Safety monitoring and early warning method and device based on water consumption characteristics of smart water meters

Abstract

The present invention relates to the field of safety monitoring and early warning, and in particular to a safety monitoring and early warning method and device based on the water consumption characteristics of smart water meters. The method comprises the following steps: obtaining historical water consumption monitoring data of elderly people living alone based on smart water meters; performing extreme value filtering and optimization processing on the historical water consumption monitoring data of elderly people living alone, thereby generating optimized water consumption monitoring data; performing periodic water consumption characteristic analysis on the optimized water consumption monitoring data to generate water consumption characteristic data; performing quantitative mining of the time series water consumption distribution characteristics of the water consumption characteristic data to construct a time series water consumption characteristic distribution map; obtaining dynamic water consumption behavior characteristics and the duration of each water consumption behavior based on the time series water consumption characteristic distribution map; performing personalized multi-time point evolution on the dynamic water consumption behavior characteristics based on the duration of each water consumption behavior, and constructing a personalized life water consumption behavior portrait. The present invention realizes real-time and efficient safety monitoring and early warning for elderly people living alone.

Description

Technical Field

[0001] The present invention relates to the field of safety monitoring and early warning, and in particular to a safety monitoring and early warning method and device based on water consumption characteristics of an intelligent water meter. Background Art

[0002] With the rapid development of Internet of Things technology, it has become possible to use smart devices to monitor the safety of elderly people living alone. In daily life, the water usage behavior of the elderly is often closely related to their living conditions and health conditions. For example, large amounts of water usage at night may mean that the elderly are active at night, while a sudden increase in water usage at noon may indicate that the elderly are taking a nap. By analyzing and monitoring the water usage characteristics of elderly people living alone, their living conditions can be understood in real time, and abnormal situations can be discovered in time, thereby improving the safety level of elderly people living alone. However, traditional water usage monitoring methods rely on manual recording and analysis, and cannot monitor the safety of elderly people living alone in real time and efficiently. Therefore, an intelligent water usage characteristic safety monitoring and early warning method for elderly people living alone is needed. Summary of the invention

[0003] In order to solve the above technical problems, the present invention proposes a safety monitoring and early warning method and device based on the water consumption characteristics of a smart water meter to solve at least one of the above technical problems.

[0004] To achieve the above object, the present invention provides a safety monitoring and early warning method based on water consumption characteristics of a smart water meter, comprising the following steps:

[0005] Step S1: obtaining historical water consumption monitoring data of elderly people living alone based on the smart water meter; performing extreme value filtering and optimization processing on the historical water consumption monitoring data of elderly people living alone, thereby generating optimized water consumption monitoring data;

[0006] Step S2: Perform periodic water consumption characteristic analysis on the optimized water consumption monitoring data to generate water consumption characteristic data; perform quantitative mining on the time series water consumption distribution characteristics of the water consumption characteristic data to construct a time series water consumption characteristic distribution map;

[0007] Step S3: based on the time series water use characteristic distribution map, dynamic water use behavior characteristics and the duration of each water use behavior are obtained; based on the duration of each water use behavior, the dynamic water use behavior characteristics are personalized and evolved at multiple time points to construct a personalized domestic water use behavior portrait;

[0008] Step S4: Acquire real-time smart water meter monitoring parameters; monitor the full-cycle water consumption fluctuation of the real-time smart water meter monitoring parameters through personalized domestic water consumption behavior portraits, and extract abnormal water consumption behavior data;

[0009] Step S5: generating an abnormal result quantification value based on the abnormal water use behavior data; performing a safety status assessment on the abnormal result quantification value based on a preset safety assessment model to generate safety status risk prediction data;

[0010] Step S6: Make safety warning decisions based on the safety status risk prediction data and construct a safety status warning strategy for elderly people living alone; perform safety monitoring and warning operations for elderly people living alone based on the safety status warning strategy for elderly people living alone.

[0011] The present invention obtains historical water consumption monitoring data of elderly people living alone to understand the historical patterns and habits of their water consumption behavior, performs extreme value filtering optimization processing to remove outliers and noise, improves the accuracy and credibility of the data, and generates optimized water consumption monitoring data that better reflects the actual water consumption of elderly people living alone, providing a reliable data basis for subsequent analysis. By analyzing the characteristics of periodic water consumption, the present invention obtains water consumption statistics of elderly people living alone, such as average water consumption, peak water consumption, water consumption cycle, etc., performs time series analysis on water consumption characteristic data, explores the time characteristics of water consumption behavior of elderly people living alone, such as peak water consumption hours, periodic changes in water consumption, etc., and constructs a time series water consumption monitoring system. The time-series water use characteristic distribution map intuitively displays the water use behavior patterns of elderly people living alone, helping to discover anomalies or abnormal trends. Through the time-series water use characteristic distribution map, the dynamic water use behavior characteristics of elderly people living alone, such as the frequency, duration, interval time, etc. of water use behavior, are extracted. Based on the personalized multi-time point evolution of the duration of each water use behavior, the water use behavior patterns of elderly people living alone at different time points are understood, the changes and habits of daily life are captured, and a personalized life water use behavior portrait is constructed to comprehensively describe the water use behavior characteristics of elderly people living alone, providing a personalized basis for subsequent safety monitoring, and obtaining real-time smart water meter monitoring parameters to understand the current water use of elderly people living alone in real time. The system can analyze the situation of water consumption, water use frequency, etc., monitor the fluctuation of water use in the whole cycle of real-time monitoring parameters through personalized life water use behavior portraits, compare them with the individual's historical behavior, find abnormalities or abnormal trends, improve the accuracy of abnormal detection, extract abnormal water use behavior data to help determine potential safety risks, such as water leakage, abnormally high frequency of water use, etc., generate abnormal result quantification values ​​based on abnormal water use behavior data, convert abnormal behavior into specific numerical indicators, and facilitate further analysis and comparison. Based on the preset safety assessment model, the abnormal result quantification values ​​are evaluated for safety status, and the abnormal behavior is compared with the safety standard to evaluate the safety status risk of elderly people living alone. The safety status risk prediction data is generated to discover potential safety problems in advance, provide a basis for taking corresponding prevention and intervention measures, and make safety warning decisions based on the safety status risk prediction data. According to the safety status risk degree of elderly people living alone, formulate corresponding warning strategies, handle safety problems in a graded manner, and build a safety status warning strategy for elderly people living alone. According to the specific situation, formulate targeted monitoring and intervention measures to improve the accuracy and effectiveness of safety warnings. By executing the safety monitoring and early warning operations for elderly people living alone, the water use behavior of elderly people living alone is monitored in real time, abnormal situations are discovered in time, and corresponding early warning measures are taken to ensure their safety.

