Intelligent water affair monitoring management system based on Internet of Things

Abstract

The invention relates to the technical field of monitoring management, in particular to an intelligent water affair monitoring management system based on the Internet of Things, the system comprises a processor and a memory, and the processor executes a computer program stored in the memory to realize the following steps: acquiring data correlation coefficients and differential correlation coefficients corresponding to different monitoring moments, obtaining a target abnormal coefficient of the first clustering cluster and a target abnormal coefficient of the second clustering cluster according to a data volume ratio and a variable coefficient in the first clustering cluster and the second clustering cluster obtained by clustering the data correlation coefficient and the difference correlation coefficient corresponding to the monitoring moment; and obtaining a target abnormal index value corresponding to the current monitoring moment according to the data correlation coefficient corresponding to the current monitoring moment and the target abnormal coefficient of the cluster to which the differential correlation coefficient belongs, and monitoring the water affair system according to the target abnormal index value. And the timeliness and the reliability of abnormal monitoring and early warning of the water affair system can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of monitoring management, and particularly relates to a smart water monitoring management system based on Internet of Things. BACKGROUND

[0002] Since the smart monitoring management of the water system based on Internet of Things is closely related to disaster risk prevention, water supply stability guarantee and water resource management efficiency improvement, it is crucial to monitor and manage the water system based on Internet of Things at present. The smart monitoring management of the water system based on Internet of Things can collect multi-dimensional data in real time and upload them to the cloud platform or local management center through the wireless communication network, so as to realize remote monitoring and management.

[0003] In the prior art, the abnormal monitoring and early warning of the water system is usually based on the relationship between the single-dimensional monitoring data related to the water system and the threshold value. However, this monitoring method has the problems of high lag or low timeliness, which may cause resource waste and economic loss, facility damage and other problems. That is, when the abnormality occurs or lasts for a period of time, the abnormal monitoring and early warning of the water system based on the relationship between the single-dimensional monitoring data and the threshold value can be monitored. For example, when there is a potential pipe explosion risk in the pipe network, the relationship between the single-dimensional monitoring data and the threshold value cannot identify the early stage of the slight abnormal fluctuation of water pressure. Only when the monitoring value exceeds the threshold value or the pipe explosion has occurred, the abnormal alarm is triggered. Therefore, how to improve the timeliness of monitoring the water system becomes a problem to be solved. SUMMARY

[0004] In order to solve the above problems, the present application provides a smart water monitoring management system based on Internet of Things, which adopts the following technical solutions: An embodiment of the present application provides a smart water monitoring management system based on Internet of Things, which comprises a processor and a memory. The processor executes the computer program stored in the memory to realize the following steps: obtaining a target water pressure data sequence and a target water flow data sequence corresponding to a monitoring time of a water system; obtaining data correlation coefficients and difference correlation coefficients corresponding to different monitoring times according to the correlation between the target water pressure data sequence and the target water flow data sequence and the correlation between the difference sequence of the target water pressure data sequence and the difference sequence of the target water flow data sequence, and obtaining target abnormal coefficients of a first clustering cluster and a second clustering cluster obtained by clustering the data correlation coefficients and the difference correlation coefficients corresponding to the monitoring time according to the data amount proportion and the variation coefficient in the first clustering cluster and the second clustering cluster; According to the target anomaly coefficient of the cluster to which the data correlation coefficient and the difference correlation coefficient corresponding to the current monitoring moment belong, a target anomaly index value corresponding to the current monitoring moment is obtained, and the water affair system is monitored according to the target anomaly index value.

[0005] Beneficial effects: The application first acquires the target water pressure data sequence and the target water flow data sequence of the water affair system at the monitoring moment; then according to the correlation between the target water pressure data sequence and the target water flow data sequence and the correlation between the difference sequence of the target water pressure data sequence and the difference sequence of the target water flow data sequence, the data correlation coefficient and the difference correlation coefficient corresponding to different monitoring moments are obtained, the target anomaly coefficient of the first cluster and the target anomaly coefficient of the second cluster are obtained according to the data amount proportion and the variation coefficient in the first cluster and the second cluster obtained by clustering the data correlation coefficient and the difference correlation coefficient corresponding to the monitoring moment; then according to the target anomaly coefficient of the cluster to which the data correlation coefficient and the difference correlation coefficient corresponding to the current monitoring moment belong, the target anomaly index value corresponding to the current monitoring moment is obtained, and finally the water affair system is monitored according to the target anomaly index value. And the application can improve the timeliness and reliability of the abnormal monitoring and early warning of the water affair system according to the data correlation coefficient and the difference correlation coefficient. BRIEF DESCRIPTION OF DRAWINGS

[0006] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, and the advantages thereof, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description only represent some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor based on these drawings.

[0007] Figure 1 The flow chart of the present application, a smart water affair monitoring and management method based on the Internet of Things. DETAILED DESCRIPTION

[0008] The technical solutions in the embodiments of the present application will be described clearly and completely in the following with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments only represent some embodiments of the present application, and not all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art belong to the scope of protection of the embodiments of the present application.

