Intelligent monitoring system for water quality of water supply network

Abstract

The invention relates to the technical field of pipe network water quality monitoring, in particular to a water supply pipe network water quality intelligent monitoring system. The system comprises: a spatial feature analysis module, which is used for collecting the water age of each monitoring point in the water supply network and various water quality parameters at each moment; acquiring a spatial consistency index of the target monitoring point at the moment by utilizing the numerical value variation amplitude of various water quality parameters of the target monitoring point and upstream and downstream adjacent monitoring points at the same moment; the flow velocity regulation factor acquisition module is used for acquiring a flow velocity regulation factor of the target monitoring point at one moment; the real abnormal index acquisition module is used for acquiring a real abnormal index of the target monitoring point at one moment; and the abnormal score correction module is used for correcting the initial abnormal score of the target monitoring point at the moment by using the flow velocity regulation factor and the real abnormal index of the target monitoring point at the moment to obtain a corrected abnormal score. The water quality of the water supply network can be accurately monitored.

Description

Technical Field

[0001] The present invention relates to the technical field of pipe network water quality monitoring, and in particular to an intelligent water quality monitoring system for a water supply pipe network. Background Art

[0002] Water quality monitoring in water supply networks is not only a basic measure to ensure residents' drinking water safety, but also an important part of urban infrastructure management and environmental protection. In this process, monitoring data serves as basic information for assessing water quality, judging abnormal events, and driving intelligent scheduling. Its accuracy and credibility directly affect the system's operating efficiency and public safety. If the monitoring data contains false alarms, false alarms, or missed alarms, it may lead to misjudgment of water pollution or delayed treatment, bringing water supply risks. On the other hand, frequent false alarms will also cause unnecessary resource investment, increase the operation and maintenance costs of on-site inspections and equipment maintenance, and reduce system response efficiency. Therefore, how to improve the accuracy and credibility of water quality monitoring data and reduce false alarms and misjudgments are key goals for improving the performance of the monitoring system.

[0003] Traditionally, water quality monitoring points are deployed in key locations within water supply networks, such as outlets, main pipelines, regional demarcation points, peripheral pipe networks, and key user areas, to collect real-time data on water quality parameters such as pH, residual chlorine, turbidity, and conductivity. However, since water supply networks typically cover an entire city or region, monitoring points are located in diverse environments, including underground wells, pumping stations, water tanks, and at the intersections of residential areas. These locations are subject to complex conditions such as large temperature fluctuations, water pressure shocks, and electromagnetic interference. These conditions can easily lead to sensor probe contamination, aging, signal loss, or drift, potentially causing data changes that resemble actual water quality anomalies. Urban water supply networks are home to a large number of monitoring devices, but sampling frequencies are typically minute-by-minute, with maintenance cycles of weeks or even months. Consequently, instrument anomalies often cannot be promptly verified on-site. Relying solely on anomaly detection algorithms (such as the Isolation Forest algorithm) may fail to effectively distinguish whether data anomalies are caused by actual water pollution or by issues with the monitoring equipment. Consequently, false alarms lead to unnecessary maintenance and spot checks, increasing operating costs. Summary of the Invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a water quality intelligent monitoring system for water supply network, and the technical solutions adopted are as follows: One embodiment of the present invention provides a water quality intelligent monitoring system for a water supply network, the system comprising: The spatial feature analysis module is used to collect the water age and various water quality parameters at each monitoring point in the water supply network at each moment; obtain the numerical variation range of various water quality parameters at each monitoring point at each moment, and use the numerical variation range of various water quality parameters at a target monitoring point and its upstream and downstream adjacent monitoring points at a moment to obtain the spatial consistency index of the target monitoring point at that moment; A flow rate adjustment factor acquisition module is used to obtain the flow rate adjustment factor of the target monitoring point at a moment based on the spatial consistency index of the target monitoring point, the water flow velocity, and the water age of the target monitoring point, the upstream adjacent monitoring point of the target monitoring point, and the last monitoring point in the water supply network; A true abnormality index acquisition module is used to obtain the true abnormality index of the target monitoring point at a moment based on various water quality parameters of the target monitoring point at a moment, various water quality parameters at the moment before the moment, and various water quality parameters at historical moments; The anomaly score correction module is used to obtain the initial anomaly score of the target monitoring point at a moment; the initial anomaly score of the target monitoring point at a moment is corrected by using the flow rate adjustment factor and the true anomaly index of the target monitoring point at a moment to obtain a corrected anomaly score.