[0012] Preferably, step S1 comprises the following steps:

[0013] Step S11: Acquire historical water consumption monitoring data based on the smart water meter;

[0014] Step S12: Performing a distribution analysis of the value range of the historical water monitoring data to obtain a normalized value range;

[0015] Step S13: Identify abnormal samples of historical water consumption monitoring data based on the normalized value range and extract abnormal monitoring samples;

[0016] Step S14: Perform extreme value filtering optimization processing based on the monitored abnormal samples to generate optimized water use monitoring data.

[0017] The present invention obtains the water use situation of the elderly living alone in the past period of time by acquiring historical water use monitoring data, including water use amount, water use frequency, etc. The historical water use monitoring data provides a data basis for subsequent analysis, helps to understand the water use behavior pattern and habit of the elderly living alone, performs monitoring data value range distribution analysis to understand the distribution of the historical water monitoring data, including statistical information such as minimum value, maximum value, average value, etc., obtains a normalized value range to determine the typical value range of the normal water use behavior of the elderly living alone, and serves as a reference for subsequent abnormal sample identification and optimization processing. Based on the normalized value range, abnormal sample identification is performed to compare the historical water use monitoring data with the normal range, and abnormal samples exceeding the normal range are found. By extracting the abnormal monitoring samples, the abnormal situation existing in the historical water use monitoring data, such as abnormally high water volume or abnormally low water volume, etc., is determined. The abnormal samples in the historical water use monitoring data are removed by extreme value filtering optimization processing, and the accuracy and credibility of the data are improved. The generated optimized water use monitoring data better reflects the actual water use situation of the elderly living alone, eliminates the influence of abnormal samples, and provides a more reliable data basis for subsequent analysis and safety monitoring.

[0018] Preferably, the specific steps of step S14 are:

[0019] Performing an abnormal type analysis on the monitored abnormal samples to obtain abnormal sample types; the abnormal sample types include true abnormal samples and error samples;

[0020] The timestamp of the error sample is extracted;

[0021] Performing error sample timing analysis based on the timestamp to generate error sample timing data;

[0022] Perform multivariate analysis on error sample time series data to identify error patterns;

[0023] Calculate the average value of historical water use monitoring data to obtain the mean water use value;

[0024] Perform mean filling correction on the error samples based on the water mean to generate error correction data;

[0025] Perform anomaly elimination processing based on real abnormal samples to obtain abnormal sample elimination data;

[0026] According to the error law, local filtering optimization is performed on the error correction data and abnormal sample elimination data to generate optimized water use monitoring data.

[0027] The present invention classifies the monitored abnormal samples through abnormal type analysis, and distinguishes between real abnormal samples and error samples. Real abnormal samples refer to actual water use abnormalities, while error samples are abnormal samples caused by measurement errors or equipment failures. The timestamps of the error samples are extracted to record the time point when the error samples occurred, which is used for subsequent time series analysis and correction processing. By performing time series analysis on the timestamps of the error samples, the error samples are organized in chronological order to form time series data. The time series data helps to understand the distribution law and change trend of the error samples, and provides a basis for subsequent analysis and correction. The correlation and laws in the error sample time series data are explored through multivariate analysis. Identifying error laws helps to understand the generation mechanism and influencing factors of error samples, and provides guidance for subsequent correction and optimization. The average value of historical water use monitoring data is calculated to obtain the average level of water consumption, the water use mean. As a reference standard, it is used to correct error samples and eliminate real abnormal samples. Through mean filling correction based on the mean water consumption, the outliers in the error samples are replaced by the mean, thereby correcting the abnormal conditions caused by measurement errors or equipment failures. The generated error-corrected data is closer to the actual water consumption situation, which improves the accuracy and credibility of the data. According to the real abnormal samples, the actual abnormal conditions are eliminated from the data to ensure the accuracy of subsequent analysis and monitoring. The obtained abnormal sample elimination data is more in line with normal water use behavior and eliminates the influence of abnormal conditions. Through local filtering optimization based on error laws, the error correction data and abnormal sample elimination data are further smoothed and corrected. The generated optimized water use monitoring data is more stable and accurate, eliminating the interference of noise and abnormal data on the analysis results, and providing a more reliable data basis for the safety monitoring and analysis of elderly people living alone based on water consumption.

