多目标约束下给水管网水质传感器选址优化算法的研究

利用EPANET2模拟扩展周期非稳态水力水质条件下管网节点污染物浓度变化,根据各个节点被注入污染物后,在管网模拟结束时得到的选址目标值的大小来确定节点有可能作为污染物注入的节点,目标值越大,该节点被选择的可能性越大.另外,本文提出了基本粒子群算法与遗传算法交叉、变异算子相结合的整数编码的优化算法来求解水质传感器选址问题,并编制了相应的计算程序.文中结合算例,以经过归一化后的节点平均坐标作为衡量选址结果的指标,得到了不同污染物注入开始时刻、注入持续时问和质量注入速率条件下传感器选址节点平均坐标的累计分布函数图,为传感器的选址提供参考.     

Forecasting Surface Water Level Fluctuations of Lake Van by Artificial Neural Networks

Lake Van in eastern Turkey has been subject to water level rise during the last decade and, consequently, the low-lying areas along the shore are inundated, giving problems to local administrators, governmental officials, irrigation activities and to people's property. Therefore, forecasting water levels of the Lake has started to attract the attention of the researchers in the country. An attempt has been made to use artificial neural networks (ANN) for modeling the temporal change water levels of Lake Van. A back-propagation algorithm is used for training. The study indicated that neural networks can successfully model the complex relationship between the rainfall and consecutive water levels. Three different cases were considered with the network trained for different arrangements of input nodes, such as current and antecedent lake levels, rainfall amounts. All of the three models yields relatively close results to each other. The neural network model is simpler and more reliable than the conventional methods such as autoregressive (AR), moving average (MA), and autoregressive moving average with exogenous input (ARMAX) models. It is shown that the relative errors for these two different models, are below 10% which is acceptable for engineering studies. In this study, dynamic changes of the lake level are evaluated. In contrast to classical methods, ANNs do not require strict assumptions such as linearity, normality, homoscadacity etc.     

排水体系建设对城市洪涝灾害的影响

地下管网、水系等是城市重要的排水体系,直接决定着城市的洪涝安全。以武汉市光谷中心城为研究区域,采用一、二维耦合的水动力数学模型,进行了不同排水体系组合条件下的多方案城市雨洪模拟,量化分析了地下管网、规划水系建设对城市洪涝灾害的消减效果。结果表明:光谷中心城现状未形成成熟排水体系时,其内涝范围、淹没面积及受灾损失均最大,规划排水体系建设后,区域洪涝淹没面积和受灾损失消减率分别达56.58%,63.74%。由于作用范围及作用方式不同,单独建设水系对区域排涝的作用低于单独建设地下管网的作用。对不同土地利用类型比较,住宅用地减灾效果对城市排水体系的建设最为敏感。当下游湖泊水位较高时,湖泊对水系存在顶托作用,其对区域洪涝灾害影响的程度与河道纵向比降、河堤高程,以及地下管网和河道的连接状况有关。     

Cascade vulnerability for risk analysis of water infrastructure

One of the major tasks in urban water management is failure-free operation for at least most of the time. Accordingly, the reliability of the network systems in urban water management has a crucial role. The failure of a component in these systems impacts potable water distribution and urban drainage. Therefore, water distribution and urban drainage systems are categorized as critical infrastructure. Vulnerability is the degree to which a system is likely to experience harm induced by perturbation or stress. However, for risk assessment, we usually assume that events and failures are singular and independent, i.e. several simultaneous events and cascading events are unconsidered. Although failures can be causally linked, a simultaneous consideration in risk analysis is hardly considered. To close this gap, this work introduces the term cascade vulnerability for water infrastructure. Cascade vulnerability accounts for cascading and simultaneous events. Following this definition, cascade risk maps are a merger of hazard and cascade vulnerability maps. In this work cascade vulnerability maps for water distribution systems and urban drainage systems based on the 'Achilles-Approach' are introduced and discussed. It is shown, that neglecting cascading effects results in significant underestimation of risk scenarios.     

Lake Level Prediction using Feed Forward and Recurrent Neural Networks

The protection of high quality fresh water in times of global climate changes is of tremendous importance since it is the key factor of local demographic and economic development. One such fresh water source is Vrana Lake, located on the completely karstified Island of Cres in Croatia. Over the last few decades a severe and dangerous decrease of the lake level has been documented. In order to develop a reliable lake level prediction, the application of the artificial neural networks (ANN) was used for the first time. The paper proposes time-series forecasting models based on the monthly measurements of the lake level during the last 38 years, capable to predict 6 or 12 months ahead. In order to gain the best possible model performance, the forecasting models were built using two types of ANN: the Long-Short Term Memory (LSTM) recurrent neural network (RNN), and the feed forward neural network (FFNN). Instead of classic lagged data set, the proposed models were trained with the set of sequences with different length created from the time series data. The models were trained with the same set of the training parameters in order to establish the same conditions for the performance analysis. Based on root mean squared error (RMSE) and correlation coefficient (R) the performance analysis shown that both model types can achieve satisfactory results. The analysis also revealed that regardless of the model types, they outperform classic ANN models based on datasets with fixed number of features and one month the prediction period. Analysis also revealed that the proposed models outperform classic time series forecasting models based on ARIMA and other similar methods .