[0009] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present application belongs.

[0010] The embodiment provides a smart water monitoring management system based on Internet of Things, which comprises a processor and a memory, and the processor executes a computer program stored in the memory to realize a smart water monitoring management method based on Internet of Things. Figure 1 As shown in the figure, the smart water monitoring management method based on Internet of Things comprises the following steps: Step S001, obtaining a target water pressure data sequence and a target water flow data sequence corresponding to a monitoring moment of a water system.

[0011] The subsequent description of the embodiment will take the abnormal monitoring and early warning process of any water system as an example. Since the water system is a comprehensive water resource management network and comprises multiple links, each link undertakes a specific function and cooperates with each other, in order to more comprehensively monitor and early warn the water system, multiple monitoring positions are usually arranged in each link of the water system, and the purpose of monitoring and early warning the water system is achieved by monitoring and analyzing each monitoring position, that is, the parameters at the monitoring position can reflect the condition of the water system. In addition, since the water system usually processes water in a water plant, pressurizes water in a pump station, transports water in a pipe network and uses water at a user end, the links of the water system comprise water plant processing, pump station pressurization, pipe network transportation and user end, and the arrangement rule of the monitoring positions of the water system is a known technology.

[0012] Since the parameters at the monitoring position that can reflect the condition of the water system comprise water pressure and water flow, the subsequent embodiment needs to monitor and analyze the water pressure and water flow at each monitoring position to achieve the purpose of abnormal detection and early warning of the water system, that is, the subsequent embodiment needs to first obtain the water pressure data and water flow data of the monitoring position at different monitoring moments by collecting the water pressure and water flow at the monitoring position by using a sensor, that is, the sensors for collecting water pressure and collecting water flow are arranged at the same monitoring position, and the water pressure and water flow at the same monitoring position are synchronously collected; since the monitoring and analyzing method of each monitoring position is the same, the subsequent description of the embodiment will take the monitoring and analyzing process of any monitoring position as an example for description, and the monitoring position is recorded as a target monitoring position, that is, the subsequent water pressure and water flow data are all data of the target monitoring position. The data collected by the sensor is uploaded to a cloud platform or a local control center through a wireless communication network, so as to ensure the real-time, integrity and reliability of the collected data, thereby providing a basis for subsequent intelligent analysis, monitoring and early warning.

[0013] In addition, since the correlation between the multi-dimensional parameters of the water system in the abnormal state or the state with abnormal trend is different from the correlation between the multi-dimensional parameters of the water system in the normal state, the embodiment mainly analyzes the correlation between the multi-dimensional parameters related to the running state of the water system to improve the timeliness of the abnormal monitoring and early warning of the water system, that is, the correlation between the water pressure and water flow data of the target monitoring position at different monitoring times is analyzed to improve the timeliness of the abnormal monitoring and early warning of the water system. In order to avoid the contingency of instantaneous data, the embodiment needs to select a period of data for correlation analysis. However, the local change characteristics at different times are different, such as the different water consumption of users in different time periods, which leads to different water pressure data fluctuation characteristics in the water pipe at different times. The local change characteristics are necessarily related to the abnormal detection and early warning. For example, when the change is unstable, a longer data segment should be selected for correlation analysis to capture short-term abnormalities. Therefore, if the same length of data is selected for correlation operation without considering the difference in local change characteristics, it will affect the timeliness of subsequent abnormal monitoring. Therefore, the embodiment needs to obtain the target water pressure data sequence and the target water flow data sequence for correlation analysis based on the local change characteristics of the target monitoring position at different monitoring times, that is, to obtain the target water pressure data sequence and the target water flow data sequence corresponding to the target monitoring position at different monitoring times, which can also be called the target water pressure data sequence and the target water flow data sequence corresponding to the water system at the monitoring time. In addition, for ease of understanding, the specific acquisition process of the target water pressure data sequence and the target water flow data sequence corresponding to the target monitoring position at the monitoring time t will be described in the following embodiment, that is, the specific acquisition process of the target water pressure data sequence and the target water flow data sequence corresponding to the monitoring time t is: Firstly, a time sequence sequence of all water pressure data collected by the sensor at the target monitoring position within a preset initial time length before the monitoring time t and at the monitoring time t is obtained, and is recorded as an initial water pressure data sequence corresponding to the monitoring time t of the target monitoring position. A time sequence sequence of all water flow data collected by the sensor at the target monitoring position within a preset initial time length before the monitoring time t and at the monitoring time t is obtained, and is recorded as an initial water flow data sequence corresponding to the monitoring time t of the target monitoring position. The initial water pressure data sequence and the initial water flow data sequence are mainly used to analyze the local change characteristics at the monitoring time t, and subsequent scale adjustment is based on the analyzed results, that is, the preset initial time length is adjusted. In specific applications, the implementer needs to set the preset initial time length according to historical experience and other actual situations. If the water pressure or water flow at the target monitoring position usually fluctuates greatly in history, the preset initial time length can be set to be shorter. If the water pressure or water flow at the target monitoring position usually fluctuates less in history, the preset initial time length can be set to be longer, so as to ensure the accuracy and reliability of the analysis of the local change characteristics. For example, the preset initial time length in this embodiment can be set to 10 minutes, so that the initial water pressure data sequence and the initial water flow data sequence corresponding to the monitoring time t of the target monitoring position correspond to a time period of ten minutes before the monitoring time t and at the monitoring time t. After obtaining the initial water pressure data sequence and the initial water flow data sequence corresponding to the monitoring time t, the change characteristic value is analyzed based on the difference between the initial water pressure data sequence and the initial water flow data sequence. The preset initial time length is adjusted based on the analysis result, the target data selection time length at the monitoring time t is obtained, and the target water pressure data sequence and the target water flow data sequence corresponding to the monitoring time t are obtained based on the obtained target data selection time length.