[0028] Preferably, the specific steps of step S2 are:

[0029] Step S21: performing periodic water usage frequency statistics on the optimized water usage monitoring data to generate a periodic water usage frequency;

[0030] Step S22: performing multi-point water consumption peak identification on the optimized water consumption monitoring data to obtain multiple water consumption peak points;

[0031] Step S23: performing periodic water consumption characteristic analysis on multiple water consumption peak points based on periodic water consumption frequency to generate water consumption characteristic data;

[0032] Step S24: Quantitatively mine the time series water consumption distribution characteristics of the water consumption characteristic data, so as to obtain the time series water consumption distribution data within the period;

[0033] Step S25: dynamically render the distribution graph of the time-series water consumption distribution data within the period, and construct a time-series water consumption characteristic distribution graph.

[0034] The present invention uses periodic water use frequency statistics to understand the number of water use times that occur within a certain time period. Based on the statistical results of periodic water use frequency, the water use habits and behavioral patterns of elderly people living alone are revealed, providing a reference for subsequent safety monitoring and analysis. By identifying multi-point water use peaks, multiple peak periods in water use monitoring data, i.e., time periods with large water use, are determined. The identification of multi-point water use peaks helps to discover water use activities of elderly people living alone, including bathing, drinking water, etc., providing a basis for subsequent analysis and safety monitoring. By analyzing the characteristics of periodic water use, the water use characteristics in each period, such as average water use, maximum water use, etc., are calculated, and the generated water use characteristics are The water consumption characteristic data reflects the water consumption of elderly people living alone in each period, providing a data basis for subsequent analysis and safety monitoring. Through quantitative mining of the time-series water consumption distribution characteristics, the distribution of water consumption in different time periods is analyzed, such as the volatility of water consumption, peak-to-valley differences, etc. The generated time-series water consumption distribution data within the period reveals the time-series characteristics of the water consumption behavior of elderly people living alone, which is helpful to discover abnormal water use. Through dynamic rendering of the distribution map, the distribution of time-series water consumption within the period is visualized. The constructed time-series water consumption characteristic distribution map intuitively presents the changing trend and abnormal situation of the water consumption behavior of elderly people living alone, providing an intuitive reference for safety monitoring and analysis.

[0035] Preferably, the specific steps of step S3 are:

[0036] Step S31: Perform dynamic water use behavior analysis on the time series water use characteristic distribution diagram to obtain dynamic water use behavior characteristics;

[0037] Step S32: based on the time series water use characteristic distribution diagram, start and stop calculation of water use behavior point by point is performed to obtain the duration of each water use behavior;

[0038] Step S33: classifying and identifying the water use behavior pattern of the dynamic water use behavior characteristics based on the duration of each water use behavior, thereby generating water use behavior pattern characteristic data;

[0039] Step S34: Perform personalized multi-time point evolution based on the water use behavior pattern characteristic data to construct a personalized domestic water use behavior portrait.

[0040] The present invention identifies different water use behavior characteristics, such as washing, cooking, and laundry, through dynamic water use behavior analysis by observing and analyzing the time series water use characteristic distribution map. Through the obtained dynamic water use behavior characteristics, the water use behavior patterns of elderly people living alone in different time periods are understood, providing a basis for subsequent safety monitoring and analysis. The start and end time points of each water use behavior are determined according to the water use behavior characteristics in the time series water use characteristic distribution map. The duration of each water use behavior obtained reflects the duration of each water use of the elderly living alone, which is helpful for analyzing the frequency and duration distribution of water use behavior. Through water use behavior pattern classification and identification, the dynamic water use behavior characteristics of the elderly living alone are classified, such as short-term water use, long-term water use, frequent water use, etc. The generated water use behavior pattern characteristic data describes the water use behavior pattern of the elderly living alone in more detail, providing data support for subsequent personalized analysis and safety monitoring. The personalized multi-time point evolution reveals the water use behavior preferences and laws of the elderly living alone at different time points by analyzing the water use behavior pattern characteristic data of the elderly living alone. The constructed personalized life water use behavior portrait provides personalized water use behavior analysis and safety monitoring for the elderly living alone, which is helpful for detecting abnormal situations and providing customized services.

[0041] Preferably, the specific steps of step S4 are:

[0042] Step S41: Acquire real-time smart water meter monitoring parameters;

[0043] Step S42: Perform full-cycle water usage fluctuation monitoring on the real-time smart water meter monitoring parameters through personalized domestic water usage behavior portraits, thereby obtaining water usage fluctuation monitoring data;

[0044] Step S43: Analyze the abnormal water use behavior of the water use fluctuation monitoring data and extract the abnormal water use behavior data.

[0045] The present invention obtains the water consumption data and other relevant parameters of the elderly living alone at the current moment, such as flow rate, pressure, etc., by acquiring the real-time monitoring parameters of the smart water meter. The acquisition of the real-time monitoring parameters provides a real-time data basis for the subsequent water use fluctuation monitoring and abnormal water use behavior analysis. Through the personalized domestic water use behavior portrait, the real-time smart water meter monitoring parameters are matched and compared with the water use behavior pattern of the elderly living alone. The full-cycle water use fluctuation monitoring analyzes the changing trends of the real-time monitoring parameters in different time periods, thereby obtaining the water use fluctuation monitoring data and revealing the water consumption fluctuation of the elderly living alone. The abnormal water use behavior analysis detects and identifies the abnormal water use behavior pattern by statistics and comparison of the water use fluctuation monitoring data. The extracted abnormal water use behavior data reflects the abnormal water use of the elderly living alone, such as water leakage, excessive waste, etc., and provides a basis for safety monitoring and early warning.