[0005] Preferably, obtaining the numerical variation range of various water quality parameters at each monitoring point at each moment includes: The difference between a water quality parameter at a monitoring point at a moment and the water quality parameter at the moment before the moment is obtained as the numerical variation range of the water quality parameter at the monitoring point at the moment.

[0006] Preferably, the spatial consistency index of the target monitoring point at a moment is obtained by using the numerical variation range of various water quality parameters of the target monitoring point and its upstream adjacent monitoring point and downstream adjacent monitoring point at a moment, including: Obtain the absolute value of the difference in the numerical change amplitude of a water quality parameter between an upstream adjacent monitoring point of the target monitoring point and the target monitoring point at the same time, and record it as the numerical change amplitude difference corresponding to the water quality parameter at the upstream adjacent monitoring point at the time, and sum the numerical change amplitude differences corresponding to various water quality parameters of all upstream adjacent monitoring points of the target monitoring point at the time to obtain the upstream cumulative difference; obtain the absolute value of the difference in the numerical change amplitude of a water quality parameter between the target monitoring point and a downstream adjacent monitoring point of the target monitoring point at the same time, and record it as the numerical change amplitude difference corresponding to the water quality parameter at the downstream adjacent monitoring point at the time; sum the numerical change amplitude differences corresponding to various water quality parameters of all downstream adjacent monitoring points of the target monitoring point at the time to obtain the downstream cumulative difference; use the first preset value to subtract the normalized value of the sum of the upstream cumulative difference and the downstream cumulative difference to obtain the spatial consistency index of the target monitoring point at the time.

[0007] Preferably, the flow rate adjustment factor of the target monitoring point at a moment is obtained based on the spatial consistency index of the target monitoring point at a moment, the water flow velocity, and the water age of the target monitoring point, the upstream adjacent monitoring point of the target monitoring point, and the last monitoring point in the water supply network, including: Obtain the difference between the water age of the target monitoring point and the average water age of all upstream adjacent monitoring points of the target monitoring point, and then divide it by the sum of the water age of the last monitoring point in the water supply network and the hyperparameter to obtain the water age characteristic value; obtain the sum of the water flow velocity at a moment in the target monitoring point and the hyperparameter, and then divide it by the maximum value of the water flow velocity of all monitoring points at that moment to obtain the water flow velocity characteristic value; multiply the spatial consistency index of the target monitoring point at that moment by the ratio of the water age characteristic value to the water flow velocity characteristic value to obtain the flow velocity adjustment factor of the target monitoring point at that moment.

[0008] Preferably, obtaining the true abnormality index of the target monitoring point at a moment according to various water quality parameters of the target monitoring point at a moment, various water quality parameters at a moment before the moment, and various water quality parameters at historical moments includes: Compare the difference between a water quality parameter at a target monitoring point at a moment and the water quality parameter at the moment before that moment, and the water quality parameter at the moment before that moment, and take the absolute value to obtain the growth rate of the water quality parameter at that moment; average the growth rates of various water quality parameters at that moment to obtain the average growth rate corresponding to that moment; obtain the absolute value of the difference between a water quality parameter at the target monitoring point at that moment and the mean of the water quality parameter at each historical moment to obtain the deviation degree of the water quality parameter at that moment; average the deviation degrees of various water quality parameters at that moment to obtain the average deviation degree corresponding to that moment; add the average deviation degree and the average growth rate to obtain the true abnormality index of the target monitoring point at that moment.

[0009] Preferably, the initial anomaly score of the target monitoring point at a moment is corrected using the flow rate adjustment factor and the true anomaly index at the target monitoring point at that moment to obtain a corrected anomaly score, including: The normalized value of the flow rate adjustment factor and the normalized value of the true anomaly index at the target monitoring point at a moment are added and averaged to obtain the true anomaly factor of the target monitoring point at that moment; an anomaly attribution function is constructed, and the true anomaly factor of the target monitoring point at that moment is substituted into the anomaly attribution function to obtain the adjustment coefficient; and the corrected anomaly score of the target monitoring point at that moment is obtained based on the adjustment coefficient and the initial anomaly score of the target monitoring point at that moment.

[0010] Preferably, the abnormal attribution function is specifically:

[0011] in, represents the adjustment factor; represents the true abnormal factor of the target monitoring point A at time t; a represents the set threshold.

