Pressure-Dependent EPANET Extension

In water distribution systems (WDSs), the available flow at a demand node is dependent on the pressure at that node. When a network is lacking in pressure, not all consumer demands will be met in full. In this context, the assumption that all demands are fully satisfied regardless of the pressure in the system becomes unreasonable and represents the main limitation of the conventional demand driven analysis (DDA) approach to WDS modelling. A realistic depiction of the network performance can only be attained by considering demands to be pressure dependent. This paper presents an extension of the renowned DDA based hydraulic simulator EPANET 2 to incorporate pressure-dependent demands. This extension is termed “EPANET-PDX” (pressure-dependent extension) herein. The utilization of a continuous nodal pressure-flow function coupled with a line search and backtracking procedure greatly enhance the algorithm’s convergence rate and robustness. Simulations of real life networks consisting of multiple sources, pipes, valves and pumps were successfully executed and results are presented herein. Excellent modelling performance was achieved for analysing both normal and pressure deficient conditions of the WDSs. Detailed computational efficiency results of EPANET-PDX with reference to EPANET 2 are included as well.     

Evaluation of Reliability Indicators for WDNs with Demand-Driven and Pressure-Driven Models

Reliability of water distribution networks (WDNs) has received much attention in recent years due to progressive aging of infrastructures and climate change. Several reliability indicators, focusing on hydraulic aspects rather than water quality, have been proposed in literature. Reliability is generally assessed resorting to well established methods coupling hydraulic simulations and stochastic techniques that describe the WDNs hydraulic performance and component availability respectively. Two main algorithms are employed to simulate WDNs: the demand driven approach (DDA) that disregards the physical relationship between actual water demand and nodal pressure, and the pressure driven approach (PDA) that explicitly incorporates it. In this paper, we show how the choice of hydraulic solver may affect reliability indicators. We modify existing quantitative indicators at nodal and network level, and define novel indicators to consider water quality aspects. These indicators are evaluated for three example WDNs; discrepancies between results obtained with the two approaches depend on network size, feeding scheme and skeletonization. Results suggest to use with caution the DDA for reliability assessment at both local and global level.     

Extending EPANET hydraulic solver capacity with rigid water column global gradient algorithm

EPANET is one of the most commonly used open-source programs in hydraulic modelling water distribution networks (WDNs), based on steady-state and extended period simulation approaches. These approaches effectively estimate flow capacity and average pressures in networks; however, EPANET is not yet fully effective in modelling incompressible unsteady flows in WDNs. In this study, the hydraulic solver capacity of EPANET 3 is extended with the Rigid Water Column Global Gradient Algorithm (RWC-GGA) to model incompressible unsteady flow hydraulics in WDNs. Moreover, we incorporated dynamically more accurate valve expressions than the existing ones in the default EPANET code and introduced a new global convergence algorithm, Convergence Tracking Control Method (CTCM), in the solver code. The RWC-GGA, CTCM, and valve expressions are tested and validated in three different WDNs varying from simple to sophisticated set-ups. The results show that incompressible unsteady flows can be modelled with RWC-CGA and dynamic valve representations. Finally, the convergence problem due to the valve motion and the pressure-dependent algorithm (PDA) is solved by the implemented global convergence algorithm, i.e. CTCM.     

基于主成分分析的GA-BP模型在城市需水预测中的应用研究

针对城市需水预测模型中需水量影响因子多、影响因子之间普遍存在多重共线问题,以及BP神经网络收敛速度慢、易陷入局部最优等缺点,提出一种由主成分分析、遗传算法及BP神经网络三者相结合的改进预测模型。以泰州市为实例,建立以主成分分析筛选需水量主要影响因子,遗传算法优化BP网络连接权值和阈值的需水预测模型,预测结果与BP神经网络预测模型预测结果做对比。结果表明:改进预测模型对泰州市2003-2014年需水量预测的平均相对误差为0.564%,最大相对误差为1.681%,精度优于BP神经网络预测模型;改进预测模型预测值与实际泰州市需水量吻合良好且训练速度更快、预测精度更高,可作为需水预测的一种有效方法。     

Pressure Driven analysis of water distribution systems for preventing siphonic flow