[0046] Preferably, the specific steps of step S42 are:

[0047] Extract features of real-time smart water meter monitoring parameters to obtain water usage interval features and water consumption features;

[0048] Convert the water usage interval characteristics and water consumption characteristics into time period water usage values ​​to generate water usage characteristic values ​​for each time period;

[0049] Through personalized domestic water use behavior portraits, the water use characteristic values ​​of each period are compared and processed throughout the entire cycle to identify the water use difference periods;

[0050] Water use fluctuation deviation analysis is performed based on water use difference periods to obtain water use fluctuation monitoring data.

[0051] The present invention extracts water use interval characteristics and water consumption characteristics from real-time smart water meter monitoring parameters through feature extraction. The water use interval characteristics reflect the time interval between water use of elderly people living alone, and the water consumption characteristics indicate the specific water consumption of each water use. The time period water use value conversion associates the water use interval characteristics and water consumption characteristics with time periods, and converts them into water use characteristic values ​​of each time period. The generated water use characteristic values ​​of each time period reflect the water use behavior of elderly people living alone in different time periods, such as early morning, morning, afternoon, evening, etc. Through personalized life water use behavior portraits, the water use characteristic values ​​of each time period are compared and analyzed with the water use behavior patterns of elderly people living alone. The full-cycle difference comparison processing identifies the differences in water use behaviors in different time periods, such as higher or lower water consumption in a certain time period. The water use fluctuation deviation analysis evaluates the water use fluctuation of elderly people living alone through statistics and analysis of water use difference time periods, and identifies water use behaviors that deviate from normal patterns. The obtained water use fluctuation monitoring data reflects the stability and variability of water use behaviors of elderly people living alone, and provides a basis for safety monitoring and the discovery of abnormal situations.

[0052] Preferably, the specific steps of step S5 are:

[0053] Step S51: numerically process the abnormal water use behavior data to generate a quantitative value of the abnormal result;

[0054] Step S52: Perform safety status assessment on the quantified value of the abnormal result based on a preset safety assessment model to generate safety status assessment data for the elderly living alone;

[0055] Step S53: Perform safety risk prediction on the safety status assessment data of the elderly living alone to generate safety status risk prediction data.

[0056] The present invention converts abnormal water use behavior data into specific numerical representations through numerical processing for subsequent analysis and processing. The generated abnormal result quantification value provides an assessment of the degree or severity of the abnormal water use behavior, and provides a data basis for subsequent safety status assessment and safety risk prediction. Based on a preset safety assessment model, the safety status of elderly people living alone can be assessed according to the abnormal result quantification value. Through the safety status assessment, a comprehensive analysis is conducted on the water use situation of elderly people living alone to assess whether there are safety hazards or abnormal situations. The safety risk prediction will be based on the safety status assessment data, combined with historical data and model analysis, to make a future risk prediction for the safety status of elderly people living alone. The generated safety status risk prediction data provides early warning of future safety risks that may arise for elderly people living alone, and helps to take corresponding measures to intervene and ensure safety.

[0057] Preferably, the specific steps of step S6 are:

[0058] Step S61: Make safety warning decisions based on safety status risk prediction data and build a safety status warning strategy for elderly people living alone;

[0059] Step S62: dynamically optimizing and modeling the personalized domestic water use behavior portrait based on the real-time smart water meter monitoring parameters to build a water use behavior safety monitoring model;

[0060] Step S63: Execute safety monitoring and early warning operations for elderly people living alone based on the safety status early warning strategy for elderly people living alone and the water use behavior safety monitoring model.

[0061] The present invention makes safety warning decisions based on safety status risk prediction data, that is, it determines whether there is a safety risk according to the prediction results and threshold settings, and formulates corresponding warning strategies, and constructs a safety status warning strategy for elderly people living alone to help monitoring personnel or relevant institutions take timely measures when safety risks are discovered to ensure the safety of elderly people living alone. By dynamically optimizing and modeling the monitoring parameters of real-time smart water meters, personalized domestic water use behavior portraits are continuously updated to more accurately describe the water use behavior patterns of elderly people living alone, and a water use behavior safety monitoring model is constructed to analyze and predict based on the personalized domestic water use behavior portraits, identify abnormal water use behaviors, and assist in safety warning decisions. Based on the safety status warning strategy for elderly people living alone and the water use behavior safety monitoring model, actual safety monitoring operations are carried out. Execution of safety monitoring and warning operations for elderly people living alone includes monitoring and warning the safety status according to the warning strategy, identifying and processing abnormal water use behaviors, timely discovering potential safety risks, and taking necessary intervention measures to ensure the safety of elderly people living alone.

[0062] In this specification, a safety monitoring and early warning device based on water consumption characteristics of a smart water meter is provided, which is used to execute the safety monitoring and early warning method based on water consumption characteristics of a smart water meter as described above, including:

[0063] The extreme value filtering module is used to obtain the historical water consumption monitoring data of the elderly living alone based on the smart water meter; the historical water consumption monitoring data of the elderly living alone is subjected to extreme value filtering optimization processing to generate optimized water consumption monitoring data;

[0064] The time series water consumption characteristic module is used to analyze the periodic water consumption characteristics of the optimized water consumption monitoring data to generate water consumption characteristic data; to quantitatively mine the time series water consumption distribution characteristics of the water consumption characteristic data to construct a time series water consumption characteristic distribution map;

[0065] The water use behavior portrait module is used to obtain dynamic water use behavior characteristics and the duration of each water use behavior based on the time series water use characteristic distribution map; based on the duration of each water use behavior, the dynamic water use behavior characteristics are personalized and evolved at multiple time points to construct a personalized domestic water use behavior portrait;

