WO2022036820A1 - 一种基于供水物联网数据同化的污水管网实时模拟方法 - Google Patents

一种基于供水物联网数据同化的污水管网实时模拟方法 Download PDF

Info

Publication number
WO2022036820A1
WO2022036820A1 PCT/CN2020/120149 CN2020120149W WO2022036820A1 WO 2022036820 A1 WO2022036820 A1 WO 2022036820A1 CN 2020120149 W CN2020120149 W CN 2020120149W WO 2022036820 A1 WO2022036820 A1 WO 2022036820A1
Authority
WO
WIPO (PCT)
Prior art keywords
pipe network
water supply
water
time
sewage pipe
Prior art date
Application number
PCT/CN2020/120149
Other languages
English (en)
French (fr)
Inventor
郑飞飞
张清周
Original Assignee
浙江大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 浙江大学 filed Critical 浙江大学
Publication of WO2022036820A1 publication Critical patent/WO2022036820A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/28—Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/18—Network design, e.g. design based on topological or interconnect aspects of utility systems, piping, heating ventilation air conditioning [HVAC] or cabling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2113/00—Details relating to the application field
    • G06F2113/08—Fluids
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2113/00—Details relating to the application field
    • G06F2113/14—Pipes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces

Definitions

  • the invention belongs to the field of municipal engineering urban water supply and drainage pipe networks, in particular to a real-time simulation technology of sewage pipe networks based on data assimilation of the Internet of Things for water supply.
  • the safe operation of the urban sewage pipe network directly affects the urban water environment, water safety and people's health.
  • the scale of sewage pipe network has been continuously expanded, the topology structure has become more complex, and the system aging has become more serious, which has brought great difficulties to the operation and management of sewage pipe network.
  • the problems that are prone to occur in the urban sewage pipe network at present mainly include pipe network deposition, stealing waste water, pipeline leakage, misconnection of rain and sewage pipes and sewage overflow. These problems have seriously threatened the urban water environment and are also the root cause of urban black and odorous water bodies. reason.
  • the direct method to solve these problems is to arrange sensors in the sewage pipe network to monitor the water depth and flow information of the pipes in real time, so as to realize the prediction, early warning and localization of abnormal events.
  • the number of sensors arranged in the pipeline network is usually very limited, and only a small area around the monitoring point can be alarmed for abnormal events (such as overflow or leakage).
  • the abnormal observation results of monitoring points may also be caused by the sudden increase of user water consumption, so the data analysis method of monitoring points that does not consider the change of water consumption is likely to lead to a high false alarm rate.
  • the key to realizing the real-time hydraulic simulation of the sewage pipe network is to obtain the real-time sewage inflow of each inspection well, but it is unrealistic to obtain flow data with such a high spatial and temporal resolution in actual engineering, which is also the card of the current real-time simulation of the sewage pipe network. Neck problem.
  • researchers have proposed many optimization methods to invert the real-time sewage inflow of inspection wells through the observation data of limited monitoring points.
  • the present invention proposes for the first time a real-time simulation method of sewage pipe network based on the data-driven water supply Internet of Things.
  • the water consumption at all times is allocated to the inspection wells of the nearest sewage pipe network, and then the optimal transfer coefficient between the water consumption of the nodes and the inflow of the inspection wells is determined by the optimization method, which effectively solves the problem of multi-solution.
  • the real-time water consumption is based on the water supply.
  • the data and the determined optimal transfer coefficients enable real-time simulation of water depth and flow parameters throughout the sewer network.
  • the innovation of the present invention is that the Internet of Things in the water supply system, which has become mature in recent years, has achieved in-depth integration and data assimilation with the sewage pipe network.
  • the Internet of Things in the water supply includes many pressure gauges, flow meters and smart water meters, which can provide users with water in real time. Quantitative information, which in turn drives the real-time simulation of the sewage network.
  • the use of the real-time hydraulic model of the sewage pipe network provides key technical support for effectively solving the problems of pipe blockage, stealing, leakage, pipe misconnection and sewage overflow in the sewage pipe network.
  • a real-time simulation method of sewage pipe network based on data-driven water supply Internet of Things the steps of which are as follows:
  • Offline module including three stages S1, S2 and S3, the execution frequency and times of offline module are determined according to actual needs,
  • S22 Initialize the water consumption of each node at a certain historical time t: For a water supply network with a given number of nodes N x , where N y nodes are installed with smart water meters (y ⁇ x), first N y The separately metered water consumption is allocated to the corresponding nodes, and the remaining water volume is proportionally allocated to the remaining N x -N y nodes according to the length of the pipeline connecting each node to the adjacent node.
  • the specific formula is as follows:
  • l r is the total length of the pipeline connected to node r
  • L T is the total length of the water supply network pipeline
  • L M is the smart water meter node The total length of the connected pipes
  • Q T is the total water supply volume of the water supply network
  • Q M is the total water volume of the smart water meter at historical time t
  • N x nodes in the pipeline network there are a total of N x nodes in the pipeline network, some of which (the number is N y ) are equipped with smart water meters, and the water consumption at historical time t is directly obtained from the smart water meters, and the other part of the nodes (N x -N y ) are not installed.
  • the water consumption of smart water meters at historical time t is unknown.
  • the water consumption of these nodes (N x -N y ) at historical time t is calculated according to formula 1-1.
  • the total initial value of all nodes in the water supply network at historical time t is calculated.
  • the water consumption is equal to the sum of the initial water consumption of each node (a total of N x nodes) at historical time t,
  • J H and J Q are the Jacobian matrices of the water supply network at the s-th iteration, respectively; where, and respectively represent the weight coefficients of the pressure monitoring point u and the flow monitoring point v; is the weight coefficient vector,
  • S27 Repeat the process S21-S26 to obtain the water consumption data of the water supply pipe network nodes with a historical time period of T (usually 2 weeks) and a time precision of ⁇ t (referring to the time difference between two t moments before and after, usually half an hour). , used for the calculation of S3;
  • K [k 1 , k 2 ,...k n ] T
  • k l is the water consumption transfer coefficient of the inspection well l in the sewage pipe network model
  • T is the total time for correction and simulation of the sewage pipe network model
  • Tw Simulate the hot start time for the sewage pipe network
  • M and N are the number of liquid level gauges and flow meters installed in the sewage pipe network, respectively, which are obtained from the sewage pipe network data acquisition system; and are the observed value and simulated value of the liquid level of the liquid level monitoring point i at the historical time t, respectively, and are the observed and simulated flow values of the flow monitoring point j in the sewage pipe network at the historical time t, and are the analog value vectors at all times of the entire time history period T of the liquid level monitoring point i and the flow monitoring point j, respectively
  • F s (D(T)) is and The combined vector of , D(T) is a T ⁇ n matrix, representing the inflow of all inspection wells n at all times
  • x represents the observed value or simulated value of the monitoring point
  • x min and x max are the upper and lower limits, which are generally obtained according to the historical data statistics of the monitoring point for a period of time (such as 30 days)
  • the present invention has the following advantages:
  • the present invention proposes a data assimilation method for the water supply system and the sewage pipe network for the first time.
  • the mapping relationship between the water consumption of the water supply pipe network node and the sewage pipe network inspection well inflow By establishing the mapping relationship between the water consumption of the water supply pipe network node and the sewage pipe network inspection well inflow, the data of the sewage pipe network inspection well inflow is effectively solved. A serious lack of problems.
  • the present invention proposes a method for real-time correction of water consumption of water supply pipe network nodes and single-objective optimization calculation method for the transfer coefficient of sewage pipe network inspection wells, innovatively realizes the real-time simulation method of sewage pipe network based on water supply data driving, and completely solves the existing problems of sewage pipe network.
  • the multi-solution problem common in the simulation technology realizes the real-time and accurate simulation of the hydraulic parameters of the liquid level and flow of the entire sewage pipe network.
  • the present invention fills the blank in the field of real-time simulation of sewage pipe network, is an important supplement to the research field of urban drainage pipe network management, provides important technical support for the management of sewage pipe network system, and has good promotion and practical engineering application. value.
  • Fig. 1 is a functional schematic diagram of the water supply pipe network and the sewage pipe network of the present invention.
  • FIG. 2 is a schematic diagram of the integration of the water supply pipe network and the sewage pipe network according to the present invention.
  • FIG. 3 is a specific implementation roadmap of the present invention.
  • Fig. 4 is the layout diagram of the sewage pipe network and water supply pipe network system and monitoring points of the embodiment BKN.
  • Fig. 5 is the layout diagram of the XZN sewage pipe network system and monitoring points of the embodiment.
  • Fig. 6 is the layout diagram of the XZN water supply pipe network system and monitoring points of the embodiment.
  • Fig. 7 is the error distribution diagram of the simulated value and the monitored value of the monitoring points of the Benk and XZN water supply pipe network of the embodiment (a total of 816 time steps in the first 17 days).
  • Figure 8 is a comparison of the water consumption of the nodes corrected by the water supply network of the embodiment and the known water consumption connected to the smart water meter: (a) Case BKN (b) Case XZN.
  • FIG. 9 is a distribution diagram of the transfer coefficient between the water consumption of an embodiment node and the inflow of the inspection well.
