Remote water meter abnormality early warning method and system based on big data

Abstract

The present application belongs to the technical field of remote water meter operation management and big data processing, and particularly relates to a remote water meter abnormality early warning method and system based on big data, which comprises the following steps: obtaining the cumulative water consumption of all remote water meters, calculating the time period fluctuation volume of any remote water meter as a target remote water meter, calculating the spatial coordination deviation volume according to the physical space distribution law of the time period fluctuation volume, calculating the instantaneous abnormality ratio according to the historical time sequence statistical characteristics of the spatial coordination deviation volume, calculating the early warning confidence according to the dynamic evolution characteristics of the instantaneous abnormality ratio, and finally combining the preset trigger judgment threshold to complete the abnormality judgment of the target remote water meter. The present application improves the self-adaptive judgment ability of the model to the complex dynamic water supply operation background and reduces the interference of the pseudo abnormal signal caused by the pipe network water hammer effect.

Description

Technical Field

[0001] This invention relates to the field of remote water meter operation management and big data processing technology. More specifically, this invention relates to a method and system for early warning of anomalies in remote water meters based on big data. Background Technology

[0002] Remote water meter anomaly early warning refers to the technology that uses time-series analysis and feature extraction of cumulative water consumption data collected by remote water meters to identify events such as water meter equipment failure, pipeline leakage, and abnormal water use, and then sends early warning signals to operation and maintenance personnel.

[0003] To promptly detect damage to remote water meters, pipeline leaks, and sudden abnormal water usage events in water supply networks, reduce leakage rates, and ensure residents' water safety, water supply companies are now widely deploying big data-based remote water meter anomaly early warning systems. These systems perform feature analysis on massive amounts of historical and real-time water usage data from remote water meters. For example, Chinese patent document CN117196446B discloses a big data-based real-time product risk monitoring platform. This platform discloses a technical solution that collects operational status data from various devices, extracts independent operational characteristics of individual devices, sets static judgment thresholds for anomaly identification, and generates equipment maintenance work orders based on the identification results.

[0004] However, in the practical application of remote water meter anomaly early warning in water supply networks, this type of technology has the problem of treating each remote water meter as an independent analysis node and lacking a rigorous physical measurement logic for the spatial correlation information between water meters. This makes it impossible to accurately distinguish between normal fluctuations in the network environment caused by changes in the network and independent anomalies of a single remote water meter while maintaining strict consistency of physical dimensions. At the same time, the existing technology relies too much on complex nonlinear models for feature fitting, resulting in high computational overhead. The feature extraction process is prone to deviating from the basic physical meaning of fluid mechanics, and the generalization ability is insufficient under complex network conditions, resulting in a high false alarm rate for anomaly early warning. Summary of the Invention

[0005] To address the aforementioned technical problem of high false alarm rates in remote water meter anomaly warning systems, this invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a remote water meter anomaly early warning method based on big data, comprising: Obtain the cumulative water consumption of all remote water meters; select any remote water meter as the target remote water meter and calculate the time-period fluctuation volume of the target remote water meter; calculate the spatial coordination deviation volume of each remote water meter based on the physical spatial distribution pattern of the time-period fluctuation volume; calculate the instantaneous anomaly ratio of each remote water meter based on the historical time-series statistical characteristics of the spatial coordination deviation volume; calculate the warning confidence level of each remote water meter based on the dynamic evolution characteristics of the instantaneous anomaly ratio; and determine the anomaly of the target remote water meter based on the relationship between the warning confidence level and the preset trigger judgment threshold.

[0007] This invention obtains the cumulative water consumption of all remote water meters, selects any remote water meter as the target remote water meter and calculates its time-period fluctuation volume, then calculates the spatial coordination deviation volume of each remote water meter based on the physical spatial distribution law of the time-period fluctuation volume, then calculates the instantaneous anomaly ratio of each remote water meter based on the historical time-series statistical characteristics of the spatial coordination deviation volume, then calculates the warning confidence level of each remote water meter based on the dynamic evolution characteristics of the instantaneous anomaly ratio, and finally determines the anomaly of the target remote water meter based on the relationship between the warning confidence level and the preset trigger judgment threshold. In this way, by combining spatial correlation information and temporal evolution characteristics, it distinguishes between independent anomalies of a single meter and overall fluctuations of the pipeline network, reduces false alarms caused by instantaneous interference such as water hammer effect, reduces the computational cost of the model, improves the adaptability of the remote water meter anomaly warning method to different water supply network conditions, and provides a reference basis for anomaly judgment for the pipeline operation and maintenance of water supply companies.