[0066] The water consumption fluctuation monitoring module is used to obtain real-time smart water meter monitoring parameters; the real-time smart water meter monitoring parameters are monitored for full-cycle water consumption fluctuations through personalized domestic water consumption behavior portraits, and abnormal water consumption behavior data is extracted;

[0067] A safety status assessment module is used to generate abnormal result quantification values ​​based on abnormal water use behavior data; perform safety status assessment on the abnormal result quantification values ​​based on a preset safety assessment model to generate safety status risk prediction data;

[0068] The safety warning decision-making module is used to make safety warning decisions based on safety status risk prediction data and to build a safety status warning strategy for elderly people living alone; based on the safety status warning strategy for elderly people living alone, the safety monitoring and warning operations for elderly people living alone are performed.

[0069] The present invention removes abnormal extreme values ​​or noise in historical water use monitoring data through extreme value filtering optimization processing, thereby improving the accuracy and credibility of the data. The generated optimized water use monitoring data can better reflect the actual water use situation of elderly people living alone, and provide a more reliable data basis for subsequent analysis and modeling. Based on the periodic water consumption feature analysis, the periodic law of water use of elderly people living alone, such as daily average water consumption, weekly average water consumption, etc., is extracted to provide more detailed feature data for subsequent water use behavior analysis. The constructed time series water use feature distribution diagram shows the water use distribution situation in different time periods, which helps to observe and analyze the water use habits and behavior patterns of elderly people living alone. Through the time series water use feature distribution diagram, dynamic water use behavior features, such as water use frequency, water use duration, etc., are extracted to reveal the water use behavior law of elderly people living alone. Based on the personalized multi-time point evolution of the duration of each water use behavior, the changing trend of water use behavior of elderly people living alone in different time periods is analyzed to construct a more detailed personalized life water use behavior portrait. The real-time intelligent water meter monitoring parameters provide the current water use situation of elderly people living alone, including water consumption, water use time and other information. The domestic water use behavior portrait monitors the full-cycle water use fluctuations of the real-time smart water meter monitoring parameters, identifies abnormal water use behaviors, such as sudden increases or decreases in water consumption, and helps to timely discover potential safety risks. By generating quantitative values ​​of abnormal results, the degree or severity of abnormal water use behaviors is quantitatively evaluated, and objective evaluation indicators for the safety status of elderly people living alone are provided. The safety status of the quantitative values ​​of abnormal results is evaluated based on the preset safety assessment model, and the abnormal water use behavior is compared with the preset safety standards to predict the safety status risk level of the elderly living alone. Based on the safety status risk prediction data, safety warning decisions are made to determine whether there are safety risks, and corresponding warning strategies are formulated. The safety status warning strategy for the elderly living alone is constructed to help monitoring personnel or relevant institutions take timely measures when safety risks are discovered to ensure the safety of the elderly living alone. The execution of safety monitoring and early warning operations for the elderly living alone includes monitoring and early warning of the safety status according to the early warning strategy, identifying and processing abnormal water use behaviors, timely discovering potential safety risks, and taking necessary intervention measures to ensure the safety of the elderly living alone. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 A schematic diagram of the steps of a safety monitoring and early warning method based on water consumption characteristics of an intelligent water meter according to the present invention;

[0071] Figure 2 Detailed implementation flow chart of step S1;

[0072] Figure 3 Detailed implementation flow chart of step S2;

[0073] Figure 4 Detailed implementation flow chart of step S3. DETAILED DESCRIPTION

[0074] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0075] The example of this application provides a safety monitoring and early warning method and device based on the water consumption characteristics of smart water meters. The execution subject of the safety monitoring and early warning method and device based on the water consumption characteristics of smart water meters includes but is not limited to the following: mechanical equipment, data processing platform, cloud server node, network upload device, etc. equipped with the system can be regarded as the general computing node of this application, and the data processing platform includes but is not limited to: at least one of an audio and image management system, an information management system, and a cloud data management system.

[0076] See also Figures 1 to 4 The present invention provides a safety monitoring and early warning method based on the water consumption characteristics of a smart water meter, comprising the following steps:

[0077] Step S1: obtaining historical water consumption monitoring data of elderly people living alone based on the smart water meter; performing extreme value filtering and optimization processing on the historical water consumption monitoring data of elderly people living alone, thereby generating optimized water consumption monitoring data;

[0078] Step S2: Perform periodic water consumption characteristic analysis on the optimized water consumption monitoring data to generate water consumption characteristic data; perform quantitative mining on the time series water consumption distribution characteristics of the water consumption characteristic data to construct a time series water consumption characteristic distribution map;

[0079] Step S3: based on the time series water use characteristic distribution map, dynamic water use behavior characteristics and the duration of each water use behavior are obtained; based on the duration of each water use behavior, the dynamic water use behavior characteristics are personalized and evolved at multiple time points to construct a personalized domestic water use behavior portrait;

[0080] Step S4: Acquire real-time smart water meter monitoring parameters; monitor the full-cycle water consumption fluctuation of the real-time smart water meter monitoring parameters through personalized domestic water consumption behavior portraits, and extract abnormal water consumption behavior data;

[0081] Step S5: generating an abnormal result quantification value based on the abnormal water use behavior data; performing a safety status assessment on the abnormal result quantification value based on a preset safety assessment model to generate safety status risk prediction data;

[0082] Step S6: Make safety warning decisions based on the safety status risk prediction data and construct a safety status warning strategy for elderly people living alone; perform safety monitoring and warning operations for elderly people living alone based on the safety status warning strategy for elderly people living alone.