[0014] Based on the above analysis, the next embodiment needs to obtain the change characteristic value first, that is, the next embodiment needs to perform first-order difference on the initial water pressure data sequence and the initial water flow data sequence corresponding to the monitoring time t respectively, to obtain the first-order difference sequence of the initial water pressure data sequence corresponding to the monitoring time t and the first-order difference sequence of the initial water flow data sequence corresponding to the monitoring time t, and the rth difference value in any sequence is the result of the r+1th data minus the rth data in the sequence; Since the difference value difference in the difference sequence can reflect the change characteristics, then according to the difference value difference in the first-order difference sequence of the initial water pressure data sequence corresponding to the monitoring time t, the water pressure change characteristic value at the monitoring time t is obtained, and according to the difference value difference in the first-order difference sequence of the initial water flow data sequence corresponding to the monitoring time t, the water flow change characteristic value at the monitoring time t is obtained, and then the preset initial time length is adjusted according to the water pressure change characteristic value and the water flow change characteristic value at the monitoring time t, to obtain the target data selection time length at the monitoring time t, and according to the target data selection time length at the monitoring time t, the target water pressure data sequence and the target water flow data sequence corresponding to the target monitoring position at the monitoring time t are obtained, and the time period corresponding to any sequence is the time period formed from the time when the first data in the corresponding sequence is collected to the time when the last data in the corresponding sequence is collected; The acquisition method of the water flow change characteristic value is the same as the acquisition method of the water pressure change characteristic value, so the subsequent process of obtaining the water flow change characteristic value will not be described in detail.

[0015] The specific process of obtaining the water pressure change characteristic value at the monitoring time t according to the difference value difference in the first-order difference sequence of the initial water pressure data sequence corresponding to the monitoring time t is as follows: First, the first-order difference sequence of the initial water pressure data sequence corresponding to the monitoring time t is denoted as sequence A, and the weighted change index value of each difference value in sequence A is obtained according to the difference between each difference value and its neighbor difference value and the time interval between the collection time of the water pressure data corresponding to each difference value and the monitoring time t, and it needs to be noted that the first difference value and the last difference value in sequence A do not need to be calculated for the weighted change index value, then the weighted change index values of all difference values belonging to sequence A except the first difference value and the last difference value in sequence A are accumulated, and the accumulation result is denoted as an accumulation index value, and the accumulation index value is negatively correlated mapped, and the mapping result is denoted as the water pressure change characteristic value at the monitoring time t, and here the negative correlation mapping is performed by using the negative exponential function with constant e as the base.

[0016] The obtaining process of the weighted change indicator value is as follows: for the i-th difference value in sequence A, i is greater than 1 and less than N, that is, i is not equal to 1 and N, N is the total amount of data in sequence A, the absolute value of the difference between the (i-1)-th difference value and the i-th difference value in sequence A is calculated and recorded as a first difference value, the absolute value of the difference between the i-th difference value and the (i+1)-th difference value in sequence A is calculated and recorded as a second difference value, the sum of the first difference value and the second difference value is recorded as a comprehensive difference value, the absolute value of the difference between the (i-1)-th difference value and the (i+1)-th difference value is recorded as a neighboring difference value, and the ratio of the comprehensive difference value to the neighboring difference value is recorded as the change indicator value of the i-th difference value. The time interval between the collection time of the water pressure data corresponding to the i-th difference value and the current monitoring time is obtained and recorded as the time difference value corresponding to the i-th difference value. The time difference value corresponding to the i-th difference value is negatively correlated and mapped, and the mapping result is recorded as the negative mapping value corresponding to the i-th difference value. The negative mapping value corresponding to the i-th difference value is normalized, and the result of the normalization is recorded as the time distance weight factor of the i-th difference value. The product of the change indicator value of the i-th difference value and the time distance weight factor of the i-th difference value is calculated and recorded as the weighted change indicator value of the i-th difference value. The water pressure data corresponding to the i-th difference value in sequence A is the (i+1)-th water pressure data in the initial water pressure data sequence corresponding to the monitoring time t, that is, the water pressure data corresponding to the first difference value in sequence A is the second water pressure data in the initial water pressure data sequence corresponding to the monitoring time t.