[0012] Preferably, obtaining the corrected anomaly score of the target monitoring point at that moment according to the adjustment coefficient and the initial anomaly score of the target monitoring point at that moment includes: The adjustment coefficient is multiplied by the initial anomaly score at that moment and normalized to obtain the corrected anomaly score at that moment.

[0013] The embodiments of the present invention have at least the following beneficial effects: the present application collects the water age and various water quality parameters at each monitoring point in the water supply network at each moment, thereby obtaining the numerical variation range of various water quality parameters at each monitoring point at each moment, and then analyzes the numerical variation range of the target monitoring point and its adjacent monitoring points to obtain the spatial consistency index of the target monitoring point at each moment; obtains the flow rate adjustment factor of the target monitoring point at a moment based on the spatial consistency index, water flow velocity, and water age of the target monitoring point and the water age of its upstream adjacent monitoring points at a moment; then uses the various water quality parameters of the target monitoring point at each moment to analyze and obtain the true abnormality index of the target monitoring point at a moment; finally obtains the initial abnormality score of the target monitoring point at a moment; uses the flow rate adjustment factor and the true abnormality index at a moment of the target monitoring point to correct the initial abnormality score of the target monitoring point at that moment to obtain a corrected abnormality score. The present application can effectively distinguish between true abnormalities and abnormalities caused by monitoring instrument problems, thereby improving the accuracy of water quality monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0015] Figure 1 This is a system block diagram of an intelligent water quality monitoring system for a water supply network provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0016] To further illustrate the technical means and effectiveness of the present invention in achieving its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features, and effectiveness of an intelligent water quality monitoring system for a water supply network proposed in accordance with the present invention. In the following description, references to different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0017] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0018] The specific scheme of the intelligent water quality monitoring system for a water supply network provided by the present invention is described in detail below with reference to the accompanying drawings.

[0019] Embodiment, the main application scenario of the present invention is: This application mainly analyzes the water quality of each monitoring point in the water supply network, thereby achieving the purpose of water quality monitoring.

[0020] See also Figure 1 , which shows a system block diagram of a water quality intelligent monitoring system for a water supply network provided by an embodiment of the present invention. The system includes the following modules: The spatial feature analysis module is used to collect the water age and various water quality parameters at each monitoring point in the water supply network at each moment; obtain the numerical change range of various water quality parameters at each monitoring point at each moment, and use the numerical change range of various water quality parameters at a target monitoring point and its upstream adjacent monitoring point and downstream adjacent monitoring point at a moment to obtain the spatial consistency index of the target monitoring point at that moment.

[0021] Obtain various water quality parameters at each monitoring point in the target area. Various water quality parameters are obtained through various monitoring sensors at the monitoring points. Generally, these parameters include the following conventional water quality monitoring parameters, such as pH value, residual chlorine, turbidity, temperature, etc.; usually, the sampling frequency of monitoring points in the same area is the same. If there is a difference in sampling frequency between monitoring points, the sampling frequency is unified, such as once a day.

[0022] Furthermore, the water age of each monitoring point is obtained and updated periodically. Water age refers to the time it takes for water to enter the water supply network, stay in the network, and reach a certain monitoring point. The specific update cycle can be set according to the scenario and actual needs. Water age usually changes slowly. In the water quality monitoring of the water supply network, it only needs to be updated once a week. If a more accurate water age is needed, the update cycle can be shortened). The method of obtaining water age can also be obtained according to the actual scenario. This application provides a conventional acquisition method for reference: a digital model of the water supply network is established through simulation software (such as EPANET), and actual water supply network parameters are set, such as the length, diameter, roughness of each pipe section, network topology, etc., water plant outlet flow, node water demand, etc., to obtain the water age of each monitoring point.

[0023] At the same time, the collected data needs to be processed, mainly including missing value processing and normalization. Missing value processing includes filling missing values ​​through interpolation methods such as sliding mean. Normalization is used to ensure that various water quality parameters have the same data scale to avoid distortion of subsequent analysis and calculation results. At the same time, one monitoring point to be analyzed is recorded as the target monitoring point.

[0024] In the water supply network, the water inside the pipes flows continuously, and water pollutants will naturally diffuse and spread along the water flow. Therefore, when water quality abnormalities occur, changes in water quality parameter data will not only occur at the target monitoring point, but will be based on the flow direction and continuity of water in the water supply pipe. The upstream and downstream adjacent monitoring points of the target monitoring point will show a relatively consistent change trend with the target monitoring point; for upstream and downstream adjacent monitoring points, such as ABCD, the upstream adjacent monitoring point of monitoring point B is monitoring point A, and the downstream adjacent monitoring point is monitoring point C. In fact, there may be multiple adjacent upstream and downstream monitoring points of monitoring point B in the water supply network, but they all represent the previous and next adjacent monitoring points of monitoring point B.