[0008] Preferably, obtaining the cumulative water consumption of all remote water meters includes: The sampling period of the high-frequency communication module is set, and the original cumulative water consumption data of the remote water meter is read through the data interface of the water supply IoT platform according to the sampling period. If the data is missing at a certain sampling time, the missing value is filled by linear interpolation to obtain the preprocessed cumulative water consumption.

[0009] Compared to existing IoT systems that directly discard abnormal data or use complex nonlinear prediction algorithms that cause a surge in backend computing power when encountering communication packet loss, this invention, while ensuring the continuity of high-frequency dynamic monitoring, quickly restores the continuous and realistic water usage conditions of water meters through a linear interpolation method with low computing power overhead. This effectively compensates for data breakpoints caused by environmental interference during wireless transmission and avoids introducing excessive human fitting bias, thus ensuring the integrity and reliability of the underlying collected data.

[0010] Preferably, the calculation of the time-period fluctuation volume of the target remote water meter includes: Select any remote water meter as the target remote water meter, calculate the difference between the cumulative water consumption of the target remote water meter at the current sampling time and the cumulative water consumption at the previous sampling time, and obtain the time-period fluctuation volume of the target remote water meter at the current sampling time.

[0011] Preferably, the spatial cooperative deviation volume satisfies the following relationship: ; In the formula, Indicates in Sampling time target remote water meter Spatial cooperative deviation volume; Indicates in Sampling time target remote water meter The volume of fluctuation over a period of time; Indicates the target remote water meter Rated flow rate; This represents the sum of the rated flow rates of all neighboring remote water meters; This indicates the total number of remote water meters in the neighborhood; Indicates in Sampling time target remote water meter The The time-period fluctuation volume of a neighboring remote water meter.

[0012] This invention utilizes the ratio of the rated flow of the target remote water meter to the sum of the rated flow of all neighboring remote water meters to proportionally allocate the sum of the time-period fluctuation volumes of all neighboring remote water meters. The absolute value of the difference between the allocation result and the time-period fluctuation volume of the target remote water meter itself is used to calculate the spatial coordination deviation volume. This allows the expected water volume fluctuations to be reasonably distributed according to the physical specifications of each remote water meter. By extracting the difference between the time-period fluctuation volume and the theoretically expected fluctuation mapping amount, the overall water volume fluctuations caused by environmental water pressure fluctuations in the local water supply network are removed. The spatial coordination deviation volume purely reflects the abnormal water consumption increment of the target remote water meter, independent of the surrounding water supply network, thereby reducing the probability of misjudging normal linkage water pressure fluctuations as remote water meter leakage during peak water consumption periods.

[0013] Preferably, the instantaneous anomaly ratio satisfies the following relationship: ; In the formula, Indicates in Sampling time target remote water meter The instantaneous anomaly ratio; Indicates in Sampling time target remote water meter Spatial cooperative deviation volume; Indicates the target remote water meter Based on sampling time Within the historical time window ending at the end, the first Spatial cooperative deviation volume at each sampling point; This indicates the total number of sampling points within the historical time window; This represents the first preset minute value.

[0014] This invention calculates the instantaneous anomaly ratio, allowing the stringency of anomaly judgment to be dynamically adjusted based on the historical operating status of the target remote water meter. For remote water meters that have been operating at a stable baseline for a long time, even a small abnormal deviation can trigger a high instantaneous anomaly ratio output. For older remote water meters operating in high-noise environments, a larger historical baseline average is used to suppress the false anomaly ratio generated at the current moment. At the same time, a first preset small value is used to prevent the loophole of the denominator being zero, thereby reducing misjudgment of the inherent aging physical state of the target remote water meter and improving the feature adaptation capability of the early warning model under the service conditions of multi-source heterogeneous remote water meter equipment.