[0017] In addition, the calculation expression of the water pressure change characteristic value at the monitoring time t is:

[0018] Wherein, W1 is the water pressure change characteristic value at the monitoring time t, N is the total amount of data in sequence A, is the i-th difference value, is the (i-1)-th difference value, is the (i+1)-th difference value, is the comprehensive difference value, is the neighboring difference value, exp() is the exponential function with constant e as the base, is the time difference value corresponding to the i-th difference value, is the negative mapping value corresponding to the i-th difference value, normalizes , and the negative mapping value corresponding to any difference value is obtained in the same way as the negative mapping value corresponding to the i-th difference value; , and The phase difference value can reflect the relative difference between the ith difference value and the neighborhood difference value, and can eliminate the difference in calculation results caused by different difference values. Since the change index value closer to the monitoring time t can reflect the change at the monitoring time t, the change index value closer to the monitoring time t can reflect the change of the water pressure at the monitoring time t. Therefore, the embodiment makes the change index value of the difference value closer to the monitoring time t more dominant in the result of the water pressure change characteristic value at the monitoring time t, and the closer the data collection time corresponding to the ith difference value is to the monitoring time t, The greater the product of the above The greater the product of the above W1, and the smaller W1 indicates that the local water pressure change at the monitoring time t is more unstable, and the larger W1 indicates that the local water pressure change at the monitoring time t is more stable.

[0019] According to the water pressure change characteristic value at the monitoring time t and the water flow change characteristic value, the preset initial time length is adjusted to obtain the specific process of the target data selection time length at the monitoring time t: the product of the minimum value in the water pressure change characteristic value at the monitoring time t and the water flow change characteristic value at the monitoring time t and the preset initial time length is calculated, and is recorded as the target data selection time length at the monitoring time t. The expression for obtaining the target data selection time length at the monitoring time t is Wherein, W2 is the water flow change characteristic value at the monitoring time t, min() is the minimum value function, and L is the preset initial time length. The smaller the water pressure change characteristic value or the water flow change characteristic value, the more unstable the local change at the monitoring time t, and the shorter the target data selection time length at the monitoring time t, i.e. the shorter the length of the target water pressure data sequence and the target water flow data sequence corresponding to the monitoring time t. Since the more unstable the local change at a certain monitoring time t is, the shorter data length should be selected at this time to more sensitively capture the local sudden change. Moreover, if a longer data length is used at this time, the short-time anomaly may be covered or the local sudden change may not be captured, affecting the final abnormal monitoring and early warning result. Conversely, when the local change at a certain monitoring time t is more stable, a longer data length should be selected at this time to capture the overall trend and improve the timeliness of abnormal monitoring and early warning. Since the water pressure change characteristic value and the water flow change characteristic value can reflect the stability of the local change, the embodiment adjusts the selected data length based on the water pressure change characteristic value and the water flow change characteristic value, and makes the water pressure change characteristic value or the water flow change characteristic value smaller, and the target data selection time length at the monitoring time t shorter.

[0020] The specific process of obtaining the target water pressure data sequence and the target water flow data sequence corresponding to the monitoring moment t of the target monitoring position is as follows: the target data selection time length at the monitoring moment t is denoted as T, the time period with a length of T before the monitoring moment t is denoted as the target time period before the monitoring moment t, the target time period before the monitoring moment t is adjacent to the monitoring moment t, all the water pressure data collected in the target time period before the monitoring moment t and the water pressure data collected at the monitoring moment t are denoted as a time sequence and are denoted as the target water pressure data sequence corresponding to the monitoring moment t, all the water flow data collected in the target time period before the monitoring moment t and the water flow data collected at the monitoring moment t are denoted as a time sequence and are denoted as the target water flow data sequence corresponding to the monitoring moment t. For example, if the target data selection time length is 15 minutes, all the water pressure data of the target monitoring position collected in the 15 minutes before the monitoring moment t and the water pressure data of the target monitoring position collected at the monitoring moment t form a time sequence, which is the target water pressure data sequence corresponding to the monitoring moment t, and the target water flow data sequence corresponding to the monitoring moment t is the same.

[0021] Therefore, the target water pressure data sequence and the target water flow data sequence corresponding to different monitoring moments of the target monitoring position can be obtained through the above process.

[0022] In step S002, the data correlation coefficients and the difference correlation coefficients corresponding to different monitoring moments are obtained according to the correlation between the target water pressure data sequence and the target water flow data sequence and the correlation between the difference sequence of the target water pressure data sequence and the difference sequence of the target water flow data sequence. The target anomaly coefficient of the first clustering cluster and the target anomaly coefficient of the second clustering cluster are obtained according to the data amount proportion and the variation coefficient in the first clustering cluster and the second clustering cluster obtained by clustering the data correlation coefficients and the difference correlation coefficients corresponding to the monitoring moment.