  • Figure 10 is the comparison of simulated and observed values for the first 17 days (correction stage) of the flow monitoring point of the embodiment: (a) Case BKN flowmeter C1 (b) Case XZN flowmeter C3.
  • Figure 11 is the comparison between the simulated and observed values of the flow monitoring point (a) C1 and the liquid level monitoring points (b) M1, (c) M2, (d) M3 of the embodiment BKN in the model verification stage (the last 14 days).
  • Figure 12 is a comparison between the simulated and observed values of flow monitoring points (a) C1, (b) C2 and liquid level monitoring points (c) M1 and (d) M5 of Example XZN in the model validation stage (later 14 days).
  • Fig. 13 is a comparison between the simulated value and the observed value of the XZN liquid level monitoring point M5 of the embodiment, and the simulated value of the water depth of 10 nearby inspection wells.
  • Offline module including three stages S1, S2 and S3, the execution frequency and times of offline module are determined according to actual needs,
  • S22 Initialize the water consumption of each node at a certain historical time t: For a water supply network with a given number of nodes N x , where N y nodes are installed with smart water meters (y ⁇ x), first N y The separately metered water consumption is allocated to the corresponding nodes, and the remaining water volume is proportionally allocated to the remaining N x -N y nodes according to the length of the pipeline connecting each node to the adjacent node.
  • the specific formula is as follows:
  • l r is the total length of the pipeline connected to node r
  • L T is the total length of the water supply network pipeline
  • L M is the smart water meter node The total length of the connected pipes
  • Q T is the total water supply volume of the water supply network
  • Q M is the total water volume of the smart water meter at historical time t
  • N x nodes in the pipeline network there are a total of N x nodes in the pipeline network, some of which (the number is N y ) are equipped with smart water meters, and the water consumption at historical time t is directly obtained from the smart water meters, and the other part of the nodes (N x -N y ) are not installed.
  • the water consumption of smart water meters at historical time t is unknown.
  • the water consumption of these nodes (N x -N y ) at historical time t is calculated according to formula 1-1.
  • the total initial value of all nodes in the water supply network at historical time t is calculated.
  • the water consumption is equal to the sum of the initial water consumption of each node (a total of N x nodes) at historical time t,
  • J H and J Q are the Jacobian matrices of the water supply network at the s-th iteration, respectively; where, and respectively represent the weight coefficients of the pressure monitoring point u and the flow monitoring point v; is the weight coefficient vector,
  • S27 Repeat the process S21-S26 to obtain the water consumption data of the water supply pipe network nodes with a historical time period of T (usually 2 weeks) and a time precision of ⁇ t (referring to the time difference between two t moments before and after, usually half an hour). , used for the calculation of S3;
  • K [k 1 , k 2 ,...k n ] T
  • k l is the water consumption transfer coefficient of the inspection well l in the sewage pipe network model
  • T is the total time for correction and simulation of the sewage pipe network model
  • Tw Simulate the hot start time for the sewage pipe network
  • M and N are the number of liquid level gauges and flow meters installed in the sewage pipe network, respectively, which are obtained from the sewage pipe network data acquisition system; and are the observed value and simulated value of the liquid level of the liquid level monitoring point i at the historical time t, respectively, and are the observed and simulated flow values of the flow monitoring point j in the sewage pipe network at the historical time t, and are the analog value vectors at all times of the entire time history period T of the liquid level monitoring point i and the flow monitoring point j, respectively
  • F s (D(T)) is and The combined vector of , D(T) is a T ⁇ n matrix, representing the inflow of all inspection wells n at all times
  • x represents the observed value or simulated value of the monitoring point
  • x min and x max are the upper and lower limits, which are generally obtained according to the historical data statistics of the monitoring point for a period of time (such as 30 days)
  • the sewage pipe network in Benk (denoted as FSS-BKN) consists of one sewage plant entrance, 64 inspection wells and 64 sewage pipes. Three liquid level gauges and one flow meter are installed in the sewage pipe network (as shown in Figure 4). shown), the sewage discharge is about 4100 tons/day; its corresponding water supply pipe network (denoted as WDS-BKN) consists of 1 water plant, 65 water demand nodes and 93 water supply pipes.
  • the sewage pipe network in Xiuzhou (denoted as FSS-XZN) consists of a sewage plant entrance, 1,214 inspection wells and 1,214 sewage pipes (as shown in Figure 5), with a total length of about 86 kilometers and a sewage discharge of about 21,500 tons.
  • the monitoring instrument records historical data without rainfall for 31 days in a month, and the time step is 30 minutes, so each monitoring point has a total of 1488 (31 ⁇ 24 ⁇ 2) time steps of data.
  • the historical data of the monitoring points of the water supply network and sewage network for 17 consecutive days were selected to determine the optimal transfer coefficient of each inspection well in the sewage network model.
  • Tw 3 days
  • the minimum value of the transfer coefficient of the inspection well is For the inspection wells corresponding to the nodal water consumption provided by the smart water meter, the maximum value For the inspection wells corresponding to the node water consumption obtained by checking the hydraulic model, considering the possible errors in the checking, the maximum value
  • the population size of the genetic algorithm in the conventional technique used was 500, the maximum number of iterations was 50000, and the default values were used for the remaining parameters.
  • Figure 7 shows the error distribution between the simulated and monitored values for the first 17 days (816 time steps) of the corrected Benk and XZN water supply network monitoring points. It can be seen from Figure 7(a) that more than 90% of the absolute errors at all pressure monitoring points in Benk are less than 0.32 meters, and the maximum value is 1.34%; in Figure 7(b), at the Benk flow monitoring points in the water supply network, 93% The relative error of the flow rate is less than 1.5%, and its maximum value is 2.4%; it can be seen from Figure 7(c) that the absolute errors at all pressure monitoring points in the water supply network XZN are less than 0.5 meters; in Figure 7(d), XZN Most of the relative errors at the flow monitoring points are less than 5%, and the maximum value is 9.27%.
  • Figure 8 is a comparison diagram of the water consumption correction value of nodes without smart water meters and the real water consumption values of nodes with smart water meters installed in two embodiments. It can be seen from the figure that the corrected water consumption of nodes and the actual water consumption recorded by smart water meters The trend of water consumption is the same, that is, the low-peak and peak periods of water consumption are the same, which shows that the correction error of water consumption of nodes in the two embodiments is not only satisfying the error requirements of the model application, but the correction results of water consumption of nodes are the same as the trend of actual water consumption, which is more scientific and reasonable. This ensures that the corrected hydraulic model can be accurately applied in practice.
  • Fig. 9 is the distribution diagram of the water consumption transfer coefficient of the sewage pipe network inspection wells in the two embodiments. It can be seen from the figure that the transfer coefficients of most sewage inspection wells are in the range of [0, 1]. The mean values of well transfer coefficients were 0.83 and 0.92, respectively, implying that 83% and 92% of the total water consumption in the water supply network entered the sewer network through inspection wells.
  • Figure 10(a) is a comparison chart of the observed value and the simulated value of the first 17 days (correction stage) of the BKN sewage pipe network flowmeter C1 of the embodiment. The errors are 4.5% and 1.16%, respectively (Fig. 10b).
  • Figure 10(c) is a comparison chart of the observed value and the simulated value of the first 17 days (correction stage) of the XZN sewage pipe network flowmeter C3 of the embodiment.
  • the maximum value and average value of the relative error between the monitoring value and the simulated value are 13.74% and 3.02%, respectively.
  • Figure 11 is a comparison diagram of the sewage flow of the flow meter C1 and the water depth observation values of the liquid level meters M1, M2 and M3 in the model verification stage (the last 14 days) of the embodiment BKN sewage pipe network, and the maximum value of the relative error of the flow is 4.91%, the maximum absolute error of water depth is 0.7cm.
  • Figure 12 is a comparison diagram of the observed and simulated values of flowmeters C1, C2 and liquid level meters M1, M5 in the model verification stage (the last 14 days) of Example XZN.
  • the maximum relative errors of flowmeters C1 and C2 are 13.05% and 13.45%, respectively.
  • the maximum absolute error between the observed value of the liquid level meter M1 and M5 and the simulated value is 1.4cm and 1.1cm, respectively.
  • Figure 13 is a comparison chart of the real-time simulated value and the actual observation value of the water depth of the XZN level meter M5 of the sewage pipe network of the embodiment at each moment of a certain day, and the real-time simulated value of the water depth of 10 nearby inspection wells.
  • Water depth if the water depth fluctuation in a certain period of time exceeds the normal range, an alarm will be issued, and then by analyzing the real-time water depth data of all nearby inspection wells, the location of abnormal events (such as stealing discharge, leakage, etc.) can be quickly determined.
  • the water consumption at each moment in the water supply network model in the same area is allocated to the nearest sewage pipe network inspection wells in the vicinity through the real-time simulation method of the sewage pipe network based on the data-driven water supply Internet of Things proposed in the present invention. , and use the optimization method to determine the transfer coefficient between the water consumption of the node and the inflow of the inspection well, and realize the real-time simulation of the liquid level and flow parameters of the entire sewage pipe network. It provides important technical support for issues such as discharge, rain and sewage misconnection, and sewage overflow, and has good promotion and practical engineering application value.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Geometry (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • Fluid Mechanics (AREA)
  • Mathematical Physics (AREA)
  • Computing Systems (AREA)
  • Algebra (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Computational Mathematics (AREA)
  • Sewage (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种基于供水物联网数据驱动的污水管网实时模拟方法，包括离线模块和实时在线模块，离线模块集成污水管网与供水管网水力模型，校正供水管网水力模型每个节点的历史用水量，建立污水管网模型校正单目标优化模型，确定每个节点用水量和检查井入流量之间的转移系数，实时在线模块实现污水管网模型水力参数的实时模拟。该方法填补了污水管网实时模拟领域空白，是对城市排水管网管理研究领域的一个重要补充，为污水管网系统的管理提供了重要的技术支撑，具有很好的推广和实际工程应用价值。