[0025] In addition, due to the differences in pipe network structure, water age, pressure head, etc. among various monitoring points, the water quality parameters are naturally inconsistent. For example, the residual chlorine value is higher at the point close to the water plant, and the residual chlorine value is lower at the point close to the end. Its static water quality data value (that is, the data value corresponding to a single moment) is affected by the spatial differences in the water supply network. Therefore, the difference between a water quality parameter at a monitoring point at a moment and the moment before that moment is used as the dynamic numerical change amplitude of the water quality parameter, and then the difference in the numerical change amplitude of the parameter between the downstream adjacent monitoring point and the target monitoring point and the difference in the numerical change amplitude of the parameter between the upstream adjacent monitoring point and the target monitoring point are calculated respectively, and their absolute values ​​are accumulated and summed. Similarly, the calculation formula for the downstream adjacent monitoring point and the target monitoring point is obtained, and the spatial consistency index of the water quality parameter of the target monitoring point is obtained comprehensively.

[0026] The method for obtaining the numerical variation range of various water quality parameters at each monitoring point at each moment is specifically as follows: obtaining the difference between a water quality parameter at a monitoring point at a moment and the water quality parameter at the moment before the moment, as the numerical variation range of the water quality parameter at the monitoring point at that moment.

[0027] Furthermore, the spatial consistency index of the target monitoring point at that moment is obtained by using the numerical variation range of various water quality parameters of the target monitoring point and its upstream adjacent monitoring points and downstream adjacent monitoring points at the same moment. Specifically, the absolute value of the difference between the numerical variation range of a water quality parameter of an upstream adjacent monitoring point of the target monitoring point and the target monitoring point at a moment is obtained, recorded as the numerical variation range difference corresponding to the water quality parameter of the upstream adjacent monitoring point at that moment, and the numerical variation range differences corresponding to various water quality parameters of all upstream adjacent monitoring points of the target monitoring point at that moment are summed up to obtain the upstream cumulative difference; the absolute value of the difference between the numerical variation range of a water quality parameter of the target monitoring point and a downstream adjacent monitoring point of the target monitoring point at the same moment is obtained, recorded as the numerical variation range difference corresponding to the water quality parameter of the downstream adjacent monitoring point at that moment; the numerical variation range differences corresponding to various water quality parameters of all downstream adjacent monitoring points of the target monitoring point at that moment are summed up to obtain the downstream cumulative difference; the spatial consistency index of the target monitoring point at that moment is obtained by subtracting the normalized value of the sum of the upstream cumulative difference and the downstream cumulative difference from the first preset value.

[0028] The specific calculation model of the spatial consistency index is: , in, It represents the spatial consistency index of water quality parameters at the target monitoring point A at time t. N represents the number of types of water quality parameters currently being monitored. Norm() represents normalization, limiting the output result to (0,1). The spatial consistency index indicates the degree of consistency between the changing trends of various water quality parameters at the target monitoring point and those of its upstream and downstream monitoring points, reflecting whether the water quality at the target monitoring point presents spatial propagation characteristics in the water supply network. A larger output value indicates a stronger spatial consistency of water quality parameter changes. U represents the number of upstream adjacent monitoring points of the target monitoring point, Indicates the target monitoring point The numerical variation range of the jth water quality parameter at the adjacent upstream monitoring point at time t is: Indicates the numerical variation of the j-th water quality parameter at the target monitoring point at time t, The target monitoring point The difference in the numerical change amplitude of the j-th water quality parameter corresponding to the upstream adjacent monitoring point at time t is used to measure the relative difference between the change trend of the upstream adjacent monitoring point and the change trend of the target monitoring point. Cumulative difference for upstream; D represents the number of downstream adjacent monitoring points of the target monitoring point, Indicates the target monitoring point The numerical variation range of the jth water quality parameter at the adjacent downstream monitoring point at time t is: The target monitoring point The difference in the numerical change amplitude of the j-th water quality parameter corresponding to the downstream adjacent monitoring point at time t is used to measure the relative difference between the change trend of the downstream adjacent monitoring point and the change trend of the target monitoring point. The upstream cumulative difference and the downstream cumulative difference are used to indicate the consistency of the data change amplitude between the upstream and downstream adjacent monitoring points and the target monitoring point. The first preset value is 1. Thus, the spatial consistency index of the target monitoring point at time t can be obtained.