[0023] Since water pressure and water flow have a close relationship with each other, for example, when the water consumption of a certain area in the pipe network suddenly increases, the water flow will rise, and due to the resistance loss of the pipeline, the local water pressure will tend to fall, and vice versa, when the water consumption decreases, the water flow decreases, and the water pressure of the pipe network will relatively increase, so it can be known that the correlation between the changes of water pressure and water flow or the differential correlation between water pressure and water flow can be used to identify abnormalities or abnormal trends, that is, the correlation between the multi-dimensional parameters of the water system in the abnormal state or the state with abnormal trend is different from the correlation between the multi-dimensional parameters of the water system in the normal state, that is, the correlation between the multi-dimensional parameters of the water system in the abnormal state or the state with abnormal trend will usually deviate from the correlation calculated in the normal state, and since the duration of the normal state is generally longer, the embodiment will first calculate the correlation coefficients at each monitoring time, then cluster, and analyze the abnormal coefficients of the clustering clusters based on the number ratio, coefficient of variation, etc. in the clustering cluster, and subsequently based on the abnormal coefficient of the clustering cluster to which the correlation coefficient at the current monitoring time belongs to obtain the abnormal index value, which can reflect the possibility of existence of abnormality or abnormal trend; based on the above description, it can be known that the embodiment first needs to calculate the correlation coefficients at multiple monitoring times, that is, the data correlation coefficients and differential correlation coefficients corresponding to different monitoring times will be obtained according to the correlation between the target water pressure data sequence and the target water flow data sequence of the target detection position at different monitoring times and the correlation between the first order difference sequence of the target water pressure data sequence and the first order difference sequence of the target water flow data sequence of the target detection position.

[0024] For the sake of understanding and description, the embodiment will be described below with the acquisition process of the data correlation coefficient and the differential correlation coefficient corresponding to the monitoring time t as an example, t cannot be the time after the current monitoring time, nor can it be the initial stage of monitoring the target detection position. The historical data in the initial stage is too small to have little reference value of correlation, for example, t can be set to not be the first ten minutes of the start of monitoring the target detection position, and the specific acquisition process of the data correlation coefficient and the differential correlation coefficient corresponding to the monitoring time t is as follows: The Spearman rank correlation coefficient between the target water pressure data sequence corresponding to the monitoring time t and the target water flow data sequence corresponding to the monitoring time t is calculated and used as the data correlation coefficient corresponding to the monitoring time t, and the Spearman rank correlation coefficient between the first order difference sequence of the target water pressure data sequence corresponding to the monitoring time t and the first order difference sequence of the target water flow data sequence corresponding to the monitoring time t is calculated and used as the differential correlation coefficient corresponding to the monitoring time t. The acquisition process of the Spearman rank correlation coefficient is known.

[0025] Next, the HDBSCAN clustering algorithm is used to cluster the data correlation coefficients corresponding to the monitoring times, and the resulting clusters are denoted as the first cluster. Then, the HDBSCAN clustering algorithm is used to cluster the differential correlation coefficients corresponding to the monitoring times, and the resulting clusters are denoted as the second cluster. Times after the current monitoring time and times belonging to the initial stage of monitoring at the target detection location are not included in this clustering. The HDBSCAN clustering algorithm is a density-based hierarchical clustering algorithm that combines the advantages of DBSCAN and hierarchical clustering. The clustering process of the HDBSCAN clustering algorithm is well-known.

[0026] After clustering is completed, the data volume ratio and coefficient of variation in the first and second clusters are used to obtain the target anomaly coefficients for each first cluster and each second cluster. Since the method for obtaining the target anomaly coefficients for the second clusters is the same as that for the first clusters, this embodiment will only describe the process of obtaining the target anomaly coefficients for the first clusters, and will not describe the specific process of obtaining the target anomaly coefficients for the second clusters. Therefore, the process of obtaining the target anomaly coefficients for the first clusters is as follows: For any first cluster R: First, obtain the ratio of the total number of correlation coefficients in the first cluster R to the total number of correlation coefficients of all data participating in the first cluster, and record it as the count ratio. Then, obtain the coefficient of variation of the first cluster R, which is the ratio of the standard deviation to the mean of all data in the first cluster R. Normalize the coefficient of variation of the first cluster R and then perform negative correlation mapping, recording the result as the mapping coefficient. The product of the count ratio and the mapping coefficient is taken as the normal characteristic value of the first cluster R. The expression for the normal characteristic value of the first cluster R is:

[0027] in, Here, MR is the normal characteristic value of the first cluster R, NR is the total number of data correlation coefficients in the first cluster R, Norm() is the normalization function, and YR is the coefficient of variation of the first cluster R. A higher MR indicates that most of the relationship between water pressure and flow rate reflects the state of this cluster, thus increasing the reliability of the cluster or the probability that it is a normal cluster. Conversely, a lower YR indicates a more stable relationship between water pressure and flow rate, further increasing the reliability of the cluster or the probability that it is a normal cluster. When the size is larger, the size of YR is smaller. The larger, therefore when The larger the value, the greater the reference confidence of the cluster, or the higher the probability that the cluster is a normal cluster; conversely, when the value decreases... The smaller the value, the lower the reliability of the cluster or the lower the probability that the cluster is a normal cluster.