Description

一种基于供水物联网数据同化的污水管网实时模拟方法 技术领域
本发明属于市政工程城市供排水管网领域，具体涉及基于供水物联网数据同化的污水管网实时模拟技术。
背景技术
城市污水管网的安全运行直接影响到城市水环境、水安全以及人民的身体健康。近些年，随着城市人口数量的快速增长，污水管网规模不断扩大，拓扑结构变得更加复杂，而且系统老化更加严重，这给污水管网的运行和管理带来了极大的困难。城市污水管网目前容易产生的问题主要包括管网沉积、废水偷排、管道泄漏、雨污管错接和污水溢流等，这些问题已严重威胁城市水环境，也是产生城市黑臭水体的根本原因。
在现有技术中，解决这些问题的直接方法是在污水管网内布置传感器，实时监测管道水深及流量信息，以实现异常事件的预报预警与定位。然而，由于污水传感器购买和维护成本极高，管网中布置的传感器数量通常非常有限，只能对监测点周围小片区域进行异常事件的(如溢出或泄露)报警。此外，监测点的异常观测结果也可能是由于用户用水量突然增加造成，因此不考虑用水量的变化的监测点数据分析方法易导致很高的误报率。更重要的是，仅仅依靠来自传感器观测不能预测未来一段时期整个污水管网运行的运行状态(水深和流量)，进而无法实现有效的预防与控制。突破这些关键难题的一个重要方法是建立污水管网水力模型以实时模拟和预测整个管网任意位置的水深和流量参数，同时结合有限监测点的数据，以实时诊断是否存在管道淤塞、泄漏、非法排放、非法连接，更重要的是，可以实时预测污水管网所有检查井的溢流情况。
实现污水管网实时水力模拟的关键是获取每个检查井的实时污水入流量，但获取如此高时空分辨率的流量数据在实际工程中是不现实的，这也是目前污水管网实时模拟的卡脖子难题。研究学者提出了诸多优化方法，通过有限监测点的观测数据反演检查井的实时污水入流量。然而，这些方法的一个重大缺陷是污水流量优化的多解性，即每个检查井入流量取值的不同组合仍能保证监测点处的模拟值与观测值相吻合，因 此，很难确定优化解是否可以表征代表污水管网真实的水力运行状况，进而无法实现污水管网的有效监测。
发明内容
为解决上述现有技术中的瓶颈问题，本发明首次提出了一种基于供水物联网数据驱动的污水管网实时模拟方法，通过集成相同区域内的供水管网模型，将供水管网节点每个时刻的用水量分配至附近最近的污水管网检查井，随后使用优化方法确定节点用水量和检查井入流量之间的最佳转移系数，有效解决了多解性问题，最后，基于供水实时用水数据和已确定的最佳转移系数实现整个污水管网水深和流量参数的实时模拟。本发明的创新之处是将近些年趋于成熟的供水系统物联网与污水管网实现深入的融合和数据同化，供水物联网包含诸多的压力计、流量计和智能水表，可实时提供用户用水量信息，进而驱动污水管网的实时模拟。本发明中，采用污水管网实时水力模型为有效解决污水管网中的管道淤塞、偷排、泄露、管道错接以及污水溢流问题提供了关键的技术支撑。
本发明解决上述技术问题所具体采用的技术方案如下：
一种基于供水物联网数据驱动的污水管网实时模拟方法，其步骤如下：
过程1：离线模块，包括S1、S2和S3三个阶段，离线模块的执行频率和次数根据实际需要确定，
S1:按步骤S11-S12集成污水管网与供水管网水力模型，
S11:基于地理信息系统GIS提供的供水管道、水库、泵站、污水管道、检查井等模型组件的参数信息，建立污水管网和供水管网水力模型(图1)，
S12:基于GIS自带的空间分析功能，建立供水管网模型节点与污水管网模型检查井之间的映射关系，实现模型中每个供水节点的排水进入与之空间距离最近的污水检查井(图2)；
S2:按步骤S21-S27校正供水管网水力模型每个节点的历史用水量，
S21:设置所需的相关参数：供水管网中所有压力监测点和流量监测点在历史某时刻t的观测值H o、Q o；误差阈值ε；最大迭代次数S和节点用水量调整范围p，
S22:初始化每个节点在在历史某时刻t的用水量：对于一个给定节点数量为N x的供水管网，其中N y个节点安装了智能水表(y＜x)，首先将N y个单独计量的用水量分配至相应的节点，剩余水量根据每个节点与相邻节点连接的管道长度按比例分配至其 余N x-N y个节点，具体公式如下：
Figure PCTCN2020120149-appb-000001
式中，
Figure PCTCN2020120149-appb-000002
为按管道长度比例分配的节点r在历史某时刻t初始化后的节点用水量，l r为与节点r相连接的管道总长度，L T为供水管网管道总长度，L M为智能水表节点连接的管道总长度；Q T为供水管网总供水量；Q M为智能水表在历史t时刻的总水量，
管网中共有N x个节点，其中有一部分节点(数量为N y)处装有智能水表，在历史t时刻的用水量根据智能水表直接获取，另外一部分节点(N x-N y)未装有智能水表在历史t时刻t的用水量未知，这些节点(N x-N y)在历史t时刻的用水量根据公式1-1计算得出，供水管网中历史t时刻所有节点的总初始用水量等于历史t时刻每个节点(总共N x个节点)初始用水量之和，
S23:计算历史t时刻压力和流量监测点的观测值与模拟值残差：运行供水管网水力模拟，计算第s次迭代时(s＝1,2,...,S)，其中，
压力监测的点观测值与模拟值残差为：
Figure PCTCN2020120149-appb-000003
流量监测点的观测值与模拟值残差为：
Figure PCTCN2020120149-appb-000004
式中，NH和NQ分别为压力和流量监测点的数量，
Figure PCTCN2020120149-appb-000005
和H u(q) s分别为压力监测点u的观测值和在第s次迭代时的模拟值(u＝1,2,...,NH)，
Figure PCTCN2020120149-appb-000006
和Q v(q) s分别为供水管网流量监测点v的观测值和在第s次迭代时的模拟值(v＝1,2,...,NH)，T表示向量的转置，
Figure PCTCN2020120149-appb-000007
为所有节点在第s次迭代时历史t时刻的节点用水量向量；
S24:计算历史t时刻节点用水量调整值，具体公式如下：
Figure PCTCN2020120149-appb-000008
式中，J H和J Q分别为第s次迭代时的供水管网雅克比矩阵；其中，
Figure PCTCN2020120149-appb-000009
Figure PCTCN2020120149-appb-000010
和
Figure PCTCN2020120149-appb-000011
分别表示压力监测点u和流量监测点v的权重系数；
Figure PCTCN2020120149-appb-000012
为权重系数向量，
S25:更新每个节点的用水量，具体公式如下：
q s+1＝q s+Δq s        1-5
Figure PCTCN2020120149-appb-000013
式中，q s+1为第s+1次迭代时每个节点在历史t时刻的用水量；
Figure PCTCN2020120149-appb-000014
和
Figure PCTCN2020120149-appb-000015
分别为某个节点r的在历史t时刻最小和最大用水量，一般取p＝10％～20％，
S26:重复过程S23～S24，直至满足||Δq s||＜ε或s＞S，一般取ε＝0.01，S＝100，
S27:重复过程S21～S26，获取历史时间周期为T(通常取2个星期)时间精度为Δt(指前后两个t时刻之间的时间差，通常取半小时)的供水管网节点用水量数据，用于S3的计算；
S3:按步骤S31-S32建立污水管网模型校正单目标优化模型，确定每个节点用水量和检查井入流量之间的转移系数，
S31:建立单目标函数，具体公式如下：
最小化函数：
Figure PCTCN2020120149-appb-000016
Figure PCTCN2020120149-appb-000017
Figure PCTCN2020120149-appb-000018
d l(t)＝k l×q l(t)        1-10
Figure PCTCN2020120149-appb-000019
式中，K＝[k 1,k 2,...k n] T，k l为污水管网模型中检查井l的用水量转移系数，T为污水管网模型校正模拟总时间，T w为污水管网模拟热启动时间，M和N分别为污水管网中安装的液位计和流量计的数量，在污水管网数据采集系统中获取；
Figure PCTCN2020120149-appb-000020
和
Figure PCTCN2020120149-appb-000021
分别为液 位监测点i在历史t时刻的液位观测值和模拟值，
Figure PCTCN2020120149-appb-000022
和
Figure PCTCN2020120149-appb-000023
为污水管网中的流量监测点j在历史t时刻的流量观测值和模拟值，
Figure PCTCN2020120149-appb-000024
和