     

Correction of Reservoir Runoff Forecast Based on Multi-scenario Division and Multi Models

Accurate runoff forecast is very important for reservoir operation. In view of the shortcomings of the existing correction models for runoff forecast, including the influence of the difference of external factors on the forecast results is not considered, and the optimal situation adaptation of different forecast models is not considered, three models, i.e., long and short-term memory neural network model (LSTM), gaussian process regression model (GPR) and support vector machine regression model (SVR), are used to forecast the relative errors of runoff forecast under different scenarios in this paper. The classification of forecast scenarios is determined based on factors such as rainfall, inflow, and foresight period, and two scenario sets are given, i.e., 12 forecast scenarios and 24 forecast scenarios. Then, a multi-model coupled runoff forecast correction method considering forecast error and forecast scenario is proposed. Through the case study of the Three Gorges Reservoir (TGR), it is found that, when the analysis is carried out based on the forecast period, the SVR model should be used for forecast correction when the foresight period is 1–5 days, and the LSTM model should be used for forecast correction when the foresight period is 6 days. The application effect of SVR and LSTM is better than GPR in the scenario set of 12 forecast scenarios. LSTM model has the highest accuracy of forecast correction in the scenario set of 24 forecast scenarios, and the mean value of the coefficient of certainty (R²) changes from 0.919 of 12 forecast scenarios to 0.931 of 24 forecast scenarios, increasing by 1.31%. The mean value of mean relative error (MRE) changes from 6.80% of 12 forecast scenarios to 5.64% of 24 forecast scenarios, a decrease of 17.06%. Finally, the best model adaptation table corresponding to different forecast scenarios of TGR is established, which has an important guiding role in the actual runoff forecast of TGR.

     

River Flow Forecasting using Recurrent Neural Networks

Forecasting a hydrologic time series has been one of the most complicated tasks owing to the wide range of data, the uncertainties in the parameters influencing the time series and also due to the non availability of adequate data. Recently, Artificial Neural Networks (ANNs) have become quite popular in time series forecasting in various fields. This paper demonstrates the use of ANNs to forecast monthly river flows. Two different networks, namely the feed forward network and the recurrent neural network, have been chosen. The feed forward network is trained using the conventional back propagation algorithm with many improvements and the recurrent neural network is trained using the method of ordered partial derivatives. The selection of architecture and the training procedure for both the networks are presented. The selected ANN models were used to train and forecast the monthly flows of a river in India, with a catchment area of 5189 km² up to the gauging site. The trained networks are used for both single step ahead and multiple step ahead forecasting. A comparative study of both networks indicates that the recurrent neural networks performed better than the feed forward networks. In addition, the size of the architecture and the training time required were less for the recurrent neural networks. The recurrent neural network gave better results for both single step ahead and multiple step ahead forecasting. Hence recurrent neural networks are recommended as a tool for river flow forecasting.     

南水北调河北省受水区污水资源化利用

根据南水北调河北省受水区城市用水与排水、受水区水资源的现状分布及开发利用情况,利用BP人工神经网络方法建立受水城市的污水排放预测模型,提出基于水质分级的受水区污水利用规划,为受水区当地地表水、地下水、外调水、污水、微咸水等多种水源的联合配置提供科学依据。     

基于城市雨洪模型的市政排水与水利排涝标准衔接研究

针对目前存在的市政排水与水利排涝两个标准的衔接仍无规范统一方法的问题,以广州市东濠涌流域为研究区域,采用城市综合流域排水模型Info Works ICM建立东濠涌流域管道、河道及地面二维耦合模型,分析计算市政排水1年一遇与水利排涝5年一遇以及市政排水1年一遇与水利排涝10年一遇两种情况下的标准衔接关系,为城市排水防涝规划设计提供技术支撑。结果表明:1年一遇市政排水标准与10年一遇水利排涝标准的组合能够满足流域涝水顺利排除的要求,但管道排水口底高程距河底高程的距离过短也会对管道的水位顶托产生一定影响,故建议城市排水管网的规划建设应至少保证排水口底高程高于河道底高程0.5 m以上。     