[0028] Then, based on the normal characterization values ​​of the first cluster R and the normal characterization values ​​of the remaining first clusters excluding the first cluster R, the target anomaly coefficient of the first cluster R is obtained; and the specific process of obtaining the target anomaly coefficient of the first cluster R based on the normal characterization values ​​of the first cluster R and the normal characterization values ​​of the remaining first clusters excluding the first cluster R is as follows: First, the set consisting of all first clusters except for the first cluster R is denoted as the set of remaining clusters corresponding to the first cluster R. The sum of the normal representation values ​​of all first clusters in the set of remaining clusters corresponding to the first cluster R is denoted as the first remaining normal representation value corresponding to the first cluster R. The ratio of the normal representation value of each first cluster in the set of remaining clusters corresponding to the first cluster R to the first remaining normal representation value corresponding to the first cluster R is denoted as the weight value of the corresponding first cluster, i.e., the weight of any first cluster in the set of remaining clusters. The weighted value is the ratio of the normal representation value of the first cluster to the first residual normal representation value corresponding to the first cluster R. Then, the weighted ratio set corresponding to the first cluster R is obtained. The j-th weighted ratio in the weighted ratio set is the ratio of the normal representation value of the j-th first cluster in the residual cluster set corresponding to the first cluster R to the normal representation value of the first cluster R, multiplied by the weight value of the j-th first cluster. Then, the cumulative result of all weighted ratios in the weighted ratio set is calculated and recorded as the target anomaly coefficient of the first cluster R. The specific calculation expression of the target anomaly coefficient of the first cluster R is:

[0029] in, Let J be the target anomaly coefficient of the first cluster R, and J be the number of clusters in the remaining cluster set corresponding to the first cluster R. This represents the normal characteristic value of the j-th first cluster in the remaining cluster set corresponding to the first cluster R; and the higher the anomaly level of a certain first cluster, the larger the difference between its normal characteristic value and the normal characteristic values ​​of other first clusters. The larger the value, the greater the anomaly of the first cluster R, and because... It is determined by comparing it with the clusters in the remaining cluster set, so when When it is larger, The greater the reference value, the more dominant it should be. As a result, In order to Normalization is performed, therefore The larger and When it is larger, The larger, and The larger the value, the higher the anomaly of the first cluster R. A higher anomaly of the first cluster R indicates that the data belonging to the first cluster R deviates more from the normal state, or that the first cluster R is more likely to be an anomalous cluster. Conversely, when the value is smaller... The smaller the value, the higher the normality of the first cluster R. The higher the normality of the first cluster R, the closer the data belonging to the first cluster R is to a normal state, or the more likely the first cluster R is to be a normal cluster.

[0030] Therefore, this embodiment can obtain the target anomaly coefficient of the first cluster and the target anomaly coefficient of the second cluster through the above process.

[0031] Step S003: Based on the target anomaly coefficient of the cluster to which the data correlation coefficient and differential correlation coefficient belong at the current monitoring time, obtain the target anomaly index value corresponding to the current monitoring time, and monitor the water system based on the target anomaly index value.

[0032] Since the target anomaly coefficient of a cluster can reflect the deviation of data belonging to the corresponding cluster from normal or the probability of anomaly, this embodiment, after obtaining the target anomaly coefficient of the cluster, obtains the target anomaly index value corresponding to the current monitoring time based on the target anomaly coefficient of the cluster to which the data correlation coefficient and differential correlation coefficient of the target monitoring node belong at the current monitoring time. Specifically, it obtains the target anomaly coefficient of the first cluster to which the data correlation coefficient of the target monitoring node belongs at the current monitoring time, plus the anomaly coefficient of the second cluster to which the differential correlation coefficient of the target monitoring node belongs at the current monitoring time, and records the result as the target anomaly index value corresponding to the target monitoring node at the current monitoring time. Furthermore, the larger the target anomaly index value corresponding to the target monitoring node at the current monitoring time, the more abnormal the change in the water pressure-flow relationship coefficient at the target monitoring node is, and the greater the probability of anomalies or abnormal trends in the water system.