Figure PCTCN2020120149-appb-000025
分别为液位监测点i和流量监测点j整个时间历史周期T所有时刻的模拟值向量，F s(D(T))为
Figure PCTCN2020120149-appb-000026
和
Figure PCTCN2020120149-appb-000027
的组合向量，D(T)为T×n矩阵，表示所有检查井n整个时间周期T所有时刻的入流量，d l(t)为检查井l在历史t时刻的入流量，q l(t)为步骤S2中检查井l对应的供水管网节点历史t时刻的用水量校正值；
Figure PCTCN2020120149-appb-000028
和
Figure PCTCN2020120149-appb-000029
分别为k l的最小值和最大值；g()为线性转换函数，用于将液位和流量转换为同一区间，即0到1区间范围，定义为：
Figure PCTCN2020120149-appb-000030
式中，x表示监测点的观测值或模拟值；x min和x max为上限和下限，一般根据监测点一段时期(如30天)的历史数据统计获取，
S32:求解单目标优化模型：使用现有技术中的遗传算法求解优化模型，得到每个检查井i的用水量转移系数k l(l＝1,...,n)；
过程2：实时在线模块，包括S4阶段，S4阶段每个时刻执行一次，
S4:按步骤S41-S43实现污水管网模型水力参数的实时模拟，
S41:根据供水管网压力计、流量计和智能水表获取当前t时刻的压力、流量和用水量数据，按照过程S2校正供水管网水力模型当前时刻t的节点用水量，
S42:根据S41获取的当前时刻t供水系统每个节点的用水量，以及S3得到的每个检查井的用水量转移系数，根据公式1-10计算污水管网中每个检查井当前时刻的入流量d l(t)，
S43:运行污水管网水力模型，实时模拟时间精度为Δt(通常取半小时)的整个污水管网的液位和流量水力参数。
相对于现有技术而言，本发明具有以下优点：
①本发明首次提出了供水系统与污水管网的数据同化方法，通过建立供水管网节点用水量与污水管网检查井入流量之间的映射关系，有效解决了污水管网检查井入流量数据严重缺乏的问题。
②本发明提出了供水管网节点用水量实时校正与污水管网检查井转移系数单目标优化计算方法，创新实现了基于供水数据驱动的污水管网实时模拟方法，彻 底解决了污水管网现有模拟技术普遍存在的多解问题，实现了整个污水管网液位和流量水力参数的实时准确模拟。
③本发明填补了污水管网实时模拟领域空白，是对城市排水管网管理研究领域的一个重要补充，为污水管网系统的管理提供了重要的技术支撑，具有很好的推广和实际工程应用价值。
附图说明
图1是本发明供水管网与污水管网功能示意图。
图2是本发明供水管网与污水管网集成示意图。
图3是本发明具体实施路线图。
图4是实施例BKN污水管网与供水管网系统及监测点布置图。
图5是实施例XZN污水管网系统及监测点布置图。
图6是实施例XZN供水管网系统及监测点布置图。
图7是实施例Benk和XZN供水管网监测点模拟值与监测值误差分布图(前17天共816个时间步长)。
图8是实施例供水管网校正的节点用水量与智能水表连接的已知用水量对比:(a)案例BKN(b)案例XZN。
图9是实施例节点用水量和检查井入流量之间的转移系数分布图。
图10是实施例流量监测点前17天(校正阶段)的模拟值与观测值对比:(a)案例BKN流量计C1(b)案例XZN流量计C3。
图11是实施例BKN在模型验证阶段(后14天)流量监测点(a)C1和液位监测点(b)M1、(c)M2、(d)M3的模拟值与观测值对比
图12是实施例XZN在模型验证阶段(后14天)流量监测点(a)C1、(b)C2和液位监测点(c)M1和(d)M5的模拟值与观测值对比。
图13是实施例XZN液位监测点M5的模拟值与观测值，以及附近10个检查井水深的模拟值对比。
具体实施方式
下面通过附图和实施例对本发明进行具体阐述，以便本领域技术人员更好地理解本发明的技术方案，需要说明的是，在实施方式中所列举的案例，其作用是为了使本领域技术人员更好地理解和实施本发明的技术方案，不应视为对本发明的限定或提前 公开。。
参见图3，一种基于供水物联网数据驱动的污水管网实时模拟方法，其步骤如下：
过程1：离线模块，包括S1、S2和S3三个阶段，离线模块的执行频率和次数根据实际需要确定，
S1:按步骤S11-S12集成污水管网与供水管网水力模型，
S11:基于地理信息系统GIS提供的供水管道、水库、泵站、污水管道、检查井等模型组件的参数信息，建立污水管网和供水管网水力模型(图1)，
S12:基于GIS自带的空间分析功能，建立供水管网模型节点与污水管网模型检查井之间的映射关系，实现模型中每个供水节点的排水进入与之空间距离最近的污水检查井(图2)；
S2:按步骤S21-S27校正供水管网水力模型每个节点的历史用水量，
S21:设置所需的相关参数：供水管网中所有压力监测点和流量监测点在历史某时刻t的观测值H o、Q o；误差阈值ε；最大迭代次数S和节点用水量调整范围p，
S22:初始化每个节点在在历史某时刻t的用水量：对于一个给定节点数量为N x的供水管网，其中N y个节点安装了智能水表(y＜x)，首先将N y个单独计量的用水量分配至相应的节点，剩余水量根据每个节点与相邻节点连接的管道长度按比例分配至其余N x-N y个节点，具体公式如下：
Figure PCTCN2020120149-appb-000031
式中，
Figure PCTCN2020120149-appb-000032
为按管道长度比例分配的节点r在历史某时刻t初始化后的节点用水量，l r为与节点r相连接的管道总长度，L T为供水管网管道总长度，L M为智能水表节点连接的管道总长度；Q T为供水管网总供水量；Q M为智能水表在历史t时刻的总水量，
管网中共有N x个节点，其中有一部分节点(数量为N y)处装有智能水表，在历史t时刻的用水量根据智能水表直接获取，另外一部分节点(N x-N y)未装有智能水表在历史t时刻t的用水量未知，这些节点(N x-N y)在历史t时刻的用水量根据公式1-1计算得出，供水管网中历史t时刻所有节点的总初始用水量等于历史t时刻每个节点(总共N x个节点)初始用水量之和，
S23:计算历史t时刻压力和流量监测点的观测值与模拟值残差：运行供水管网水 力模拟，计算第s次迭代时(s＝1,2,...,S)，其中，
压力监测的点观测值与模拟值残差为：
Figure PCTCN2020120149-appb-000033
流量监测点的观测值与模拟值残差为：
Figure PCTCN2020120149-appb-000034
式中，NH和NQ分别为压力和流量监测点的数量，
Figure PCTCN2020120149-appb-000035
和H u(q) s分别为压力监测点u的观测值和在第s次迭代时的模拟值(u＝1,2,...,NH)，
Figure PCTCN2020120149-appb-000036
和Q v(q) s分别为供水管网流量监测点v的观测值和在第s次迭代时的模拟值(v＝1,2,...,NH)，T表示向量的转置，
Figure PCTCN2020120149-appb-000037
为所有节点在第s次迭代时历史t时刻的节点用水量向量；
S24:计算历史t时刻节点用水量调整值，具体公式如下：
Figure PCTCN2020120149-appb-000038
式中，J H和J Q分别为第s次迭代时的供水管网雅克比矩阵；其中，
Figure PCTCN2020120149-appb-000039
Figure PCTCN2020120149-appb-000040
和
Figure PCTCN2020120149-appb-000041
分别表示压力监测点u和流量监测点v的权重系数；
Figure PCTCN2020120149-appb-000042
为权重系数向量，
S25:更新每个节点的用水量，具体公式如下：
q s+1＝q s+Δq s        1-5
Figure PCTCN2020120149-appb-000043
式中，q s+1为第s+1次迭代时每个节点在历史t时刻的用水量；
Figure PCTCN2020120149-appb-000044
和
Figure PCTCN2020120149-appb-000045
分别为某个节点r的在历史t时刻最小和最大用水量，一般取p＝10％～20％，
S26:重复过程S23～S24，直至满足||Δq s||＜ε或s＞S，一般取ε＝0.01，S＝100，
S27:重复过程S21～S26，获取历史时间周期为T(通常取2个星期)时间精度为Δt(指前后两个t时刻之间的时间差，通常取半小时)的供水管网节点用水量数据，用于S3的计算；
S3:按步骤S31-S32建立污水管网模型校正单目标优化模型，确定每个节点用水量和检查井入流量之间的转移系数，
S31:建立单目标函数，具体公式如下：
最小化函数：
Figure PCTCN2020120149-appb-000046