The analysis of the water distribution network is complicated and requires several assumptions to simplify its problem definition. Demand Driven Analysis (DDA) is typically used to analyse the network assuming that all network nodes can deliver the required demand regardless of the available pressure. In the case of analysing an existing network under deficit condition such as pipe breakage or extra demand required for firefighting, assumptions used to simulate the network with DDA is not valid. Node Head Flow Relationship (NHFR) should be considered through Pressure Driven Analysis (PDA) to analyse the network. Most PDA methods assume that the networks are airtight which means that if the pressure at any demand node is negative, delivered demand will be equal to zero and the flow is permitted in the connected pipes (Siphonic flow). This assumption is hydraulically incorrect since the air is allowed to get into the connected pipes and prevent their flow leading to node isolation. In this paper, a new Pressure Driven Analysis to Prevent Siphonic Flow (PDA-SF) approach is proposed to analyze the network under deficit conditions and consider isolating the nodes that show available head less than node elevation. The PDA-SF was tested and compared to previous methods in four case studies under steady state analysis or extended period simulation. The case studies cover also different network conditions whether node isolation is needed or not. The PDA-SF was able to solve different networks where other methods failed to achieve the required demand or service pressure. The new PDA-SF method shall enable peers and modelers to better simulate and analysis water distribution networks.     

异构多种群粒子群优化算法在水位流量关系拟合中的应用

通过8个复杂函数对一种异构多种群粒子群优化算法进行仿真验证,并与传统单种群粒子群优化算法进行对比。针对水位流量关系拟合中相关参数难以确定的不足,利用异构多种群粒子群优化算法优化水位流量关系相关参数,以云南省龙潭站、西洋站水位流量关系拟合为例进行实例研究,并与粒子群优化算法、最小二乘法拟合结果进行对比。结果表明:异构多种群粒子群优化算法收敛精度远远优于粒子群优化算法,具有较好的计算鲁棒性和全局寻优能力。该算法对龙潭站和西洋站水位流量关系拟合的平均相对误差绝对值分别仅为0.27%和0.50%,拟合精度优于粒子群优化算法和最小二乘法。利用异构多种群粒子群优化算法优化水位流量关系可以获得更好的拟合效果。     

基于人工生态系统优化算法的组合生长需水预测模型

为提高需水预测精度,拓展生长模型在需水预测中的应用,提出基于人工生态系统优化(AEO)算法的组合生长需水预测模型。结合实例,选取6个标准测试函数在不同维度条件下对AEO算法进行仿真验证,并与鲸鱼优化算法(WOA)、灰狼优化(GWO)算法、教学优化(TLBO)算法和传统粒子群优化(PSO)算法的仿真结果进行比较。基于Weibull、Richards、Usher 3种单一生长模型构建Weibull-Richards-Usher、Weibull-Richards、Weibull-Usher、Richards-Usher 4种组合生长模型,利用AEO算法同时对组合模型参数和权重系数进行优化,提出AEO-Weibull-Richards-Usher、AEO-Weibull-Richards、AEO-Weibull-Usher、AEO-Richards-Usher需水预测模型,并构建AEO-Weibull、AEO-Richards、AEO-Usher、AEO-SVM、AEO-BP模型作对比,以上海市需水预测为例进行实例验证,利用实例前30组和后8组统计资料对各组合模型进行训练和预测。结果表明,在不同维度条件下,AEO算法寻优精度优于WOA、GWO、TLBO、PSO算法,具有较好的寻优精度和全局搜索能力。4种组合模型对实例预测的平均相对误差绝对值、平均绝对误差分别在0.94%~1.17%、0.30亿~0.37亿m³之间,预测精度优于AEO-Weibull等其他5种模型。4种组合模型均具有较好的预测精度和泛化能力,表明AEO算法能同时有效优化组合生长模型参数和权重系数,基于AEO算法的组合生长模型用于需水预测是可行和有效的。     