[0083] The present invention obtains historical water consumption monitoring data of elderly people living alone to understand the historical patterns and habits of their water consumption behavior, performs extreme value filtering optimization processing to remove outliers and noise, improves the accuracy and credibility of the data, and generates optimized water consumption monitoring data that better reflects the actual water consumption of elderly people living alone, providing a reliable data basis for subsequent analysis. By analyzing the characteristics of periodic water consumption, the present invention obtains water consumption statistics of elderly people living alone, such as average water consumption, peak water consumption, water consumption cycle, etc., performs time series analysis on water consumption characteristic data, explores the time characteristics of water consumption behavior of elderly people living alone, such as peak water consumption hours, periodic changes in water consumption, etc., and constructs a time series water consumption monitoring system. The time-series water use characteristic distribution map intuitively displays the water use behavior patterns of elderly people living alone, helping to discover anomalies or abnormal trends. Through the time-series water use characteristic distribution map, the dynamic water use behavior characteristics of elderly people living alone, such as the frequency, duration, interval time, etc. of water use behavior, are extracted. Based on the personalized multi-time point evolution of the duration of each water use behavior, the water use behavior patterns of elderly people living alone at different time points are understood, the changes and habits of daily life are captured, and a personalized life water use behavior portrait is constructed to comprehensively describe the water use behavior characteristics of elderly people living alone, providing a personalized basis for subsequent safety monitoring, and obtaining real-time smart water meter monitoring parameters to understand the current water use of elderly people living alone in real time. The system can analyze the situation of water consumption, water use frequency, etc., monitor the fluctuation of water use in the whole cycle of real-time monitoring parameters through personalized life water use behavior portraits, compare them with the individual's historical behavior, find abnormalities or abnormal trends, improve the accuracy of abnormal detection, extract abnormal water use behavior data to help determine potential safety risks, such as water leakage, abnormally high frequency of water use, etc., generate abnormal result quantification values ​​based on abnormal water use behavior data, convert abnormal behavior into specific numerical indicators, and facilitate further analysis and comparison. Based on the preset safety assessment model, the abnormal result quantification values ​​are evaluated for safety status, and the abnormal behavior is compared with the safety standard to evaluate the safety status risk of elderly people living alone. The safety status risk prediction data is generated to discover potential safety problems in advance, provide a basis for taking corresponding prevention and intervention measures, and make safety warning decisions based on the safety status risk prediction data. According to the safety status risk degree of elderly people living alone, formulate corresponding warning strategies, handle safety problems in a graded manner, and build a safety status warning strategy for elderly people living alone. According to the specific situation, formulate targeted monitoring and intervention measures to improve the accuracy and effectiveness of safety warnings. By executing the safety monitoring and early warning operations for elderly people living alone, the water use behavior of elderly people living alone is monitored in real time, abnormal situations are discovered in time, and corresponding early warning measures are taken to ensure their safety.

[0084] In the embodiment of the present invention, refer to Figure 1 , is a schematic flow chart of the steps of a safety monitoring and early warning method based on water consumption characteristics of a smart water meter according to the present invention. In this example, the steps of the safety monitoring and early warning method based on water consumption characteristics of a smart water meter include:

[0085] Step S1: obtaining historical water consumption monitoring data of elderly people living alone based on the smart water meter; performing extreme value filtering and optimization processing on the historical water consumption monitoring data of elderly people living alone, thereby generating optimized water consumption monitoring data;

[0086] In this embodiment, a smart water meter is installed to collect water use data of the residence of an elderly person living alone in real time, including water consumption, water use time and other information. The collected data is transmitted to a cloud database through the Internet of Things technology to establish historical water use monitoring data of the elderly living alone. The frequency of data collection can be adjusted appropriately according to demand. Usually, data is collected multiple times a day to obtain more detailed water use information. The historical water use monitoring data is analyzed to identify abnormal values ​​or extreme values, such as a sudden increase in water consumption on a certain day. Statistical methods, such as the standard deviation method and the Z-score method, are used to filter and correct these abnormal values. The filtered data is more in line with the normal distribution of daily water use of the elderly, and noise and interference factors are greatly reduced. Statistical analysis is performed on the water consumption data, such as calculating the mean value and standard deviation. Etc. These statistical indicators are used to identify outliers. According to the results of statistical analysis, reasonable thresholds or ranges are set to determine which water consumption data are outliers. The historical water consumption monitoring data are traversed to check whether the water consumption data exceeds the set threshold or range. If exceeded, it is marked as an outlier. The data marked as outliers are processed and optimized by deletion, correction or interpolation. Deleting outliers means excluding them from the optimized data. Correcting outliers is corrected according to the trend of surrounding data. Interpolation methods are filled according to the changing trend of surrounding data. After completing the extreme value filtering optimization process, the optimized water consumption monitoring data is obtained, in which outliers and noise have been removed, and relatively accurate and reliable water consumption data is retained.