Figure PCTCN2020120149-appb-000047
Figure PCTCN2020120149-appb-000048
d l(t)＝k l×q l(t)        1-10
Figure PCTCN2020120149-appb-000049
式中，K＝[k 1,k 2,...k n] T，k l为污水管网模型中检查井l的用水量转移系数，T为污水管网模型校正模拟总时间，T w为污水管网模拟热启动时间，M和N分别为污水管网中安装的液位计和流量计的数量，在污水管网数据采集系统中获取；
Figure PCTCN2020120149-appb-000050
和
Figure PCTCN2020120149-appb-000051
分别为液位监测点i在历史t时刻的液位观测值和模拟值，
Figure PCTCN2020120149-appb-000052
和
Figure PCTCN2020120149-appb-000053
为污水管网中的流量监测点j在历史t时刻的流量观测值和模拟值，
Figure PCTCN2020120149-appb-000054
和
Figure PCTCN2020120149-appb-000055
分别为液位监测点i和流量监测点j整个时间历史周期T所有时刻的模拟值向量，F s(D(T))为
Figure PCTCN2020120149-appb-000056
和
Figure PCTCN2020120149-appb-000057
的组合向量，D(T)为T×n矩阵，表示所有检查井n整个时间周期T所有时刻的入流量，d l(t)为检查井l在历史t时刻的入流量，q l(t)为步骤S2中检查井l对应的供水管网节点历史t时刻的用水量校正值；
Figure PCTCN2020120149-appb-000058
和
Figure PCTCN2020120149-appb-000059
分别为k l的最小值和最大值；g()为线性转换函数，用于将液位和流量转换为同一区间，即0到1区间范围，定义为：
Figure PCTCN2020120149-appb-000060
式中，x表示监测点的观测值或模拟值；x min和x max为上限和下限，一般根据监测点一段时期(如30天)的历史数据统计获取，
S32:求解单目标优化模型：使用现有技术中的遗传算法求解优化模型，得到每个检查井i的用水量转移系数k l(l＝1,...,n)；
过程2：实时在线模块，包括S4阶段，S4阶段每个时刻执行一次，
S4:按步骤S41-S43实现污水管网模型水力参数的实时模拟，
S41:根据供水管网压力计、流量计和智能水表获取当前t时刻的压力、流量和用水量数据，按照过程S2校正供水管网水力模型当前时刻t的节点用水量，
S42:根据S41获取的当前时刻t供水系统每个节点的用水量，以及S3得到的每个检查井的用水量转移系数，根据公式1-10计算污水管网中每个检查井当前时刻的入流量d l(t)，
S43:运行污水管网水力模型，实时模拟时间精度为Δt(通常取半小时)的整个污水管网的液位和流量水力参数。下面基于该方法，将其与具体实施例结合，以展示其具体技术效果，方法的具体步骤不再赘述。
实施例
下面将本发明的上述方法分别应用到Benk和Xiuzhou两个城市的污水管网。城市Benk的污水管网(记为FSS-BKN)由1个污水厂入口、64个检查井和64根污水管道组成，污水管网中安装了3个液位计和一个流量计(如图4所示)，排污量约为4100吨/天；其相应的供水管网(记为WDS-BKN)由1个水厂、65个需水量节点和93根供水管道组成，供水管网中安装了3个压力计和1个流量计和40个智能水表，供水量约为4800吨/天；如图4所示，虚线箭头表示的是根据最短距离确定的供水管网节点与污水管网检查井之间的对应关系。城市Xiuzhou的污水管网(记为FSS-XZN)由1个污水厂入口、1214个检查井和1214根污水管道组成(如图5所示)，总长度约86千米，排污量约21500吨/天，污水管网中安装了3个流量计和8个液位计；其相应的供水管网(记为WDS-XZN)由1个水厂、1个泵站、1119个节点和1137根供水管道组成(如图6所示)，供水量约23150吨/天，服务人口约10.75万人，供水管网中安装有5个流量计、8个压力计和525个智能水表。
每个实例中，监测仪表记录了某月31天无降雨情况下的历史数据，时间步长为30分钟，因此每个监测点共有1488(31×24×2)个时间步长的数据。在过程1离线模式中，选取前面连续17天的供水管网和污水管网监测点历史数据(时间步长30分钟)，用于确定污水管网模型中每个检查井的最佳转移系数。污水管网检查井转移系数优化计算中，污水管网模型热启动时间T w＝3天，17天中剩余的14天用于确定转移系数。使用31天中最后连续14天供水管网和污水管网监测点历史数据，进行过程2中的污水管网每个时刻t的实时在线模拟结果验证。
供水管网节点用水量校核中，针对每个实施例，校正误差阈值ε＝0.1；最大迭代次数S＝100和节点用水量调整范围p＝20％。污水管网检查井转移系数优化计算中，检查井转移系数最小值为
Figure PCTCN2020120149-appb-000061
对于由智能水表提供的节点用水量所对应的检查井，最大值
Figure PCTCN2020120149-appb-000062
而对于通过水力模型校核得出的节点用水量所对应的检查井，考虑校核可能的误差，最大值
Figure PCTCN2020120149-appb-000063
所使用的常规技术中的遗传算法种群数量为500，最大迭代次数为50000，其余参数使用默认值。
图7为校正后的Benk和XZN供水管网监测点前17天(816个时间步长)的模拟值与监测值误差分布图。从图7(a)中可以看出Benk所有压力监测点处超过90％的绝对误差小于0.32米，最大值为1.34％；图7(b)中，供水管网Benk流量监测点处，93％的流量相对误差小于1.5％，其最大值为2.4％；从图7(c)中可以看出供水管网XZN所有压力监测点处的绝对误差均小于0.5米；图7(d)中，XZN流量监测点处大多数相对误差小于5％，最大值为9.27％。图8为两个实施例未装有智能水表的节点用水量校正值与安装智能水表的节点用水量真实值对比图，从图中可以看出，校正的节点用水量与智能水表记录的实际用水量趋势相同，即用水低峰和高峰时段相同，说明了两个实施例节点用水量校正误差除了满足模型应用的误差要求之外，节点用水量校正结果与实际用水量趋势相同，更具有科学合理性，保证了校正好的水力模型能够精确的进行实际应用。
图9为两个实施例污水管网检查井用水量转移系数分布图，从图中可以看出绝大多数的污水检查井转移系数位于[0,1]范围内，实施例BKN和XZN所有检查井转移系数的平均值分别为0.83和0.92，意味着83％和92％的供水管网总用水量通过检查井进入了污水管网。图10(a)为实施例BKN污水管网流量计C1前17天(校正阶段)的观测值与模拟值对比图，所有时刻监测值与模拟值的相对误差均小于5％，最大误差和平均误差分别为4.5％和1.16％(图10b)。图10(c)为实施例XZN污水管网流量计C3前17天(校正阶段)的观测值与模拟值对比图，监测值与模拟值相对误差最大值和平均值分别为13.74％和3.02％(图10d)。图11为实施例BKN污水管网在模型验证阶段(后14天)流量计C1的污水流量和液位计M1、M2和M3的水深观测值与模拟值对比图，其中流量相对误差最大值为4.91％，水深最大绝对误差为0.7cm。图12为实施例XZN在模型验证阶段(后14天)流量计C1、C2和液位计M1、M5观测值与模拟值对比图，流量计C1和C2相对误差最大值分别为13.05％和13.45％，液位计M1和 M5观测值与模拟值最大绝对误差分别为1.4cm和1.1cm。由此可知，两个实施例污水管网检查井转移系数校正结果更具有科学合理性，均能保证监测点处的模型模拟值与实际观测值相吻合。
图13为实施例污水管网XZN液位计M5某天每个时刻的水深实时模拟值与实际观测值，以及附近10个检查井水深的实时模拟值对比图，通过监测M5每个时刻的实时水深，若某个时间段水深波动超出正常范围，则进行报警，然后通过分析附近所有检查井的实时水深数据，可以快速确定异常事件(如偷排、泄漏等)的发生位置。
由此可知，通过本发明所提出的基于供水物联网数据驱动的污水管网实时模拟方法，将相同区域内的供水管网模型中每个时刻的用水量分配至附近最近的污水管网检查井，并使用优化方法确定节点用水量和检查井入流量之间的转移系数，实现了整个污水管网液位和流量参数的实时模拟，为有效解决污水管网中的淤塞、泄漏、沉积、非法排放、雨污错接、污水溢流等问题提供了重要技术支撑，具有很好的推广和实际工程应用价值。
以上所述的实施例只是本发明的一种较佳的方案，然其并非用以限制本发明。有关技术领域的普通技术人员，在不脱离本发明的精神和范围的情况下，还可以做出各种变化和变型。因此凡采取等同替换或等效变换的方式所获得的技术方案，均落在本发明的保护范围内。