Simulation of Water Distribution Network under Pressure-Deficient Condition

Pressure deficient condition occurs in the water distribution network (WDN) when the nodal demands are in excess of the design discharge as in the case of fire demand, pump failure, pipe breaks, valve failure etc. It causes either no-flow or partial-flow depending upon the available pressure head at the nodes. To evaluate the nodal flows in such condition, node flow analysis (NFA) gives reasonable results in comparison to demand-driven analysis (DDA) and head-dependent analysis (HDA). The NFA works on the predefined pressure-discharge relationship to evaluate the nodal flows. However, this approach requires human intervention and hence cannot be applied to large WDN. Recently, modified pressure-deficient network algorithm (M-PDNA) has been developed by Babu and Mohan (2012) for pressure-deficient analysis with EPANET toolkit. However, it requires modification of the source code of EPANET. In this study a relationship with the M-PDNA and node flow analysis (Gupta and Bhave 1996) has been investigated and it is found that M-PDNA is the simplified version of NFA. Further, the working principle of M-PDNA has been investigated with suitable examples of Babu and Mohan (2012). The theoretical basis of M-PDNA has not been investigated in terms of head-discharge relationship. Herein, a head-discharge relationship based on the working principal of M-PDNA is proposed. Some of the toolkits are also readily available to modify demand driven solver of EPANET 2 to suit for the pressure-driven analysis and then it can be used for analysing pressure deficient network. Also in this study, a modification in M-PDNA approach is proposed which does not require the use of EPANET toolkit which is found to be capable of simulating both pressure-sufficient and pressure-deficient conditions in a single hydraulic simulation. Using the proposed approach, pressure-deficient condition is analysed with constant and variable demand pattern.     

多策略人工蜂群算法在梯级水电站优化调度中的应用

梯级水电站优化调度问题的准确、快速求解,是水利学科领域需解决的基本问题。针对该问题,提出了一种新的多策略人工蜂群算法。为更好地平衡算法的全局搜索与局部搜索能力,新算法在两个具有代表性的解搜索策略基础上,对其融合构成新的搜索策略,同时保留了原有的两个解搜索策略。新算法的三个候选解搜索策略,增强了对各类优化问题求解的适应性。为验证新算法的适应性及可行性,不仅在经典的基准测试函数中对其进行测试,并且将其应用于梯级水电站优化调度问题。实验结果表明,新算法具有适应性强、收敛速度快等优点。     

基于GRA-MEA-BP耦合模型的城市需水预测研究

针对城市需水量影响因子多、BP神经网络收敛速度慢、精度低、易陷入局部最优等问题,提出灰色关联分析、思维进化算法、BP神经网络三者耦合的改进预测模型,利用灰色关联分析(GRA)筛选需水量主要影响因子,采用全局搜索能力极强的思维进化算法(MEA)优化BP神经网络的权值和阈值,从而构建GRA-MEA-BP耦合需水预测模型,同时建立BP神经网络模型作为对比。实例应用结果表明,GRA-MEA-BP耦合模型具有更高的预测精度和预测速度,可作为一种有效的需水预测模型。     

Dynamic modeling of thermal conditions for hot-water district-heating networks

To investigate the dynamic characteristics of the thermal conditions of hot-water district-heating networks, a dynamic modeling method is proposed with consideration of the heat dissipations in pipes and the characteristic line method is adopted to solve it. Besides, the influences of different errors, space steps and initial values on the convergence of the dynamic model results are analyzed for a model network. Finally, a part of a certain city district-heating system is simulated and the results are compared with the actual operation data in half an hour from 6 secondary heat stations. The results indicate that the relative errors for the supply pressure and temperature in 5 stations are all within 2%, except in one station, where the relative error approaches 4%. So the proposed model and algorithm are validated.     

Noniterative Implementation of Pressure-Dependent Demands Using the Hydraulic Analysis Engine of EPANET 2

To analyze water distribution networks under pressure-deficient conditions, most of the available hydraulic simulators, including EPANET 2, must be either modified by embedding pressure-dependent demands in the governing network equations or run repeatedly with successive adjustments made to specific parameters until a sufficient hydraulic consistency is obtained. This paper presents and discusses a simple technique that implements the square root relationship between the nodal demand and the nodal pressure using EPANET 2 tools and allows a water distribution network with pressure-dependent demands to be solved in a single run of the unmodified snapshot hydraulic analysis engine of EPANET 2. In this technique, artificial strings made up of a flow control valve, a pipe with a check valve, and a reservoir are connected to the demand nodes before running the engine, and the pressure-dependent demands are determined as the flows in the strings. The resistance of the artificial pipes is chosen such that the demands are satisfied in full at a desired nodal pressure. The proposed technique shows reasonable convergence as evidenced by its testing on example networks.     