[0015] Preferably, the early warning confidence level satisfies the following relationship: ; In the formula, Indicates in Sampling time target remote water meter The confidence level of the early warning; Indicates in Sampling time target remote water meter The instantaneous anomaly ratio; Indicates in Sampling time target remote water meter The instantaneous anomaly ratio; This represents the second preset micro value.

[0016] In calculating the early warning confidence level of a target remote water meter, this invention combines the relative difference between the instantaneous anomaly ratio at the current sampling time and the instantaneous anomaly ratio at the previous sampling time. This difference is used to generate a trend adjustment coefficient to weight and scale the current instantaneous anomaly ratio. This not only objectively assesses the severity of the current anomaly deviation but also dynamically captures the evolution direction of the target remote water meter's leakage status. When the instantaneous anomaly ratio is in a rapidly rising phase, the early warning confidence level is proactively amplified. When a single transient oscillation occurs in the water supply network and the instantaneous anomaly ratio begins to decline, the early warning confidence level is reasonably suppressed. This reduces the overreaction of the monitoring system to non-continuous physical phenomena such as one-time water pressure surges in the water supply network.

[0017] Preferably, the step of determining the anomaly of the target remote water meter based on the relationship between the warning confidence level and the preset trigger judgment threshold includes: Set a trigger threshold to determine the target remote water meter. The warning confidence level at the sampling time is compared with the trigger judgment threshold. If the target remote water meter is in... If the warning confidence level at the sampling time is greater than or equal to the trigger judgment threshold, the system outputs an alarm pop-up to the management terminal and generates a device maintenance dispatch order containing the geographical coordinates of the target remote water meter; if the target remote water meter is in If the confidence level of the early warning at the sampling time is less than the trigger judgment threshold, the system will not intervene and will remain in a silent monitoring state.

[0018] Preferably, the determination of the neighborhood remote water meter includes: Centered on the target remote water meter, a preset radius distance is defined in the water supply network, and the remaining remote water meters within the preset radius distance are regarded as neighboring remote water meters of the target remote water meter.

[0019] Preferably, obtaining the sum of the rated flow rates of the neighboring remote water meters includes: The rated flow rate of the remote water meter is read from the nameplate ledger of the remote water meter device, and the sum of the rated flow rates of all neighboring remote water meters of the target remote water meter is calculated.

[0020] Secondly, the present invention provides a remote water meter anomaly early warning system based on big data, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned remote water meter anomaly early warning method based on big data is implemented.

[0021] By adopting the above technical solution, the above-mentioned remote water meter anomaly early warning method based on big data is generated into a computer program and stored in a memory so that it can be loaded and executed by a processor. In this way, a terminal device can be made based on the memory and processor for convenient use.

[0022] The beneficial effects of this invention are as follows: Addressing the monitoring challenges caused by uneven water pressure distribution and aging differences in remote water meters within water supply networks, this invention integrates water usage data from the target remote water meter and neighboring remote water meters to calculate the spatial coordination deviation volume. This eliminates interference from common water pressure fluctuations caused by overall water supply network scheduling or peak water usage. Furthermore, this invention introduces a historical time window, combining the temporal statistical characteristics of the target remote water meter to assess the instantaneous anomaly ratio and early warning confidence. It integrates current water volume deviation characteristics with long-term equipment background noise to establish a dynamically evolving evaluation scale. This not only tracks the evolution trend of remote water meter leakage but also suppresses false jump signals caused by short-term oscillations in the water supply network, thereby reducing the frequency of misjudgments in complex operating conditions. This improves the robustness of big data-based remote water meter leakage monitoring and enhances the refined operation level of water supply companies. Attached Figure Description

[0023] Figure 1 The flowchart of the remote water meter anomaly early warning method based on big data in this invention is illustrated in the schematic diagram. Figure 2 This is a schematic diagram of the physical spatial distribution of remote water meters in the pipeline network; Figure 3 This is a schematic diagram illustrating the evolution of the fluctuation volume of the target remote water meter. Detailed Implementation

[0024] This invention discloses a remote water meter anomaly early warning method based on big data, referring to... Figure 1 This includes steps S100-S500: S100: Obtain the cumulative water consumption of all remote water meters and calculate the fluctuation volume.