Claims (1)

  1. 一种基于供水物联网数据驱动的污水管网实时模拟方法，其特征在于，步骤如下：
    过程1：离线模块，包括S1、S2和S3三个阶段，离线模块的执行频率和次数根据实际需要确定，
    S1:按步骤S11-S12集成污水管网与供水管网水力模型，
    S11:基于地理信息系统GIS提供的供水管道、水库、泵站、污水管道、检查井等模型组件的参数信息，建立污水管网和供水管网水力模型(图1)，
    S12:基于GIS自带的空间分析功能，建立供水管网模型节点与污水管网模型检查井之间的映射关系，实现模型中每个供水节点的排水进入与之空间距离最近的污水检查井(图2)；
    S2:按步骤S21-S27校正供水管网水力模型每个节点的历史用水量，
    S21:设置所需的相关参数：供水管网中所有压力监测点和流量监测点在历史某时刻t的观测值H o、Q o；误差阈值ε；最大迭代次数S和节点用水量调整范围p，
    S22:初始化每个节点在在历史某时刻t的用水量：对于一个给定节点数量为N x的供水管网，其中N y个节点安装了智能水表(y＜x)，首先将N y个单独计量的用水量分配至相应的节点，剩余水量根据每个节点与相邻节点连接的管道长度按比例分配至其余N x-N y个节点，具体公式如下：
    Figure PCTCN2020120149-appb-100001
    式中，
    Figure PCTCN2020120149-appb-100002
    为按管道长度比例分配的节点r在历史某时刻t初始化后的节点用水量，l r为与节点r相连接的管道总长度，L T为供水管网管道总长度，L M为智能水表节点连接的管道总长度；Q T为供水管网总供水量；Q M为智能水表在历史t时刻的总水量，
    管网中共有N x个节点，其中有一部分节点(数量为N y)处装有智能水表，在历史t时刻的用水量根据智能水表直接获取，另外一部分节点(N x-N y)未装有智能水表在历史t时刻t的用水量未知，这些节点(N x-N y)在历史t时刻的用水量根据公式1-1计算得出，供水管网中历史t时刻所有节点的总初始用水量等于历史t时刻每个节点(总 共N x个节点)初始用水量之和，
    S23:计算历史t时刻压力和流量监测点的观测值与模拟值残差：运行供水管网水力模拟，计算第s次迭代时(s＝1,2,...,S)，其中，
    压力监测的点观测值与模拟值残差为：
    Figure PCTCN2020120149-appb-100003
    流量监测点的观测值与模拟值残差为：
    Figure PCTCN2020120149-appb-100004
    式中，NH和NQ分别为压力和流量监测点的数量，
    Figure PCTCN2020120149-appb-100005
    和H u(q) s分别为压力监测点u的观测值和在第s次迭代时的模拟值(u＝1,2,...,NH)，
    Figure PCTCN2020120149-appb-100006
    和Q v(q) s分别为供水管网流量监测点v的观测值和在第s次迭代时的模拟值(v＝1,2,...,NH)，T表示向量的转置，
    Figure PCTCN2020120149-appb-100007
    为所有节点在第s次迭代时历史t时刻的节点用水量向量；
    S24:计算历史t时刻节点用水量调整值，具体公式如下：
    Figure PCTCN2020120149-appb-100008
    式中，J H和J Q分别为第s次迭代时的供水管网雅克比矩阵；其中，
    Figure PCTCN2020120149-appb-100009
    Figure PCTCN2020120149-appb-100010
    和
    Figure PCTCN2020120149-appb-100011
    分别表示压力监测点u和流量监测点v的权重系数；
    Figure PCTCN2020120149-appb-100012
    为权重系数向量，
    S25:更新每个节点的用水量，具体公式如下：
    q s+1＝q s+Δq s   1-5
    Figure PCTCN2020120149-appb-100013
    式中，q s+1为第s+1次迭代时每个节点在历史t时刻的用水量；
    Figure PCTCN2020120149-appb-100014
    和
    Figure PCTCN2020120149-appb-100015
    分别为某个节点r的在历史t时刻最小和最大用水量，一般取p＝10％～20％，
    S26:重复过程S23～S24，直至满足||Δq s||＜ε或s＞S，一般取ε＝0.01，S＝100，
    S27:重复过程S21～S26，获取历史时间周期为T(通常取2个星期)时间精度为Δt (指前后两个t时刻之间的时间差，通常取半小时)的供水管网节点用水量数据，用于S3的计算；
    S3:按步骤S31-S32建立污水管网模型校正单目标优化模型，确定每个节点用水量和检查井入流量之间的转移系数，
    S31:建立单目标函数，具体公式如下：
    最小化函数：
    Figure PCTCN2020120149-appb-100016
    Figure PCTCN2020120149-appb-100017
    Figure PCTCN2020120149-appb-100018
    d l(t)＝k l×q l(t)  1-10
    Figure PCTCN2020120149-appb-100019
    式中，K＝[k 1,k 2,...k n] T，k l为污水管网模型中检查井l的用水量转移系数，T为污水管网模型校正模拟总时间，T w为污水管网模拟热启动时间，M和N分别为污水管网中安装的液位计和流量计的数量，在污水管网数据采集系统中获取；
    Figure PCTCN2020120149-appb-100020
    和
    Figure PCTCN2020120149-appb-100021
    分别为液位监测点i在历史t时刻的液位观测值和模拟值，
    Figure PCTCN2020120149-appb-100022
    和
    Figure PCTCN2020120149-appb-100023
    为污水管网中的流量监测点j在历史t时刻的流量观测值和模拟值，
    Figure PCTCN2020120149-appb-100024
    和
    Figure PCTCN2020120149-appb-100025
    分别为液位监测点i和流量监测点j整个时间历史周期T所有时刻的模拟值向量，F s(D(T))为
    Figure PCTCN2020120149-appb-100026
    和
    Figure PCTCN2020120149-appb-100027
    的组合向量，D(T)为T×n矩阵，表示所有检查井n整个时间周期T所有时刻的入流量，d l(t)为检查井l在历史t时刻的入流量，q l(t)为步骤S2中检查井l对应的供水管网节点历史t时刻的用水量校正值；
    Figure PCTCN2020120149-appb-100028
    和
    Figure PCTCN2020120149-appb-100029
    分别为k l的最小值和最大值；g()为线性转换函数，用于将液位和流量转换为同一区间，即0到1区间范围，定义为：
    Figure PCTCN2020120149-appb-100030
    式中，x表示监测点的观测值或模拟值；x min和x max为上限和下限，一般根据监测点一段时期(如30天)的历史数据统计获取，
    S32:求解单目标优化模型：使用现有技术中的遗传算法求解优化模型，得到每个检查井i的用水量转移系数k l(l＝1,...,n)；
    过程2：实时在线模块，包括S4阶段，S4阶段每个时刻执行一次，
    S4:按步骤S41-S43实现污水管网模型水力参数的实时模拟，
    S41:根据供水管网压力计、流量计和智能水表获取当前t时刻的压力、流量和用水量数据，按照过程S2校正供水管网水力模型当前时刻t的节点用水量，
    S42:根据S41获取的当前时刻t供水系统每个节点的用水量，以及S3得到的每个检查井的用水量转移系数，根据公式1-10计算污水管网中每个检查井当前时刻的入流量d l(t)，
    S43:运行污水管网水力模型，实时模拟时间精度为Δt(通常取半小时)的整个污水管网的液位和流量水力参数。
PCT/CN2020/120149 2020-08-18 2020-10-10 一种基于供水物联网数据同化的污水管网实时模拟方法 WO2022036820A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010832786.9A CN112182984B (zh) 2020-08-18 2020-08-18 一种基于供水物联网数据驱动的污水管网实时模拟方法
CN202010832786.9 2020-08-18