[0033] Therefore, the embodiment can obtain the target abnormal index value corresponding to each monitoring position at the current monitoring moment through the above process. Then, the water system is monitored according to the target abnormal index value corresponding to the current monitoring moment. The specific process is as follows: whether the target abnormal index value corresponding to all monitoring positions of the water system at the current monitoring moment is located in the preset normal interval corresponding to the monitoring position at the current monitoring moment. If not, it is determined that the water system at the current monitoring moment has an operation abnormality or an abnormal trend, and an alarm needs to be given immediately, and maintenance and troubleshooting need to be performed in time. If yes, it is determined that the water system at the current monitoring moment is normal, and can continue to operate normally. The preset normal interval is obtained by a three-sigma detection method. The specific determination process of the preset normal interval corresponding to any monitoring position at the current monitoring moment is as follows: the target abnormal index value corresponding to the monitoring position at all historical monitoring moments is obtained, and is recorded as the historical target abnormal index value corresponding to the monitoring position. The method for obtaining the target abnormal index value in the embodiment is the same. The standard deviation and the mean value of all historical target abnormal index values corresponding to the monitoring position are calculated. The mean value is taken as a central reference point, and a range of plus or minus three times the standard deviation is taken as the preset normal interval corresponding to the monitoring position at the current monitoring moment. That is, the preset normal interval corresponding to the monitoring position at the current monitoring moment is , is the mean value of all historical target abnormal index values corresponding to the monitoring position, is the standard deviation of all historical target abnormal index values corresponding to the monitoring position.

[0034] Thus, the embodiment completes the abnormal monitoring and early warning of the water system.

[0035] In summary, the embodiment first obtains the target water pressure data sequence and the target water flow data sequence corresponding to the monitoring moment of the water system. Then, the data correlation coefficient and the difference correlation coefficient corresponding to different monitoring moments are obtained according to the correlation between the target water pressure data sequence and the target water flow data sequence and the correlation between the difference sequence of the target water pressure data sequence and the difference sequence of the target water flow data sequence. The target abnormal coefficient of the first clustering cluster and the target abnormal coefficient of the second clustering cluster are obtained according to the data amount proportion and the variation coefficient in the first clustering cluster and the second clustering cluster obtained by clustering the data correlation coefficient and the difference correlation coefficient corresponding to the monitoring moment. Then, the target abnormal index value corresponding to the current monitoring moment is obtained according to the target abnormal coefficient of the clustering cluster to which the data correlation coefficient and the difference correlation coefficient corresponding to the current monitoring moment belong. Finally, the water system is monitored according to the target abnormal index value. According to the data correlation coefficient and the difference correlation coefficient, the embodiment can improve the timeliness and reliability of the abnormal monitoring and early warning of the water system.

[0036] The above-described embodiments are only used to illustrate the technical solutions of the present application, but not limit them; although the present application is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A smart water monitoring and management system based on the Internet of Things, comprising a processor and a memory, characterized in that, The processor executes the computer program stored in the memory to perform the following steps: Acquire the target water pressure data sequence and target water flow data sequence of the water system at the monitoring time; Based on the correlation between the target water pressure data sequence and the target water flow data sequence, and the correlation between the differential sequence of the target water pressure data sequence and the differential sequence of the target water flow data sequence, the data correlation coefficient and differential correlation coefficient corresponding to different monitoring times are obtained. Based on the data volume ratio and variation coefficient of the first cluster and the second cluster obtained by clustering the data correlation coefficient and differential correlation coefficient corresponding to the monitoring times, the target anomaly coefficient of the first cluster and the target anomaly coefficient of the second cluster are obtained. Based on the target anomaly coefficient of the cluster to which the data correlation coefficient and differential correlation coefficient belong at the current monitoring time, the target anomaly index value corresponding to the current monitoring time is obtained, and the water system is monitored based on the target anomaly index value.

2. The smart water monitoring and management system based on the Internet of Things as described in claim 1, characterized in that, The method for obtaining the target water pressure data sequence and the target water flow rate data sequence includes: Acquire the initial water pressure data sequence and initial water flow data sequence of the water system at the monitoring time; For any given monitoring time, the water pressure change characteristic value at that monitoring time is obtained based on the difference in the first-order difference sequence of the initial water pressure data sequence corresponding to that monitoring time. Similarly, the water flow change characteristic value at that monitoring time is obtained based on the difference in the first-order difference sequence of the initial water flow data sequence corresponding to that monitoring time. The target data selection time length at that monitoring time is then obtained based on the water pressure change characteristic value and the water flow change characteristic value. Finally, the target water pressure data sequence and the target water flow data sequence corresponding to that monitoring time are obtained based on the target data selection time length. The method for obtaining the water flow change characteristic value is the same as the method for obtaining the water pressure change characteristic value.

