmental Science and Engineering, Tianjin University, No. 135 Yaguan Road, Jinnan District, Tianjin, 300350, China
2 shared publications - Guolei Zheng
School of Environmental Science and Engineering, Tianjin University, Tianjin 300350, China
1 shared publication - Jing Cheng
School of Environmental Science and Engineering, Tianjin University, Tianjin 300350, China
1 shared publication - Xingqi Wu
School of Environmental Science and Engineering, Tianjin University, Tianjin 300350, China
1 shared publication - Xingqi Wu
School of Environmental Science and Engineering, Tianjin University, Tianjin 300350, China
1 shared publication
Latest Publications
- Open Access
Burst Diagnosis Multi-Stage Model for Water Distribution Networks Based on Deep Learning Algorithms
Pipe bursts in water distribution networks (WDNs) pose significant threats to the safety of distribution networks, driving attention to deep learning-based burst detection and localization. However, the applicability of different pressure features still needs to be compared and verified. A large number of nodes challenges deep learning with the excessive number of classification categories and low recognition accuracy. To address these problems, this paper extracts different burst pressure features, including pressure value, pressure difference, and pressure fluctuation ratio, and inputs one of these features into a Burst Diagnosis Multi-Stage Model (BDMM) based on three CS-LSTMs (a combination of the Cuckoo Search algorithm and a long short-term memory network). The first model addresses a binary classification problem, outputting labels indicating whether a pipe burst has occurred. The second one solves a multi-classification problem, outputting the label of the burst partition, and the third model also solves a multi-classification problem, outputting the ID of the bursting junction. The model is tested on a real network and outperforms ELM. For basic burst identification tasks using CS-LSTM, differences among the three features are minimal, while pressure difference and pressure fluctuation ratio exhibit superior performance to pressure value when resolving more complex problems like burst junction localization.
Open AccessBurst Diagnosis Multi-Stage Model for Water Distribution Networks Based on Deep Learning Algorithms
Pipe bursts in water distribution networks (WDNs) pose significant threats to the safety of distribution networks, driving attention to deep learning-based burst detection and localization. However, the applicability of different pressure features still needs to be compared and verified. A large number of nodes challenges deep learning with the excessive number of classification categories and low recognition accuracy. To address these problems, this paper extracts different burst pressure features, including pressure value, pressure difference, and pressure fluctuation ratio, and inputs one of these features into a Burst Diagnosis Multi-Stage Model (BDMM) based on three CS-LSTMs (a combination of the Cuckoo Search algorithm and a long short-term memory network). The first model addresses a binary classification problem, outputting labels indicating whether a pipe burst has occurred. The second one solves a multi-classification problem, outputting the label of the burst partition, and the third model also solves a multi-classification problem, outputting the ID of the bursting junction. The model is tested on a real network and outperforms ELM. For basic burst identification tasks using CS-LSTM, differences among the three features are minimal, while pressure difference and pressure fluctuation ratio exhibit superior performance to pressure value when resolving more complex problems like burst junction localization.
- Open Access
Leakage Detection in Water Distribution Networks Based on Multi-Feature Extraction from High-Frequency Pressure Data
Leakage detection is an important task to ensure the operational safety of water distribution networks. Leakage characteristic extraction based on high-frequency data has been widely used for leakage detection in experimental networks. However, the accuracy of single-feature-based methods is limited by the interference of background pressure fluctuations in networks. In addition, the setting of leakage diagnosis thresholds has been insufficiently studied, but influences leakage detection performance greatly. Hence, a new method of leakage detection is proposed based on multi-feature extraction. The multi-features of leakage are composed of instantaneous characteristics (ICs) and trend characteristics (TCs), which constitute comprehensive leakage information. The levels of the instantaneous and trend pressure drops in background pressure fluctuations in network environments are quantified for the setting of leakage diagnosis thresholds. In addition, ICs and TCs are used for leakage degree prediction. The proposed method was applied to an experimental network. Compared with the single-feature-based method and the cumulative sum (CUSUM) method, the proposed method achieved increases of 6.01% and 13.66% in F-Scores, respectively, and showed better adaptability to background pressure fluctuations in complex network environments.
Open AccessLeakage Detection in Water Distribution Networks Based on Multi-Feature Extraction from High-Frequency Pressure Data
Leakage detection is an important task to ensure the operational safety of water distribution networks. Leakage characteristic extraction based on high-frequency data has been widely used for leakage detection in experimental networks. However, the accuracy of single-feature-based methods is limited by the interference of background pressure fluctuations in networks. In addition, the setting of leakage diagnosis thresholds has been insufficiently studied, but influences leakage detection performance greatly. Hence, a new method of leakage detection is proposed based on multi-feature extraction. The multi-features of leakage are composed of instantaneous characteristics (ICs) and trend characteristics (TCs), which constitute comprehensive leakage information. The levels of the instantaneous and trend pressure drops in background pressure fluctuations in network environments are quantified for the setting of leakage diagnosis thresholds. In addition, ICs and TCs are used for leakage degree prediction. The proposed method was applied to an experimental network. Compared with the single-feature-based method and the cumulative sum (CUSUM) method, the proposed method achieved increases of 6.01% and 13.66% in F-Scores, respectively, and showed better adaptability to background pressure fluctuations in complex network environments.
- Open Access
Pressure Sensor Placement in Water Supply Network Based on Graph Neural Network Clustering Method
Pressure sensor placement is critical to system safety and operation optimization of water supply networks (WSNs). The majority of existing studies focuses on sensitivity or burst identification ability of monitoring systems based on certain specific operating conditions of WSNs, while nodal connectivity or long-term hydraulic fluctuation is not fully considered and analyzed. A new method of pressure sensor placement is proposed in this paper based on Graph Neural Networks. The method mainly consists of two steps: monitoring partition establishment and sensor placement. (1) Structural Deep Clustering Network algorithm is used for clustering analysis with the integration of complicated topological and hydraulic characteristics, and a WSN is divided into several monitoring partitions. (2) Then, sensor placement is carried out based on burst identification analysis, a quantitative metric named “indicator tensor” is developed to calculate hydraulic characteristics in time series, and the node with the maximum average partition perception rate is selected as the sensor in each monitoring partition. The results showed that the proposed method achieved a better monitoring scheme with more balanced distribution of sensors and higher coverage rate for pipe burst detection. This paper offers a new robust framework, which can be easily applied in the decision-making process of monitoring system establishment.
Open AccessPressure Sensor Placement in Water Supply Network Based on Graph Neural Network Clustering Method
Pressure sensor placement is critical to system safety and operation optimization of water supply networks (WSNs). The majority of existing studies focuses on sensitivity or burst identification ability of monitoring systems based on certain specific operating conditions of WSNs, while nodal connectivity or long-term hydraulic fluctuation is not fully considered and analyzed. A new method of pressure sensor placement is proposed in this paper based on Graph Neural Networks. The method mainly consists of two steps: monitoring partition establishment and sensor placement. (1) Structural Deep Clustering Network algorithm is used for clustering analysis with the integration of complicated topological and hydraulic characteristics, and a WSN is divided into several monitoring partitions. (2) Then, sensor placement is carried out based on burst identification analysis, a quantitative metric named “indicator tensor” is developed to calculate hydraulic characteristics in time series, and the node with the maximum average partition perception rate is selected as the sensor in each monitoring partition. The results showed that the proposed method achieved a better monitoring scheme with more balanced distribution of sensors and higher coverage rate for pipe burst detection. This paper offers a new robust framework, which can be easily applied in the decision-making process of monitoring system establishment.