Publications (1)

Publication Number Publication Date
WO2022036820A1 true WO2022036820A1 (zh) 2022-02-24

Family

ID=73919430

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/120149 WO2022036820A1 (zh) 2020-08-18 2020-10-10 一种基于供水物联网数据同化的污水管网实时模拟方法

Country Status (2)

Country Link
CN (1) CN112182984B (zh)
WO (1) WO2022036820A1 (zh)

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114741831A (zh) * 2022-06-13 2022-07-12 成都同飞科技有限责任公司 一种供水管网水力压强预警分布的模拟方法及系统
CN114818211A (zh) * 2022-04-12 2022-07-29 长江生态环保集团有限公司 一种基于mike+构建污水收集系统量化模拟技术方法
CN115861011A (zh) * 2023-02-15 2023-03-28 山东优嘉环境科技有限公司 一种基于多源数据融合的智慧城市优化管理方法和系统
CN116102096A (zh) * 2023-01-17 2023-05-12 中节能国祯环保科技股份有限公司 一种城市污水厂网一体化的控制方法及系统
CN116150930A (zh) * 2023-04-20 2023-05-23 天津智云水务科技有限公司 基于3d地图信息的供水管网水力模型节点流量初始化方法
CN116662767A (zh) * 2023-08-01 2023-08-29 陕西中科绿能能源研究院有限公司 基于多传感器的蒸发冷却机组系统温度数据智能采集方法
CN116976046A (zh) * 2023-06-12 2023-10-31 中国建筑第四工程局有限公司 基于管网追踪的污水监控方法及系统
CN117034044A (zh) * 2023-10-09 2023-11-10 山东朝启电子科技有限公司 一种基于数据智能的智能表井管理系统
CN117057819A (zh) * 2023-08-15 2023-11-14 泰华智慧产业集团股份有限公司 一种雨水管网污水排放溯源分析方法及系统
CN117522199A (zh) * 2023-11-08 2024-02-06 重庆大学 一种城市供水管网综合安全评价及监测点选址方法
CN117610793A (zh) * 2024-01-22 2024-02-27 成都秦川物联网科技股份有限公司 一种智能水表远程维护方法、装置及物联网系统
CN117805338A (zh) * 2024-03-01 2024-04-02 广东省建筑设计研究院有限公司 一种建筑供水管网水质实时在线监测方法及系统

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113011903A (zh) * 2021-04-26 2021-06-22 成都环极科技有限公司 基于gis和水力模型的水污染精准溯源方法
CN113781276B (zh) * 2021-07-12 2024-01-19 浙江大学 一种污水管网流量不确定性分析方法
CN113626959B (zh) * 2021-07-12 2023-12-08 浙江大学 一种基于三维地理信息的污水管网水力模型建立方法
WO2023284060A1 (zh) * 2021-07-12 2023-01-19 浙江大学 一种基于地理三维信息的污水管网流量不确定性分析方法
CN113803646B (zh) * 2021-09-01 2023-04-07 长江生态环保集团有限公司 适用于缺陷管道排查的排水管网监测布点方法
CN113818536A (zh) * 2021-11-03 2021-12-21 江苏四联自动化科技有限公司 一种污水管网排水能力综合调度方法及系统
CN114477329B (zh) * 2022-02-22 2023-01-06 江苏舜维环境工程有限公司 一种水泥厂集成化水处理装置

Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102032935A (zh) * 2010-12-07 2011-04-27 杭州电子科技大学 城市排水汇流管网污水泵站流量的软测量方法
CN102592008A (zh) * 2011-12-30 2012-07-18 华东师范大学 基于gis技术的城市排水管网设计系统及方法
US20140156232A1 (en) * 2012-11-30 2014-06-05 International Business Machines Corporation Methods, systems and computer program storage devices for generating a flooding forecast
CN104598698A (zh) * 2015-02-13 2015-05-06 北京清控人居环境研究院有限公司 在线构架系统以及基于在线构架的供水管网模型建模方法
CN107506536A (zh) * 2017-08-07 2017-12-22 浙江大学 一种针对多水源供水管网的供水优化分区方法
CN107563007A (zh) * 2017-08-07 2018-01-09 浙江大学 一种节点流量和管道阻力系数同时调整的供水管网模型快速校正方法
CN107633340A (zh) * 2017-07-21 2018-01-26 浙江大学 一种基于压力监测的供水管网漏失区域快速识别方法
CN108197725A (zh) * 2017-12-06 2018-06-22 浙江大学 基于需水量先验信息的供水管网节点需水量校核方法
CN108332059A (zh) * 2018-01-16 2018-07-27 浙江大学 服务于供水管网爆管监测的测压点优化布置方法
CN108664684A (zh) * 2018-03-16 2018-10-16 浙江大学 一种基于瞬变模拟的供水管网串联管道简化方法
CN108758357A (zh) * 2018-06-01 2018-11-06 浙江大学 一种基于供水管网压力变化值分析的爆管定位方法
CN109237311A (zh) * 2018-09-13 2019-01-18 上海万朗水务科技有限公司 污水处理厂排水管道管网信息系统
CN110929359A (zh) * 2019-11-20 2020-03-27 合肥泽众城市智能科技有限公司 基于pnn神经网络和swmm技术的管网淤积风险预测建模方法