改进的遗传算法在水库调度中的应用

为了改善遗传算法在水库优化调度中的应用效果,采用自适应遗传算法和广度变异模块相结合的分层收敛算法:第一层采用广度变异和外部存档的方式改善种群的多样性;第二层嵌套广度变异模块,并采用自适应遗传算法进行全局搜索。通过比较自适应遗传算法和分层进化算法,结果显示:基于遗传算法的分层算法具有高效的全局搜索能力,避免了自适应遗传算法陷入局部最优的缺陷,在一定收敛条件下得到了更接近全局最优的目标值。     

基于新型蝙蝠算法-投影寻踪模型的文山州水量分配方法

基于公平和效率原则,构建文山州水量分配指标体系和水量分配投影寻踪(PP)模型。针对PP模型最佳投影方向难以确定的不足,利用新型蝙蝠算法(NBA)搜寻PP模型最佳投影方向,构建NBA-PP水量分配模型对文山州8县(市)水量进行分配。通过5个典型测试函数对NBA算法进行仿真验证,仿真结果与基本蝙蝠算法(BA)、人工蜂群算法(ABC)、布谷鸟搜索(CS)算法和差分进化算法(DE)进行对比。结果表明:通过引入生境选择策略及自适应补偿回声多普勒效应机制的NBA算法能有效平衡全局搜索和局部开发能力,寻优效果优于DE、CS、ABC和BA算法,具有较快的收敛速度、较高的寻优精度和较好的收敛稳定性与收敛可靠性;NBA-PP模型水量分配结果较目前分类权重法分配结果更科学、客观。     

Modelling Water Distribution Systems with Deficient Pressure: An Improved Iterative Methodology

In the last three decades, many researchers have proposed different models for water distribution network (WDN) hydraulic analysis by head-driven analysis (HDA). By considering a pressure-discharge relationship (PDR), head-driven analysis (HDA) can avoid deviation caused by traditional demand-driven analysis (DDA) under abnormal conditions. Generally, there are three types of HDA models: 1) models achieved by embedding a PDR into DDA, 2) models using EPANET structures such as emitter or tank to take place of PDR, 3) models aiming at modifying nodal outflows to satisfy PDR based on EPANET. Among these models, modifying nodal outflows is flexible to simulate network with different PDRs and specify parameters related to PDR. Most of the models use iterative algorithms to solve HDA problems; however, present ways to ensure convergence of models are still inadequate. The purpose of this paper is to present a new way to meet the iterative convergence when modifying nodal outflows based on PDR and leakage. This new methodology has been incorporated into the hydraulic network solver EPANET and is formalized algorithmically as EPANET-IMNO. Then two typical networks are used to test EPANET-IMNO, and the results demonstrate that EPANET-IMNO can converge well and applied successfully both in static simulation and extended period simulation. Different pressure deficiency conditions are tested to further confirm the flexibility and the convergence of EPANET-IMNO. Furthermore, quality analysis results back that pressure reduction can be a practical way in contamination accident response.     

基于遗传学习算法和BP算法的神经网络在矿坑涌水量计算中的应用

本文采用遗传学习算法和误差反向传播算法(BP)相结合的混合算法来训练前馈人工神经网络(BPN)，即先用遗传学习算法进行全局训练，再用BP算法进行精确训练，使网络收敛速度加快和避免局部极小。作为实例，本文将该方法运用于多维时序问题。根据山东省黑旺铁矿的矿坑充水条件建立了一个网络，以矿坑充水的各种控制因素相关资料作为样本，对网络进行训练并用训练好的网络预测矿坑涌水量。网络的训练速度及预测结果表明，该算法收敛速度较快，预测精度很高，为矿坑涌水量预报提供了一种新思路和新方法。     