[0025] It should be noted that since the raw data transmitted back by the remote water meter is the cumulative water consumption, the wireless IoT transmission process is susceptible to environmental interference, resulting in data interruptions. Directly using the raw cumulative value cannot obtain the water consumption changes for a single time period. Using complex nonlinear repair methods introduces additional computational load and human bias. Linear interpolation minimizes computational load while restoring continuous operating conditions, and differential conversion is the most direct way to convert cumulative values ​​into time period changes. Therefore, this invention first performs linear interpolation repair on the raw data, and then calculates the time period fluctuation volume by using the difference between the cumulative values ​​of adjacent time periods, thereby improving the data integrity of the basic input variables and enhancing the extraction accuracy of water consumption change characteristics for different time periods.

[0026] Specifically, a sampling period is set for the high-frequency communication module. The raw cumulative water consumption data from the remote water meter is read according to the sampling period via the data interface of the water supply IoT platform. If data is missing at a certain sampling moment, the missing value is filled using linear interpolation to obtain the preprocessed cumulative water consumption. It should be noted that the linear interpolation method is existing technology and will not be elaborated upon here. The sampling period is set to 15 minutes to meet the dynamic monitoring requirements of the water supply network.

[0027] Select any remote water meter as the target remote water meter, calculate the difference between the cumulative water consumption of the target remote water meter at the current sampling time and the cumulative water consumption at the previous sampling time, and obtain the time-period fluctuation volume of the target remote water meter at the current sampling time.

[0028] At this point, the time-period fluctuation volume of all remote water meters has been obtained.

[0029] S200, Calculate the spatial coordination deviation volume of each remote water meter.

[0030] It should be noted that in interconnected local water supply networks, pressure regulation at pumping stations or high-flow-rate water use in adjacent areas can trigger water pressure transmission, causing synchronous fluctuations in multiple water meters within the area. Simply looking at the absolute fluctuation of a single meter cannot distinguish whether the anomaly is due to its own inherent abnormality or the transmission of surrounding water pressure. Using complex spatial correlation models deviates from the fundamentals of fluid mechanics. The rated flow rate is the water flow capacity parameter calibrated at the water meter's factory, and the total fluctuation distributed according to the flow capacity conforms to the fluid distribution law of the pipe network. Therefore, this invention calculates the expected fluctuation of the target water meter by proportion to the rated flow rate, and then subtracts this from the actual fluctuation to obtain the spatially coordinated deviation volume. This reduces the interference of group common-mode fluctuations on the independent anomaly judgment of a single meter.

[0031] Specifically, based on the physical spatial distribution pattern of the time-period fluctuation volume, the spatial coordination deviation volume of all remote water meters is calculated, including: Centered on the target remote water meter, a preset radius is defined within the water supply network, and the remaining remote water meters within this radius are considered neighboring remote water meters. For example, the preset radius is set to 500m to match the typical physical transmission range of water pressure fluctuations.

[0032] Figure 2 This is a schematic diagram of the physical spatial distribution of remote water meters in the pipeline network. The diagram shows the process of defining a preset radius distance with the target remote water meter as the center, and establishing the other remote water meters within this radius distance as neighboring remote water meters.

[0033] The rated flow rate of the remote water meter is read from the nameplate ledger of the remote water meter device, which is used as the basic water supply capacity constant of the pipeline network, and the sum of the rated flow rates of all neighboring remote water meters of the target remote water meter is calculated.

[0034] The spatial coordination deviation volume of the target remote water meter satisfies the following relationship: ; In the formula, Indicates in Sampling time target remote water meter Spatial cooperative deviation volume; Indicates in Sampling time target remote water meter The volume of fluctuation over a period of time; Indicates the target remote water meter Rated flow rate; This represents the sum of the rated flow rates of all neighboring remote water meters; This indicates the total number of remote water meters in the neighborhood; Indicates in Sampling time target remote water meter The The time-period fluctuation volume of a neighboring remote water meter.