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107016622B (zh) * 2017-03-22 2020-08-04 浙江大学 一种含大用户用水信息的城市供水管网节点需水量反演方法

Patent Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102032935A (zh) * 2010-12-07 2011-04-27 杭州电子科技大学 城市排水汇流管网污水泵站流量的软测量方法
CN102592008A (zh) * 2011-12-30 2012-07-18 华东师范大学 基于gis技术的城市排水管网设计系统及方法
US20140156232A1 (en) * 2012-11-30 2014-06-05 International Business Machines Corporation Methods, systems and computer program storage devices for generating a flooding forecast
CN104598698A (zh) * 2015-02-13 2015-05-06 北京清控人居环境研究院有限公司 在线构架系统以及基于在线构架的供水管网模型建模方法
CN107633340A (zh) * 2017-07-21 2018-01-26 浙江大学 一种基于压力监测的供水管网漏失区域快速识别方法
CN107563007A (zh) * 2017-08-07 2018-01-09 浙江大学 一种节点流量和管道阻力系数同时调整的供水管网模型快速校正方法
CN107506536A (zh) * 2017-08-07 2017-12-22 浙江大学 一种针对多水源供水管网的供水优化分区方法
CN108197725A (zh) * 2017-12-06 2018-06-22 浙江大学 基于需水量先验信息的供水管网节点需水量校核方法
CN108332059A (zh) * 2018-01-16 2018-07-27 浙江大学 服务于供水管网爆管监测的测压点优化布置方法
CN108664684A (zh) * 2018-03-16 2018-10-16 浙江大学 一种基于瞬变模拟的供水管网串联管道简化方法
CN108758357A (zh) * 2018-06-01 2018-11-06 浙江大学 一种基于供水管网压力变化值分析的爆管定位方法
CN109237311A (zh) * 2018-09-13 2019-01-18 上海万朗水务科技有限公司 污水处理厂排水管道管网信息系统
CN110929359A (zh) * 2019-11-20 2020-03-27 合肥泽众城市智能科技有限公司 基于pnn神经网络和swmm技术的管网淤积风险预测建模方法

Cited By (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114818211A (zh) * 2022-04-12 2022-07-29 长江生态环保集团有限公司 一种基于mike+构建污水收集系统量化模拟技术方法
CN114818211B (zh) * 2022-04-12 2024-05-10 长江生态环保集团有限公司 一种基于mike+构建污水收集系统量化模拟技术方法
CN114741831A (zh) * 2022-06-13 2022-07-12 成都同飞科技有限责任公司 一种供水管网水力压强预警分布的模拟方法及系统
CN114741831B (zh) * 2022-06-13 2022-08-23 成都同飞科技有限责任公司 一种供水管网水力压强预警分布的模拟方法及系统
CN116102096B (zh) * 2023-01-17 2024-03-01 中节能国祯环保科技股份有限公司 一种城市污水厂网一体化的控制方法及系统
CN116102096A (zh) * 2023-01-17 2023-05-12 中节能国祯环保科技股份有限公司 一种城市污水厂网一体化的控制方法及系统
CN115861011B (zh) * 2023-02-15 2023-05-05 山东优嘉环境科技有限公司 一种基于多源数据融合的智慧城市优化管理方法和系统
CN115861011A (zh) * 2023-02-15 2023-03-28 山东优嘉环境科技有限公司 一种基于多源数据融合的智慧城市优化管理方法和系统
CN116150930A (zh) * 2023-04-20 2023-05-23 天津智云水务科技有限公司 基于3d地图信息的供水管网水力模型节点流量初始化方法
CN116976046B (zh) * 2023-06-12 2024-04-05 中国建筑第四工程局有限公司 基于管网追踪的污水监控方法及系统
CN116976046A (zh) * 2023-06-12 2023-10-31 中国建筑第四工程局有限公司 基于管网追踪的污水监控方法及系统
CN116662767B (zh) * 2023-08-01 2023-10-13 陕西中科绿能能源研究院有限公司 基于多传感器的蒸发冷却机组系统温度数据智能采集方法
CN116662767A (zh) * 2023-08-01 2023-08-29 陕西中科绿能能源研究院有限公司 基于多传感器的蒸发冷却机组系统温度数据智能采集方法
CN117057819A (zh) * 2023-08-15 2023-11-14 泰华智慧产业集团股份有限公司 一种雨水管网污水排放溯源分析方法及系统
CN117034044A (zh) * 2023-10-09 2023-11-10 山东朝启电子科技有限公司 一种基于数据智能的智能表井管理系统
CN117034044B (zh) * 2023-10-09 2024-02-02 山东朝启电子科技有限公司 一种基于数据智能的智能表井管理系统
CN117522199A (zh) * 2023-11-08 2024-02-06 重庆大学 一种城市供水管网综合安全评价及监测点选址方法
CN117610793A (zh) * 2024-01-22 2024-02-27 成都秦川物联网科技股份有限公司 一种智能水表远程维护方法、装置及物联网系统
CN117610793B (zh) * 2024-01-22 2024-04-12 成都秦川物联网科技股份有限公司 一种智能水表远程维护方法、装置及物联网系统
CN117805338A (zh) * 2024-03-01 2024-04-02 广东省建筑设计研究院有限公司 一种建筑供水管网水质实时在线监测方法及系统
CN117805338B (zh) * 2024-03-01 2024-05-28 广东省建筑设计研究院有限公司 一种建筑供水管网水质实时在线监测方法及系统

Also Published As

Publication number Publication date
CN112182984B (zh) 2022-04-26
CN112182984A (zh) 2021-01-05

Similar Documents

Publication Publication Date Title
WO2022036820A1 (zh) 一种基于供水物联网数据同化的污水管网实时模拟方法
US20230205943A1 (en) Real-time simulation method of sewage pipe network based on water supply iot data assimilation
US10795382B2 (en) Method and apparatus for model-based control of a water distribution system
WO2023284060A1 (zh) 一种基于地理三维信息的污水管网流量不确定性分析方法
KR100828968B1 (ko) 지리정보시스템과 연계된 하수관거 유지관리방법 및 그기능을 탑재한 하수관거 유지관리시스템
CN110168339A (zh) 子网络水力建模的系统和方法
KR101105192B1 (ko) 관망해석에 의한 상수도 누수혐의구간 선정방법 및 그 기능을 탑재한 상수도 통합관리운영시스템
SG188534A1 (en) System and method for water distribution modelling
CN108197725B (zh) 基于需水量先验信息的供水管网节点需水量校核方法
CN110010209B (zh) 一种二次供水储水设备出口余氯浓度的计算方法
CN113569438B (zh) 基于多源降雨融合和实时校正的城市洪涝模型构建方法
JP2023169099A (ja) 流入量予測システム
CN113626959B (zh) 一种基于三维地理信息的污水管网水力模型建立方法
CN112632733B (zh) 面向监测信息最大化的供水管网新增流量计的布置方法
CN103617563A (zh) 一种基于地统计空间分析理论的供水管网无监测节点压力确定方法
CN108615098A (zh) 基于贝叶斯生存分析的供水管网管道爆管风险预测方法
CN110569571A (zh) 基于极限学习机的城市供水管网爆管预警方法
CN113221440B (zh) 一种排水系统监测点优化布置与实时全局反演方法
CN112949221A (zh) 一种龙头水龄分析方法
CN115270372A (zh) 一种基于深度序列模型的排水管网淤积判断方法
CN113781276B (zh) 一种污水管网流量不确定性分析方法
JP4326408B2 (ja) 下水流入量予測装置
CN116150930B (zh) 基于3d地图信息的供水管网水力模型节点流量初始化方法
Lauer Water quality in the distribution system
JP3903653B2 (ja) 配水施設制御装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20950024

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20950024

Country of ref document: EP

Kind code of ref document: A1