[0029] The flow rate adjustment factor acquisition module is used to obtain the flow rate adjustment factor of the target monitoring point at a moment based on the spatial consistency index, water flow velocity, and water age of the target monitoring point, the upstream adjacent monitoring point of the target monitoring point, and the last monitoring point in the water supply network.

[0030] In a water supply network, water quality parameters change as water flows through the network, especially due to natural attenuation processes such as residual chlorine consumption and microbial growth. This attenuation process is closely related to the residence time of water in the network (i.e., water age increment). The longer the water stays in the network, the more the disinfectant (such as residual chlorine) added to the water supply system will naturally decay over time, making water quality more susceptible to abnormalities. Therefore, the water age of the target monitoring point is analyzed with the average water age of all its upstream adjacent monitoring points, and the difference is calculated to reflect the water age increment of the target monitoring point water in the pipe network. At the same time, due to the possible differences in the diameter of the water supply pipe network and the actual path may be more complex and the flow state may be unstable (such as local backflow, dead corners, and turbulence), the actual water flow velocity in each part may vary greatly due to its structural reasons. A long water age increment may indicate that the water flows slowly, or it may also indicate that this section of the path is long. However, if the current point flow velocity is fast, it means that the water body is in continuous renewal. Even if there is a certain water age increment, the possibility of water quality abnormality is still weak. Therefore, the ratio of the water flow velocity of the target monitoring point to the maximum water flow velocity of the entire network is used to reflect the relative size of the water flow velocity of the target monitoring point. Compared with the relative residence time of the water at the target monitoring point in the pipe network, the "water age increment" and the "water flow velocity" are coupled to reflect the credibility of the water quality abnormality at the current monitoring point.

[0031] The flow rate adjustment factor of the target monitoring point at a moment is obtained based on the spatial consistency index, water flow velocity, and water age of the target monitoring point, the upstream adjacent monitoring point of the target monitoring point, and the last monitoring point in the water supply network. Specifically, the difference between the water age of the target monitoring point and the average water age of all upstream adjacent monitoring points of the target monitoring point is obtained, and then the water age characteristic value is obtained by dividing it by the sum of the water age of the last monitoring point in the water supply network and the hyperparameter; the sum of the water flow velocity and the hyperparameter at a moment of the target monitoring point is obtained, and then the water flow velocity characteristic value is obtained by dividing it by the maximum value of the water flow velocity of all monitoring points at that moment; the flow rate adjustment factor of the target monitoring point at that moment is obtained by multiplying the spatial consistency index of the target monitoring point at that moment by the ratio of the water age characteristic value and the water flow velocity characteristic value.

[0032] The calculation model of the flow rate adjustment factor of the target monitoring point is as follows: , in, represents the velocity adjustment factor based on spatial consistency at the target monitoring point A at time t; Indicates the water age of target monitoring point A, represents the average water age of the i-th upstream adjacent monitoring point among all upstream adjacent monitoring points of the target monitoring point, Reflects the water age increment of the water body at the target monitoring point, Indicates the water age of the last monitoring point in the entire water supply network. It represents the water age characteristic value, α is a hyperparameter, which is a very small value to prevent the denominator from being 0. The reference setting is The water age characteristic value reflects the proportion of the water age increment of the target monitoring point water body in the entire water supply network. The size of the ratio is positively correlated with the water age increment from the upstream adjacent monitoring point to the target monitoring point. represents the water flow velocity at the target monitoring point A at time t, It represents the maximum value of the water flow velocity at all monitoring points at time t, is the characteristic value of water flow velocity, which indicates the proportion of water flow velocity at the target monitoring point in the entire pipe network; The ratio of these two parts represents the water retention intensity corresponding to the water flow renewal capacity under the water flow rate of the target monitoring point. The larger the result, the more likely the water at that point is to accumulate, the longer the retention time is, and the more likely the water quality is to be truly abnormal. However, the ratio of these two parts only evaluates the credibility of water quality anomalies from the perspective of hydraulic conditions. Even if the flow rate at a certain monitoring point is slow and the water age is high, it cannot be ruled out that its anomaly is caused by non-real water quality change factors such as equipment drift and local disturbance. In order to further improve the ability to identify real water quality anomalies, it is multiplied by the spatial consistency index of the target monitoring point to evaluate whether there is a synergistic trend in the water quality flow characteristics of the target monitoring point upstream and downstream, and the final output result is obtained comprehensively. .