Chaotic Evolutionary Algorithms for Multi-Reservoir Optimization

The water sharing dispute in a multi-reservoir river basin forces the water resources planners to have an integrated operation of multi-reservoir system rather than considering them as a single reservoir system. Thus, optimizing the operations of a multi-reservoir system for an integrated operation is gaining importance, especially in India. Recently, evolutionary algorithms have been successfully applied for optimizing the multi-reservoir system operations. The evolutionary optimization algorithms start its search from a randomly generated initial population to attain the global optimal solution. However, simple evolutionary algorithms are slower in convergence and also results in sub-optimal solutions for complex problems with hardbound variables. Hence, in the present study, chaotic technique is introduced to generate the initial population and also in other search steps to enhance the performance of the evolutionary algorithms and applied for the optimization of a multi-reservoir system. The results are compared with that of a simple GA and DE algorithm. From the study, it is found that the chaotic algorithm with the general optimizer has produced the global optimal solution (optimal hydropower production in the present case) within lesser generations. This shows that coupling the chaotic algorithm with evolutionary algorithm will enrich the search technique by having better initial population and also converges quickly. Further, the performances of the developed policies are evaluated for longer run using a simulation model to assess the irrigation deficits. The simulation results show that the model satisfactorily meets the irrigation demand in most of the time periods and the deficit is very less.     

Penalty-Free Feasibility Boundary Convergent Multi-Objective Evolutionary Algorithm for the Optimization of Water Distribution Systems

This paper presents a new penalty-free multi-objective evolutionary approach (PFMOEA) for the optimization of water distribution systems (WDSs). The proposed approach utilizes pressure dependent analysis (PDA) to develop a multi-objective evolutionary search. PDA is able to simulate both normal and pressure deficient networks and provides the means to accurately and rapidly identify the feasible region of the solution space, effectively locating global or near global optimal solutions along its active constraint boundary. The significant advantage of this method over previous methods is that it eliminates the need for ad-hoc penalty functions, additional ??boundary search?? parameters, or special constraint handling procedures. Conceptually, the approach is downright straightforward and probably the simplest hitherto. The PFMOEA has been applied to several WDS benchmarks and its performance examined. It is demonstrated that the approach is highly robust and efficient in locating optimal solutions. Superior results in terms of the initial network construction cost and number of hydraulic simulations required were obtained. The improvements are demonstrated through comparisons with previously published solutions from the literature.     

改进差分进化算法在水库优化调度中的应用研究

差分进化算法在求解水库优化调度时,进化后期种群多样性急剧下降,导致算法无法跳出局部最优解而出现“早熟”收敛。针对该问题,该文对算法的贪婪选择策略进行改进,使其以一定的概率动态接受稍差解作为子代个体,从而提高算法的种群多样性;同时,提出种群基因重生策略,进一步改善种群进化的基因信息结构。将改进的差分进化算法应用于清江梯级发电调度问题,并与差分进化算法、模拟退火算法求解结果进行对比。模拟结果表明,改进算法具有更强的全局搜索能力,求解梯级水库优化调度问题更具有优势。     

Reducing Irrigation Deficiencies Based Optimizing Model for Multi-Reservoir Systems Utilizing Spider Monkey Algorithm

Continuous droughts and water scarcity have led to the need for optimal exploitation of dams’ reservoirs. Thus, the new meta-heuristic algorithm, spider monkey, is suggested for complex modeling of the multi-reservoir system in Iran with the aim of decreasing irrigation deficiencies. Golestan and Voshmgir dams’ operations are optimized with the spider monkey algorithm. The algorithm based on the exchange of information between local and global leaders with the other monkeys which improves the convergence speed. Average deficiencies for Golestan dam is computed as 2.1 and 1.9 MCM by spider monkey algorithm while it is respectively computed as 6.7, 16.4, 11.1, 4.1, 14.6, 19 MCM by particle swarm algorithm, harmony search algorithm, imperialist competitive algorithm, water cycle algorithm, genetic algorithm, and standards operation policy method. Also, the computation time of the spider monkey algorithm is 50 and 47 s for the Golestan and Voshmgir dams while the genetic algorithm optimizes the problem in 172.6 s and 112 s and the particle swarm algorithm needs 117.4 s and 100 s for the Golestan and Voshmgir, respectively. Also, root means square error (RMSE) and mean absolute error between demand and released water for the spider monkey algorithm have the least values among the applied evolutionary algorithms. Thus, the spider monkey algorithm is suggested as an appropriate method for optimizing the operation policy for the dam and reservoir systems.