3. The smart water monitoring and management system based on the Internet of Things as described in claim 2, characterized in that, The method for obtaining the water pressure change characteristic value at the monitoring time includes: The first-order difference sequence of the initial water pressure data sequence corresponding to the monitoring time is denoted as sequence A; For the i-th difference value in sequence A, where i is greater than 1 and less than N, and N is the total amount of data in sequence A, the absolute value of the difference between the (i-1)-th difference value and the i-th difference value in sequence A is recorded as the first difference value, the absolute value of the difference between the i-th difference value and the (i+1)-th difference value in sequence A is recorded as the second difference value, the sum of the first difference value and the second difference value is recorded as the comprehensive difference value, the absolute value of the difference between the (i-1)-th difference value and the (i+1)-th difference value is recorded as the adjacent difference value, the ratio of the comprehensive difference value to the adjacent difference value is recorded as the change index value of the i-th difference value, the result of negatively correlated mapping and normalization of the time interval between the water pressure data collection time and the monitoring time corresponding to the i-th difference value is recorded as the time distance weight factor of the i-th difference value, and the product of the change index value and the time distance weight factor is recorded as the weighted change index value of the i-th difference value. The result of negatively correlated mapping of the weighted change index values ​​of all difference values ​​in sequence A is recorded as the water pressure change characteristic value at the monitoring time.

4. The smart water monitoring and management system based on the Internet of Things as described in claim 3, characterized in that, The target data selection time length at the monitoring time is the product of the minimum value of the water pressure change characteristic value and the water flow change characteristic value and the preset initial time length.

5. The smart water monitoring and management system based on the Internet of Things as described in claim 2, characterized in that, A method for selecting a time length based on the target data at the monitoring time, and obtaining the target water pressure data sequence and target water flow data sequence corresponding to the monitoring time, includes: For any given monitoring time, the target data selection time length at that monitoring time is denoted as T. The time period of length T before the monitoring time is denoted as the target time period before the monitoring time. All water pressure data collected within the target time period before the monitoring time and the water pressure data collected at the monitoring time are combined to form a time series, which is denoted as the target water pressure data sequence corresponding to the monitoring time. All water flow data collected within the target time period before the monitoring time and the water flow data collected at the monitoring time are combined to form a time series, which is denoted as the target water flow data sequence corresponding to the monitoring time.

6. The smart water monitoring and management system based on the Internet of Things as described in claim 1, characterized in that, The method for obtaining the data correlation coefficient and the difference correlation coefficient includes: For any given monitoring time, the Spearman rank correlation coefficient between the target water pressure data sequence and the target water flow data sequence corresponding to the monitoring time is taken as the data correlation coefficient corresponding to the monitoring time, and the Spearman rank correlation coefficient between the first difference sequence of the target water pressure data sequence and the first difference sequence of the target water flow data sequence corresponding to the monitoring time is taken as the difference correlation coefficient corresponding to the monitoring time.

7. The smart water monitoring and management system based on the Internet of Things as described in claim 1, characterized in that, Methods for obtaining the target anomaly coefficient of clusters include: For any first cluster, the ratio of the total number of data correlation coefficients in the first cluster to the total number of data correlation coefficients participating in the cluster is denoted as the quantity ratio. The product of the result of normalizing the coefficient of variation of the first cluster and performing negative correlation mapping and the quantity ratio is used as the normal characterization value of the first cluster. Based on the normal characterization value of the first cluster and the normal characterization values ​​of other first clusters besides the first cluster, the target anomaly coefficient of the first cluster is obtained. The target anomaly coefficient of the second cluster is obtained in the same way as the target anomaly coefficient of the first cluster.

8. The smart water monitoring and management system based on the Internet of Things as described in claim 7, characterized in that, A method for obtaining the target anomaly coefficient of the first cluster based on the normal characterization values ​​of the first cluster and the normal characterization values ​​of other first clusters besides the first cluster includes: The set of all first clusters other than the first cluster is denoted as the residual cluster set. The sum of the normal representation values ​​of all first clusters in the residual cluster set is denoted as the first residual normal representation value. The ratio of the normal representation value of each first cluster in the residual cluster set to the first residual normal representation value is denoted as the weight value of the corresponding first cluster. A set of weighted ratios corresponding to the first cluster is obtained. The j-th weighted ratio in the set is the result of multiplying the ratio of the normal representation value of the j-th first cluster in the residual cluster set to the normal representation value of the first cluster by the weight value of the j-th first cluster. j is the position order of the data or cluster in the corresponding set, and the value of j ranges from 1 to J, where J is the number of clusters in the residual cluster set. The cumulative result of the weighted ratios in the set is denoted as the target anomaly coefficient of the first cluster.

9. The smart water monitoring and management system based on the Internet of Things as described in claim 1, characterized in that, The method for obtaining the target anomaly indicator value corresponding to the current monitoring time includes: The sum of the target anomaly coefficient of the first cluster to which the data correlation coefficient belongs at the current monitoring time and the target anomaly coefficient of the second cluster to which the differential correlation coefficient belongs at the current monitoring time is denoted as the target anomaly index value at the current monitoring time.

10. The smart water monitoring and management system based on the Internet of Things as described in claim 1, characterized in that, A method for monitoring a water system based on the target abnormal indicator value includes: If the target abnormal indicator value does not belong to the preset normal range, an abnormal alarm will be triggered.

Images