[0033] The real abnormality index acquisition module is used to obtain the real abnormality index of the target monitoring point at a moment based on various water quality parameters of the target monitoring point at a moment, various water quality parameters at the moment before the moment, and various water quality parameters at historical moments.

[0034] The above method obtains the flow rate adjustment factor at the target monitoring point at time t. Further, it is necessary to conduct a coordinated analysis of the changes in multiple parameters of the target monitoring point to obtain the true anomaly index.

[0035] The real abnormality index of the target monitoring point at a moment is obtained based on the various water quality parameters of the target monitoring point at a moment, the various water quality parameters of the moment before the moment, and the various water quality parameters of the historical moments. Specifically, the difference between a water quality parameter of the target monitoring point at a moment and the water quality parameter of the moment before the moment and the water quality parameter of the moment before the moment are compared, and the absolute value is taken to obtain the growth rate of the water quality parameter at the moment; the growth rates of the various water quality parameters at the moment are averaged to obtain the average growth rate corresponding to the moment; the absolute value of the difference between a water quality parameter of the target monitoring point at the moment and the mean of the water quality parameter of the moment in history is obtained to obtain the deviation degree of the water quality parameter at the moment; the deviation degrees of the various water quality parameters at the moment are averaged to obtain the average deviation degree corresponding to the moment; the average deviation degree and the average growth rate are added to obtain the real abnormality index of the target monitoring point at the moment.

[0036] The calculation model of the true anomaly index is as follows: , in, The real anomaly index is obtained by collaborative calculation of multiple parameters at the target monitoring point A at time t. It indicates whether the changes of other water quality parameters have collaborative characteristics with the abnormal water quality parameter at time t. The larger the output value, the more dimensions of the multiple water quality parameters at the target monitoring point have significant abnormal behaviors at time t, and the more collaborative the abnormal behaviors are among the multiple parameters, the more likely it is that there is an actual water quality problem. N represents the number of types of water quality parameters, represents the jth water quality parameter at the target monitoring point at time t, represents the j-th water quality parameter of the target monitoring point at time t-1, which is also the j-th water quality parameter at the previous moment. The size of represents the fluctuation range of the j-th water quality parameter at time t and the data at the previous moment. It represents the growth rate of the jth water quality parameter at time t. Because the water quality parameters themselves have spatial and site differences in the water supply network structure, the simple numerical changes cannot be directly compared between the monitoring points. It is a growth rate calculation method that uses the growth rate to reflect the change of data; finally, the growth rates of various water quality parameters at time t are averaged to obtain the average growth rate corresponding to that time. ,The larger the value is, the more water quality parameters at the target monitoring point at that moment have undergone significant changes; Represents the average value of the j-th water quality parameter of the target monitoring point at each moment in history, that is, the average value of the j-th water quality parameter at each moment before time t; because the water quality parameter abnormality caused by abnormal water quality behavior is usually a gradual change process after the start, the change rate of adjacent data in this stage may not be drastic, that is, the output result of the previous item can detect the data change behavior at the beginning of the abnormality, but may not be able to effectively respond after the start. Therefore, the normal mode of the j-th water quality parameter under normal conditions is represented by its mean value for the previous period of historical data, and then the difference between time t and its normal mode is calculated to reflect the data change after the start of the abnormal behavior. The historical data referenced in this application is the historical data of the past month, which can be adjusted according to actual conditions; is the deviation degree corresponding to the j-th water quality parameter at time t, is the average deviation degree corresponding to time t.

[0037] Thus, the true abnormality index of the target monitoring point at a moment is obtained.

[0038] The anomaly score correction module is used to obtain the initial anomaly score of the target monitoring point at a moment; the initial anomaly score of the target monitoring point at a moment is corrected by using the flow rate adjustment factor and the true anomaly index of the target monitoring point at a moment to obtain a corrected anomaly score.

[0039] The main purpose of this application is to correct the initial anomaly score obtained by isolation forest. Therefore, it is necessary to use isolation forest to obtain the initial anomaly score of each monitoring point at each time. Specifically, the various water quality parameters of a monitoring point at a time are taken as a sample, and then multiple isolation trees are constructed. The samples are divided in each tree, and the path length of each sample in each tree is calculated to obtain the average path length. The initial anomaly score of each monitoring point at each time is calculated based on the average path length and the anomaly score calculation formula. It should be noted that this method is a well-known technology and will not be elaborated on here.