[0087] Step S2: Perform periodic water consumption characteristic analysis on the optimized water consumption monitoring data to generate water consumption characteristic data; perform quantitative mining on the time series water consumption distribution characteristics of the water consumption characteristic data to construct a time series water consumption characteristic distribution map;

[0088] In this embodiment, the time interval or time period of the water use data is determined to be hours, days, weeks, etc., the optimized water use monitoring data is grouped according to the time interval or time period, the water use data in each period is counted, and the water use data in each period is statistically analyzed, including calculating the average value, maximum value, minimum value, total value and other indicators. These indicators reflect the typical characteristics of water use in the period, and further analyze the distribution of the periodic water use data, such as drawing histograms, box plots, etc., to understand the distribution range and skewness of water use. Based on the periodic water use characteristic data, a time series water use characteristic distribution diagram is drawn, in which the horizontal axis represents time and the vertical axis represents water use. The numerical value at each time point represents the water consumption characteristic indicator value within the time period. Different chart types, such as line charts, bar charts, etc., are selected to intuitively display the changes of water consumption characteristics over time. The time series water consumption characteristic distribution diagram is analyzed to observe the changing trend and periodic pattern of water consumption characteristic indicators over time, find recurring patterns, peaks and valleys, and changes in long-term trends, and use statistical methods or time series analysis methods to further explore the characteristics of time series water consumption distribution, for example, calculate indicators such as mean, standard deviation, trend line, or apply smoothing methods, periodic decomposition methods, etc. for analysis.

[0089] Step S3: based on the time series water use characteristic distribution map, dynamic water use behavior characteristics and the duration of each water use behavior are obtained; based on the duration of each water use behavior, the dynamic water use behavior characteristics are personalized and evolved at multiple time points to construct a personalized domestic water use behavior portrait;

[0090] In this embodiment, the starting point and the ending point of the water use behavior are identified according to the time-series water use characteristic distribution diagram, the start and the end of the water use behavior are defined according to a specific threshold value or pattern, the duration of each water use behavior is calculated, that is, the end time of the water use behavior minus the start time, and the characteristics of each water use behavior are extracted, including water consumption, water use time period, water use frequency, etc. These characteristics are used to describe the pattern and characteristics of the water use behavior, the duration of each water use behavior is analyzed, the distribution of the duration of the water use behavior in different time periods is observed, and a histogram or a box plot is drawn to represent the distribution of the duration. According to the distribution of the duration, the division points of different time periods are determined, for example, according to the percentile or quantile of the duration, and the dynamic water use behavior characteristics are grouped according to the time period, for example, the water use behavior is divided into three time periods of short, medium and long time. In each time period, the personalized evolution of the dynamic water use behavior characteristics is analyzed and described respectively, the average water consumption, water use frequency and other indicators in each time period are calculated, and the differences between different time periods are compared, thereby revealing the personalized characteristics and evolution laws of water use behavior in different time periods.

[0091] Step S4: Acquire real-time smart water meter monitoring parameters; monitor the full-cycle water consumption fluctuation of the real-time smart water meter monitoring parameters through personalized domestic water consumption behavior portraits, and extract abnormal water consumption behavior data;

[0092] In this embodiment, real-time smart water meter monitoring parameters are obtained. The real-time smart water meter provides instant water use monitoring data, including parameters such as water consumption and water use time. These parameters are obtained through the connection or data interface with the smart water meter device and monitored in real time or near real time. Based on the personalized domestic water use behavior portrait, the typical pattern and characteristics of water use behavior are determined, including indicators such as average water consumption and water use frequency in each time period, and the water use parameters of the real-time smart water meter are monitored. For example, the current water consumption is obtained once at a certain time interval, and the real-time water use parameters are compared with the personalized domestic water use behavior portrait. The difference between the current water consumption and the typical pattern is calculated, and changes such as water use frequency in the time period are considered. According to the degree of difference and a pre-set threshold, it is determined whether the current water use behavior is abnormal. If the water use behavior exceeds the normal range, it is marked as abnormal water use behavior. According to the time point or time period of the abnormal water use behavior, the corresponding data is extracted from the real-time smart water meter monitoring parameters. The extracted data includes information such as the timestamp, water consumption, and water use time of the abnormal water use behavior. The abnormal water use behavior data is further analyzed, such as calculating the duration of the abnormal water use behavior, the change trend of water consumption, etc.

[0093] Step S5: generating an abnormal result quantification value based on the abnormal water use behavior data; performing a safety status assessment on the abnormal result quantification value based on a preset safety assessment model to generate safety status risk prediction data;

[0094] In this embodiment, the abnormal water use behavior data is preprocessed, including data cleaning, denoising and format conversion, etc., to ensure the accuracy and consistency of the data. According to predefined indicators and rules, the abnormal water use behavior data is converted into corresponding quantitative values, for example, the duration of abnormal water use behavior, the change range of water consumption, etc. are calculated, and the input and output of the safety assessment model are defined. The input includes the quantitative value of the abnormal result and other parameters related to the safety status. The output is the safety status assessment result or the safety status risk prediction data. According to the preset safety assessment model, the quantitative value of the abnormal result is input into the model for evaluation. The model uses machine learning, statistical analysis or expert rules and other methods to perform the evaluation. According to the output of the safety assessment model, the safety status risk prediction data is generated. The data includes information such as the level of the safety status, the degree of risk, and recommended measures. According to the safety status risk prediction data, the current safety status level and risk degree are determined, and judgment is made according to predefined standards or indicators. The main factors leading to the safety status risk and the characteristics of the abnormal water use behavior are analyzed, and the correlation, trend and causes of the abnormal behavior are explored, and corresponding explanations and suggestions are provided, such as providing corresponding measures and improvement suggestions for specific safety status risk levels.

[0095] Step S6: Make safety warning decisions based on the safety status risk prediction data and construct a safety status warning strategy for elderly people living alone; perform safety monitoring and warning operations for elderly people living alone based on the safety status warning strategy for elderly people living alone.