ANN for Hybrid Energy System Evaluation: Methodology and WSS Case Study

Water Supply Systems (WSS) are large consumers of energy mainly used in pumping stations and treatment plants. Therefore, the improvement of energy efficiency is a major priority for water utilities. The current research work presents a new methodology and a computational algorithm based on renewable energy concepts, hydraulic system behaviour, pressure control and neural networks for the determination of the best hybrid energy configuration to be applied in a typical water supply system. The Artificial Neural Network (ANN) created to determine the best hybrid system uses scenarios with only grid supply, grid combined with hydro turbine, with wind turbine and mutual solutions with hydro and wind turbine. The ANN is trained based on values obtained from a configuration and economical simulator model (CES), as well as from a hydraulic and power simulator model (HPS). The results obtained show this ANN advanced computational model is useful for decision support solutions in the plan of sustainable hybrid energy systems that can be applied in water supply systems or other existent hydro systems allowing the improvement of the global energy efficiency. A real case study is analysed to determine the best sustainable hybrid energy solution in a small WSS of Portugal.     

Continuous monitoring in sewer networks an approach for quantification of pollution loads from CSOs into surface water bodies.

Different approaches for quantification of pollution loads discharged from combined sewer networks into surface water bodies have been observed over the last few years and decades, but a large number of unresolved problems still remain. Many monitoring campaigns have been based on manual or automated spot sampling - with the long known limitations of this method such as sampling errors and errors due to sample conservation, transport and preparation. On the other hand, only recently have sensors became available which are suitable for continuous application in sewer networks. A large number of practical problems still have to be solved before continuous monitoring in sewer networks will be successful. Additionally, most of the applicable sensors are based on surrogate methods which results in a considerable effort for reference measurements for sensor calibration. Finally, it has to be considered that, depending on the sewer network topography, deposition and remobilisation of pollutants varies considerably, which limits the generality of monitoring results and, subsequently, their applicability as a base for the design of storm water tanks or combined sewer overflows (CSO). A monitoring station for continuous monitoring of load discharges from a CSO has been installed and operated for more than one year. The design and equipment of the measurement station, operational experiences and results are given in this paper.     

Enhancing urban infrastructure investment planning practices for a changing climate.

Climate change raises many concerns for urban water management because of the effects on all aspects of the hydrological cycle. Urban water infrastructure has traditionally been designed using historical observations and assuming stationary climatic conditions. The capability of this infrastructure, whether for storm-water drainage, or water supply, may be over- or under-designed for future climatic conditions. In particular, changes in the frequency and intensity of extreme rainfall events will have the most acute effect on storm-water drainage systems. Therefore, it is necessary to take future climatic conditions into consideration in engineering designs in order to enhance water infrastructure investment planning practices in a long time horizon. This paper provides the initial results of a study that is examining ways to enhance urban infrastructure investment planning practices against changes in hydrologic regimes for a changing climate. Design storms and intensity-duration-frequency curves that are used in the engineering design of storm-water drainage systems are developed under future climatic conditions by empirically adjusting the general circulation model output, and using the Gumbel distribution and the Chicago method. Simulations are then performed on an existing storm-water drainage system from NE Calgary to investigate the resiliency of the system under climate change.     

Artificial intelligence technologies in surface water quality monitoring

Abstract:

Water quality monitoring networks are designed to detect, evaluate, arid quantify past, present, and emerging water quality problems and trends. A review of strategies to design monitoring networks suggests that a structured and consistent design methodology is for the most part missing. Monitoring network design is a complex task that requires an optimal configuration to ensure the maximum information extraction from the water quality data collected. In order to attain an optimal, and ultimately cost‐effective, network design, complementary design techniques or tools are needed. An overview of potentially applicable artificial intelligence technologies, as well as a literature review of promising research undertaken in the water resources area with respect to artificial intelligence techniques are presented. The artificial intelligence technologies examined include expert systems, artificial neural networks, genetic algorithms, and fuzzy logic systems.     

Risk Based Analysis for Contamination Event Selection and Optimal Sensor Placement for Intermittent Water Distribution Network Security

Water distribution networks are vulnerable to various contamination events that may be accidental or purposeful. Sensors are required for online monitoring of water quality to safeguard human health. Since sensors are costly, their numbers must be limited that makes sensor locations crucial in the water monitoring system. This paper aims at location of sensors in intermittent water distribution system which are more prone to accidental contamination due to contaminants ingress into the pipe lines because of low pressures during non supply hours. Considering deployment of limited number of sensors, the novelty of the paper is to propose a methodology for selection of contamination events with associated risk to be used in design of sensor network. Integrated risk assessment model is used to identify risk prone areas that may lead to possible contamination events. A Genetic Algorithm based methodology is suggested for optimal location of water quality sensors to maximize the detection likelihood of the contamination events within the acceptable time from the risk prone areas to improve network security. A comparison of sensor network design is made by considering contamination events occurring with: (i) equal probability at all the nodes; (ii) equal probability at risk prone nodes; and (iii) probability of occurrences based on quantified risk, to show that identification of risk prone areas and selection of contamination events results in reduction of computational work and more sensible placement of sensors.     