[0040] The above obtains the flow rate adjustment factor of the target monitoring point at a certain moment. The flow rate adjustment factor indicates the degree of coordinated trend of the water quality flow characteristics of the target monitoring point in its upstream and downstream. At the same time, since the water quality parameter anomalies caused by the water quality itself are usually manifested in the coordinated changes of multiple water quality parameters, and the water quality parameter changes caused by the monitoring instrument usually only occur in a single water quality parameter data, the real anomaly index obtained by multi-parameter collaborative calculation combined with the flow rate adjustment factor can obtain the real anomaly factor of the water quality parameter at a certain moment, and then construct the anomaly attribution function, and adjust the initial anomaly score according to the function output result.

[0041] Thus, the initial anomaly score of the target monitoring point at a certain moment is corrected using the velocity adjustment factor and the true anomaly index at that moment to obtain a corrected anomaly score. Specifically, the normalized value of the velocity adjustment factor and the normalized value of the true anomaly index at the target monitoring point at that moment are added and averaged to obtain the true anomaly factor of the target monitoring point at that moment. An anomaly attribution function is constructed, and the true anomaly factor of the target monitoring point at that moment is substituted into the anomaly attribution function to obtain an adjustment coefficient. The corrected anomaly score of the target monitoring point at that moment is obtained based on the adjustment coefficient and the initial anomaly score of the target monitoring point at that moment.

[0042] The specific anomaly attribution function is:

[0043] in, Represents the output value of the anomaly attribution function, which is also the adjustment coefficient of the target monitoring point A at time t; represents the actual abnormal factor of the target monitoring point A at time t; a represents the set threshold value, and the reference value of this application is 0.65. Implementers can adjust it according to actual conditions and actual needs.

[0044] Finally, the adjusted anomaly score for the target monitoring point at time t is calculated based on the adjustment coefficient and the initial anomaly score at the target monitoring point at that time. Specifically, the adjusted anomaly score at time t is multiplied by the adjustment coefficient and normalized. A higher anomaly score indicates a greater likelihood of water quality anomaly.

[0045] The above method obtains the real corrected anomaly score. Through the corrected anomaly score, various water quality parameters of the target monitoring points in the actual water supply network are monitored, which improves the monitoring system's ability to identify data abnormal behaviors caused by real water quality anomalies, reduces system false alarms caused by anomalies caused by monitoring equipment, effectively reduces unnecessary operation and maintenance and spot checks caused by false alarms, increases operating costs, improves the system's perception and judgment capabilities of real water pollution incidents, and improves water quality safety management and operation and maintenance optimization.

[0046] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0047] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0048] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An intelligent water quality monitoring system for a water supply network, characterized in that: The system includes: The spatial feature analysis module is used to collect the water age and various water quality parameters at each monitoring point in the water supply network at each moment; obtain the numerical variation range of various water quality parameters at each monitoring point at each moment, and use the numerical variation range of various water quality parameters at a target monitoring point and its upstream and downstream adjacent monitoring points at a moment to obtain the spatial consistency index of the target monitoring point at that moment; A flow rate adjustment factor acquisition module is used to obtain the flow rate adjustment factor of the target monitoring point at a moment based on the spatial consistency index of the target monitoring point, the water flow velocity, and the water age of the target monitoring point, the upstream adjacent monitoring point of the target monitoring point, and the last monitoring point in the water supply network; A true abnormality index acquisition module is used to obtain the true abnormality index of the target monitoring point at a moment based on various water quality parameters of the target monitoring point at a moment, various water quality parameters at the moment before the moment, and various water quality parameters at historical moments; The anomaly score correction module is used to obtain the initial anomaly score of the target monitoring point at a moment; the initial anomaly score of the target monitoring point at a moment is corrected by using the flow rate adjustment factor and the true anomaly index of the target monitoring point at a moment to obtain a corrected anomaly score.

2. The water quality intelligent monitoring system for a water supply network according to claim 1, characterized in that: The method of obtaining the numerical variation range of various water quality parameters at each monitoring point at each time includes: The difference between a water quality parameter at a monitoring point at a moment and the water quality parameter at the moment before the moment is obtained as the numerical variation range of the water quality parameter at the monitoring point at the moment.