[0035] In this relation, This represents the actual total fluctuation volume of a local water supply network at all neighboring remote water meters due to the influence of common environmental water pressure. The larger the actual total fluctuation volume, the more drastic the overall water pressure fluctuation or overall water consumption change within the local network area; conversely, the smaller the volume, the more stable the overall water flow within the local network area. This indicates that water is allocated to the target remote water meter according to the proportion of basic water supply capacity. The larger the expected fluctuation volume mapping value, the more significant the target remote water meter. The greater the fluctuation difference that the target remote water meter should bear due to changes in the overall physical water pressure environment of the pipeline network, the better; conversely, the smaller the fluctuation difference, the more likely it is to be affected by changes in the target remote water meter. Theoretically, the less it is affected by fluctuations in water pressure in the surrounding pipe network. This indicates the target remote water meter after removing the background of local water pressure changes in the pipe network. The independent deviation volume of the actual fluctuation volume over a given period relative to the theoretically expected fluctuation volume is the volume of water meter that is being measured. The larger this independent deviation volume, the greater the deviation. The greater the actual independent leakage or sudden increase in water volume that exceeds theoretical expectations, the better; conversely, the less leakage or sudden increase in water volume, the more likely it is to be a problem with the target remote water meter. The greater the independent water reduction, the lower it is than theoretically expected.

[0036] Figure 3 This is a schematic diagram of the fluctuation volume evolution of the target remote water meter. The diagram shows the evolution trend of the time-period fluctuation volume and spatial coordination deviation volume of the target remote water meter at different sampling times, and a significant deviation occurs after the point of physical damage.

[0037] At this point, the spatial coordinated deviation volume of all remote water meters was obtained.

[0038] S300, Calculate the instantaneous anomaly rate of each remote water meter.

[0039] It should be noted that due to inherent differences in the basic operating states of different water meters, the level of routine micro-fluctuations of the same water meter also varies at different times. Using a uniform fixed threshold will generate a large number of false alarms for water meters with large background fluctuations and miss alarms for water meters with small background fluctuations. A dynamic threshold requires a benchmark that can represent the routine fluctuation level of water meters. The arithmetic mean within the historical time window is the simplest to calculate and can reflect long-term routine fluctuations. Therefore, this invention compares the current spatial cooperative deviation volume with the overall reference scale that includes the historical background to obtain a dimensionless instantaneous anomaly ratio, thereby improving the model's adaptive ability to the differences in basic fluctuations of different water meters and different time periods.

[0040] Specifically, based on the historical time-series statistical characteristics of spatial coordination deviation volume, the instantaneous anomaly ratio of all remote water meters is calculated, including: Set based on sampling time The historical time window is the endpoint. For example, the historical time window is set to 24 hours, which includes the past 96 sampling points.

[0041] The instantaneous anomaly ratio of the target remote water meter satisfies the following relationship: ; In the formula, Indicates in Sampling time target remote water meter The instantaneous anomaly ratio; Indicates in Sampling time target remote water meter Spatial cooperative deviation volume; Indicates the target remote water meter Based on sampling time Within the historical time window ending at the end, the first Spatial cooperative deviation volume at each sampling point; This indicates the total number of sampling points within the historical time window; This represents the first preset microvalue, used to prevent the denominator from being 0, and is set to 0.001.

[0042] In this relation, Indicates the target remote water meter Within a historical time window, the basic fluctuation benchmark is naturally generated due to its own physical aging or minor disturbances in the local pipeline network. The larger the basic fluctuation benchmark, the more unstable the underlying physical state of the target remote water meter is during its historical operation, and the more persistently high the system background noise exists. Conversely, it indicates that the internal measurement status of the target remote water meter is extremely stable during the historical time period and follows the cooperative evolution law of the surrounding pipeline network. This represents the overall reference scale used to assess the actual comprehensive deterioration of the target remote water meter, which combines the current actual abnormal deviation characteristics with long-term historical background noise data. The larger the overall reference scale, the more serious the independent deviation is, or the higher the inherent historical noise of the target remote water meter. This is used as a dynamic evaluation benchmark to suppress false anomaly signals. Conversely, it indicates that the target remote water meter has not experienced any obvious physical abnormalities at present, and its historical pipeline operating environment is relatively clean.