Manage Sewer In-Line Storage Control Using Hydraulic Model and Recurrent Neural Network

This paper described manage sewer in-line storage control for the city of Drammen, Norway. The purpose of the control is to use the free space of the pipes to reduce overflow at the wastewater treatment plant (WWTP). This study combined the powerful sides of the hydraulic model and neural networks. A detailed hydraulic model was developed to identify which part of the sewer system have more free space. Subsequently, the effectiveness of the proposed control solution was tested. Simulation results showed that intentionally control sewer with free space could significantly reduce overflow at the WWTP. At last, in order to enhance better decision making and give enough response time for the proposed control solution, Recurrent Neural Network (RNN) was employed to forecast flow. Three RNN architectures, namely Elman, NARX (nonlinear autoregressive network with exogenous inputs) and a novel architecture of neural networks, LSTM (Long Short-Term Memory), were compared. The LSTM exhibits the superior capability for time series prediction.     

Performance-costs evaluation for urban storm drainage.

The design process of urban stormwater systems incorporating BMPs involves more complexity unlike the design of classic drainage systems for which just the technique of pipes is likely to be used. This paper presents a simple decision aid methodology and an associated software (AvDren) concerning urban stormwater systems, devoted to the evaluation and the comparison of drainage scenarios using BMPs according to different technical, sanitary, social environmental and economical aspects. This kind of tool is particularly interesting so as to help the decision makers to select the appropriate alternative and to plan the investments especially for developing countries, with important sanitary problems and severe budget restrictions.     

基于GRACE和GRACE-FO的黄河流域陆地水储量及影响因素分析EI北大核心CSCD

基于GRACE和GRACE-FO卫星陆地水储量遥感数据,采用长短期记忆(LSTM)神经网络模型,结合水量平衡方程和全球陆地数据同化系统(GLDAS)重建GRACE与GRACE-FO间的陆地水储量变化量,分析黄河流域2002年4月至2020年3月陆地水储量变化特征,探究影响陆地水储量变化的环境因子。结果表明:LSTM模型可以有效填补GRACE与GRACE-FO间的陆地水储量变化量;黄河流域陆地水储量呈明显下降趋势,上、中、下游下降趋势依次增大,陆地水储量与地下水储量的变化特征高度相关;黄河流域上、中、下游年陆地水储量变化量与年降水量和年干燥度指数呈极显著相关关系,表明黄河流域陆地水储量变化受到降水和蒸散发的影响。     

图神经网络驱动的流域洪水预报技术

本研究提出了一个图指导的时空关联预报模型（GSCPM，graph-guided spatiotemporal correlation prediction model），针对性地解决流域洪水预报中的时空关系建模和滞后影响问题。该模型通过多个长短期记忆网络（LSTM）编码每个监测点历史属性的时间关联特征，随后利用图卷积神经网络（GCN）挖掘监测点间的地理空间依赖。此外，提出了雨量滞后特征、泄洪量滞后特征和上游水位滞后特征用以挖掘变量滞后效应。本文在现实流域数据集上进行了广泛的实验，通过跟LSTM、RNN 等模型的比较，证明了GSCPM 模型的优越性，适合在流域洪水预报中推广使用。     

Leakage Identification in Water Distribution Networks with Error Tolerance Capability

Leakages in water distribution networks have caused considerable waste of water resources. Thus, this study proposes a novel method for hydraulically monitoring and identifying regions where leakages occur in near-real time. A large network is first divided into several identification regions. To exploit a strong constructive and discriminative power, sparse coding is used, thereby adaptively coding the information embedded in observed pressures efficiently and succinctly. And a linear classifier is trained to determine the most likely leakage regions. A benchmark case is presented in this study to demonstrate the effectiveness of the proposed method. Results indicate that the proposed method can identify leakage events with enhanced tolerance capability for measurement errors. The method is also partially effective for identifying two simultaneous leakages. Certain practical advice in balancing the number of sensors and regions is also discussed to enhance the application potential of this method.

     

Entropy-Based Sensor Placement Optimization for Waterloss Detection in Water Distribution Networks

The work presented herein addresses the problem of sensor placement optimization in urban water distribution networks by use of an entropy-based approach, for the purpose of efficient and economically viable waterloss incident detection. The proposed method is applicable to longitudinal rather than spatial sensing, thus to devices such as acoustic, pressure, or flow sensors acting on pipe segments. The method utilizes the maximality, subadditivity and equivocation properties of entropy, coupled with a statistical definition of the probability of sensing within a pipe segment, to assign an entropy metric to each pipe segment and subsequently optimize the location of sensors in the network based on maximizing the total entropy in the network. The method proposed is a greedy-search heuristic.