3. The water quality intelligent monitoring system for a water supply network according to claim 1, characterized in that: The method of obtaining the spatial consistency index of the target monitoring point at a moment by using the numerical variation range of various water quality parameters of the target monitoring point and its upstream adjacent monitoring point and downstream adjacent monitoring point at a moment includes: Obtain the absolute value of the difference in the numerical change amplitude of a water quality parameter between an upstream adjacent monitoring point of the target monitoring point and the target monitoring point at the same time, and record it as the numerical change amplitude difference corresponding to the water quality parameter at the upstream adjacent monitoring point at the time, and sum the numerical change amplitude differences corresponding to various water quality parameters of all upstream adjacent monitoring points of the target monitoring point at the time to obtain the upstream cumulative difference; obtain the absolute value of the difference in the numerical change amplitude of a water quality parameter between the target monitoring point and a downstream adjacent monitoring point of the target monitoring point at the same time, and record it as the numerical change amplitude difference corresponding to the water quality parameter at the downstream adjacent monitoring point at the time; sum the numerical change amplitude differences corresponding to various water quality parameters of all downstream adjacent monitoring points of the target monitoring point at the time to obtain the downstream cumulative difference; use the first preset value to subtract the normalized value of the sum of the upstream cumulative difference and the downstream cumulative difference to obtain the spatial consistency index of the target monitoring point at the time.

4. The water quality intelligent monitoring system for a water supply network according to claim 1, characterized in that: The method of obtaining the flow rate adjustment factor of the target monitoring point at a moment according to the spatial consistency index of the target monitoring point at a moment, the water flow velocity, and the water age of the target monitoring point, the upstream adjacent monitoring point of the target monitoring point, and the last monitoring point in the water supply network includes: Obtain the difference between the water age of the target monitoring point and the average water age of all upstream adjacent monitoring points of the target monitoring point, and then divide it by the sum of the water age of the last monitoring point in the water supply network and the hyperparameter to obtain the water age characteristic value; obtain the sum of the water flow velocity at a moment in the target monitoring point and the hyperparameter, and then divide it by the maximum value of the water flow velocity of all monitoring points at that moment to obtain the water flow velocity characteristic value; multiply the spatial consistency index of the target monitoring point at that moment by the ratio of the water age characteristic value to the water flow velocity characteristic value to obtain the flow velocity adjustment factor of the target monitoring point at that moment.

5. The water quality intelligent monitoring system for a water supply network according to claim 1, characterized in that: The method of obtaining the true abnormality index of the target monitoring point at a moment according to various water quality parameters of the target monitoring point at a moment, various water quality parameters at a moment before the moment, and various water quality parameters at historical moments includes: Compare the difference between a water quality parameter at a target monitoring point at a moment and the water quality parameter at the moment before that moment, and the water quality parameter at the moment before that moment, and take the absolute value to obtain the growth rate of the water quality parameter at that moment; average the growth rates of various water quality parameters at that moment to obtain the average growth rate corresponding to that moment; obtain the absolute value of the difference between a water quality parameter at the target monitoring point at that moment and the mean of the water quality parameter at each historical moment to obtain the deviation degree of the water quality parameter at that moment; average the deviation degrees of various water quality parameters at that moment to obtain the average deviation degree corresponding to that moment; add the average deviation degree and the average growth rate to obtain the true abnormality index of the target monitoring point at that moment.

6. The water quality intelligent monitoring system for a water supply network according to claim 1, characterized in that: The method of correcting the initial anomaly score of the target monitoring point at a moment by using the velocity adjustment factor and the true anomaly index at the target monitoring point at the moment to obtain the corrected anomaly score includes: The normalized value of the flow rate adjustment factor and the normalized value of the true anomaly index at the target monitoring point at a moment are added and averaged to obtain the true anomaly factor of the target monitoring point at that moment; an anomaly attribution function is constructed, and the true anomaly factor of the target monitoring point at that moment is substituted into the anomaly attribution function to obtain the adjustment coefficient; and the corrected anomaly score of the target monitoring point at that moment is obtained based on the adjustment coefficient and the initial anomaly score of the target monitoring point at that moment.

7. The water quality intelligent monitoring system for a water supply network according to claim 6, characterized in that: The abnormal attribution function is specifically: ， in, represents the adjustment factor; represents the true abnormal factor of the target monitoring point A at time t; a represents the set threshold.

8. The water quality intelligent monitoring system for a water supply network according to claim 6, characterized in that: Obtaining the corrected anomaly score of the target monitoring point at the moment according to the adjustment coefficient and the initial anomaly score of the target monitoring point at the moment includes: The adjustment coefficient is multiplied by the initial anomaly score at that moment and normalized to obtain the corrected anomaly score at that moment.

Images