[0043] At this point, the instantaneous anomaly rates of all remote water meters were obtained.

[0044] S400, calculate the early warning confidence level of each remote water meter.

[0045] It should be noted that water hammer in pipelines can cause instantaneous flow rate jumps, manifesting as a rapid decline after a single-point anomaly. Looking only at the instantaneous anomaly ratio can misinterpret these instantaneous jump signals as genuine anomalies. Real leakage does not appear and disappear suddenly; it exhibits a continuous upward trend. The relative growth rate between adjacent time points directly reflects whether the anomaly is continuously worsening or rapidly declining. Therefore, this invention uses a dynamic trend multiplier to weight the instantaneous anomaly ratio, obtaining a warning confidence level that comprehensively reflects both the degree of anomaly and the trend of deterioration. This suppresses false anomaly false alarms caused by instantaneous jump signals such as water hammer.

[0046] Specifically, based on the dynamic evolution characteristics of the instantaneous anomaly ratio, the early warning confidence level of all remote water meters is calculated, and the early warning confidence level of the target remote water meter satisfies the following relationship: ; In the formula, Indicates in Sampling time target remote water meter The confidence level of the early warning; Indicates in Sampling time target remote water meter The instantaneous anomaly ratio; Indicates in Sampling time target remote water meter The instantaneous anomaly ratio; This represents the second preset microvalue, used to prevent the denominator from being 0, and is set to 0.001.

[0047] In this relation, This indicates the abnormal relative increase in the value of the target remote water meter between adjacent sampling times, caused by physical structural damage or deterioration of the pipeline network. The larger the abnormal relative increase, the more rapidly the actual leakage or abnormal operating condition of the target remote water meter is aggravated in a short period of time, and it has a physical inertia of continuous deterioration. Conversely, it indicates that the abnormal physical deviation of the target remote water meter is stabilizing, or it is a one-time jump caused by short-term pipeline pressure oscillation, and it is falling back and dissipating.

[0048] At this point, the warning confidence levels for all remote water meters were obtained.

[0049] S500: Determine the anomalies of remote water meters based on the warning confidence level.

[0050] It should be noted that since the early warning system needs to filter out the abnormal events that need to be handled from massive amounts of real-time monitoring data, not setting a threshold will cause all fluctuations to trigger alarms, which will be impossible for maintenance personnel to handle. Setting thresholds based on experience will be affected by subjective factors. Thresholds obtained based on a large number of historical abnormal samples can reduce unnecessary alarms without missing key anomalies. Therefore, this invention achieves automatic judgment and graded response of abnormal events by comparing the early warning confidence level with the preset trigger judgment threshold, thereby reducing the number of invalid alarms and improving the practical value of the early warning system.

[0051] Specifically, a trigger threshold is set to determine the target remote water meter. The warning confidence level at the sampling time is compared with the trigger judgment threshold. If the target remote water meter is in... If the warning confidence level at the sampling time is greater than or equal to the trigger judgment threshold, the system outputs an alarm pop-up to the management terminal and generates a device maintenance dispatch order containing the geographical coordinates of the target remote water meter; if the target remote water meter is in If the confidence level of the early warning at the sampling time is less than the trigger threshold, the system will not intervene and will remain in a silent monitoring state. For example, the trigger threshold is set to 0.08, which is determined based on statistical analysis of a large number of historical abnormal samples from urban water supply networks, and serves as the judgment boundary for outputting alarm commands.

[0052] This completes the anomaly detection of the target remote water meter.

[0053] This invention also discloses a remote water meter anomaly early warning system based on big data, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the remote water meter anomaly early warning method based on big data according to this invention is implemented.

[0054] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0055] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.

Claims

1. A remote water meter anomaly early warning method based on big data, characterized in that, include: S100: Obtain the cumulative water consumption of all remote water meters; Use any remote water meter as the target remote water meter and calculate the time-period fluctuation volume of the target remote water meter. S200. Calculate the spatial coordination deviation volume of each remote water meter based on the physical spatial distribution law of the fluctuation volume during the time period. S300. Calculate the instantaneous anomaly ratio of each remote water meter based on the historical time-series statistical characteristics of the spatial cooperative deviation volume. S400. Calculate the early warning confidence level of each remote water meter based on the dynamic evolution characteristics of the instantaneous anomaly ratio. S500. Based on the relationship between the warning confidence level and the preset trigger judgment threshold, anomaly judgment is made on the target remote water meter.

2. The remote water meter anomaly early warning method based on big data according to claim 1, characterized in that, The process of obtaining the cumulative water consumption of all remote water meters includes: The sampling period of the high-frequency communication module is set, and the original cumulative water consumption data of the remote water meter is read through the data interface of the water supply IoT platform according to the sampling period. If the data is missing at a certain sampling time, the missing value is filled by linear interpolation to obtain the preprocessed cumulative water consumption.

3. The remote water meter anomaly early warning method based on big data according to claim 1, characterized in that, The calculation of the time-period fluctuation volume of the target remote water meter includes: Select any remote water meter as the target remote water meter, calculate the difference between the cumulative water consumption of the target remote water meter at the current sampling time and the cumulative water consumption at the previous sampling time, and obtain the time-period fluctuation volume of the target remote water meter at the current sampling time.

4. The remote water meter anomaly early warning method based on big data according to claim 1, characterized in that, The spatial cooperative deviation volume satisfies the following relationship: ； In the formula, Indicates in Sampling time target remote water meter Spatial cooperative deviation volume; Indicates in Sampling time target remote water meter The volume of fluctuation over a period of time; Indicates the target remote water meter Rated flow rate; This represents the sum of the rated flow rates of all neighboring remote water meters; This indicates the total number of remote water meters in the neighborhood; Indicates in Sampling time target remote water meter The The time-period fluctuation volume of a neighboring remote water meter.

5. The remote water meter anomaly early warning method based on big data according to claim 1, characterized in that, The instantaneous anomaly ratio satisfies the following relationship: ； In the formula, Indicates in Sampling time target remote water meter The instantaneous anomaly ratio; Indicates in Sampling time target remote water meter Spatial cooperative deviation volume; Indicates the target remote water meter Based on sampling time Within the historical time window ending at the end, the first Spatial cooperative deviation volume at each sampling point; This indicates the total number of sampling points within the historical time window; This represents the first preset minute value.

6. The remote water meter anomaly early warning method based on big data according to claim 1, characterized in that, The warning confidence level satisfies the following relationship: ； In the formula, Indicates in Sampling time target remote water meter The confidence level of the early warning; Indicates in Sampling time target remote water meter The instantaneous anomaly ratio; Indicates in Sampling time target remote water meter The instantaneous anomaly ratio; This represents the second preset micro value.

7. The remote water meter anomaly early warning method based on big data according to claim 1, characterized in that, The step of determining the anomaly of the target remote water meter based on the relationship between the warning confidence level and the preset trigger judgment threshold includes: Set a trigger threshold to determine the target remote water meter. The warning confidence level at the sampling time is compared with the trigger judgment threshold. If the target remote water meter is in... If the warning confidence level at the sampling time is greater than or equal to the trigger judgment threshold, the system outputs an alarm pop-up to the management terminal and generates a device maintenance dispatch order containing the geographical coordinates of the target remote water meter; if the target remote water meter is in If the confidence level of the early warning at the sampling time is less than the trigger judgment threshold, the system will not intervene and will remain in a silent monitoring state.

8. The remote water meter anomaly early warning method based on big data according to claim 4, characterized in that, The determination of the neighboring remote water meter includes: Centered on the target remote water meter, a preset radius distance is defined in the water supply network, and the remaining remote water meters within the preset radius distance are regarded as neighboring remote water meters of the target remote water meter.

9. The remote water meter anomaly early warning method based on big data according to claim 4, characterized in that, The acquisition of the sum of the rated flow rates of the neighboring remote water meters includes: The rated flow rate of the remote water meter is read from the nameplate ledger of the remote water meter device, and the sum of the rated flow rates of all neighboring remote water meters of the target remote water meter is calculated.

10. A remote water meter anomaly early warning system based on big data, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement the remote water meter anomaly early warning method based on big data according to any one of claims 1-9.

Images