[0096] In this embodiment, rules and standards for safety warning decision-making are defined, including warning actions corresponding to different safety status risk levels, threshold settings for triggering warnings, etc., and judgment is made based on safety status risk prediction data to determine whether a safety warning is triggered. According to pre-set rules, the relationship between the current safety status risk level and the warning threshold is compared. If a safety warning is triggered, the corresponding warning action is executed, including sending an alarm notification, automatically dialing an emergency contact number, etc., analyzing and understanding the specific needs and risk factors of elderly people living alone, considering factors such as age, health status, and living environment, and identifying safety risks. According to the analysis results, a safety status warning strategy for elderly people living alone is formulated, for example, warning rules and handling measures are set for abnormal water use, long-term inactivity, and too high or too low temperature, and the execution method and process of the warning strategy are determined, including who receives the alarm notification, how to contact the emergency contact, how to coordinate relevant rescue measures, etc., and safety monitoring and warning devices and equipment are deployed, including smart sensors, surveillance cameras, emergency buttons and other equipment, which are used to monitor the safety of elderly people living alone in real time and monitor the safety status of elderly people living alone. According to the rules and indicators defined in the safety status warning strategy for elderly people living alone, the behavior and environment of elderly people living alone are monitored and analyzed. According to the security warning decision and warning strategy, the corresponding warning action is triggered. If an abnormality or risk situation is detected, the corresponding warning and response are carried out according to the pre-set processing measures.

[0097] In this embodiment, refer to Figure 2 , is a flowchart of detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include:

[0098] Step S11: Acquire historical water consumption monitoring data based on the smart water meter;

[0099] Step S12: Performing a distribution analysis of the value range of the historical water monitoring data to obtain a normalized value range;

[0100] Step S13: Identify abnormal samples of historical water consumption monitoring data based on the normalized value range and extract abnormal monitoring samples;

[0101] Step S14: Perform extreme value filtering optimization processing based on the monitored abnormal samples to generate optimized water use monitoring data.

[0102] In this embodiment, a smart water meter system is deployed, smart water meter equipment is installed, and its normal operation and data collection function are ensured to be available. Historical water consumption monitoring data is obtained through the smart water meter system. The smart water meter system provides records and storage of historical water consumption, including timestamps and water consumption values. Distribution analysis of the value range of monitoring data is performed. Statistics and distribution analysis are performed on historical water monitoring data to understand the frequency and distribution of different value ranges, determine the normalized value range, and determine the normalized value range based on the analysis results, that is, the normal range of water consumption. Based on statistical methods, empirical rules or expert knowledge, abnormal samples of historical water consumption monitoring data are identified, and historical water consumption is classified into The water consumption monitoring data is compared with the normalized value range, abnormal samples beyond the range are identified, and monitoring abnormal samples are extracted. The identified monitoring abnormal samples are extracted from the historical water consumption monitoring data for subsequent processing and optimization, and extreme value filtering is performed. For monitoring abnormal samples, extreme value filtering is performed to eliminate abnormal values ​​caused by equipment failure, abnormal conditions or erroneous readings, and generate optimized water consumption monitoring data. Based on the monitoring abnormal samples that have been processed by extreme value filtering and combined with the original historical water consumption monitoring data, optimized water consumption monitoring data is generated. These data more accurately reflect the actual water use situation and reduce misjudgment or misleading caused by abnormal samples.

[0103] In this embodiment, the specific steps of step S14 are:

[0104] Performing an abnormal type analysis on the monitored abnormal samples to obtain abnormal sample types; the abnormal sample types include true abnormal samples and error samples;

[0105] The timestamp of the error sample is extracted;

[0106] Performing error sample timing analysis based on the timestamp to generate error sample timing data;

[0107] Perform multivariate analysis on error sample time series data to identify error patterns;

[0108] Calculate the average value of historical water use monitoring data to obtain the mean water use value;

[0109] Perform mean filling correction on the error samples based on the water mean to generate error correction data;

[0110] Perform anomaly elimination processing based on real abnormal samples to obtain abnormal sample elimination data;

[0111] According to the error law, local filtering optimization is performed on the error correction data and abnormal sample elimination data to generate optimized water use monitoring data.

[0112] In this embodiment, the extracted abnormal samples are analyzed to distinguish between true abnormal samples and error samples. According to the abnormal degree of water consumption, change trend and other characteristics, the abnormal samples are divided into true abnormalities caused by equipment failure, true abnormalities caused by changes in the living habits of the elderly, and error samples caused by instrument errors. If a sample shows that the water consumption suddenly surges to 500L within one day, it can be judged as a true abnormality caused by equipment failure. When the sample shows that the water consumption fluctuates abnormally in the morning and evening, it can be judged as a true abnormality caused by changes in the living habits of the elderly. When the sample shows that the water consumption fluctuates abnormally in the morning and evening, it can be judged as a true abnormality caused by changes in the living habits of the elderly. There are large deviations in some time periods, which can be judged as error samples caused by instrument errors. For the identified error samples, the timestamp information of their occurrence is extracted. For example, some error samples appear between 6 and 8 in the morning, and some appear between 2 and 4 in the afternoon. The extracted error sample timestamps are arranged in chronological order to form time series data of the error samples. Combined with the water consumption data at each time point, a data set reflecting the time series changes of the error samples is constructed to obtain a time series data similar to [6:00, 30L], [6:30, 25L], [7:00, 35L], etc. The error sample time series data is statistically analyzed, such as calculating the mean value, variance, etc., and combined with environmental factor data, such as temperature and humidity, to perform correlation analysis. Through these analyses, it is preliminarily found that the error samples are associated with certain time periods or environmental factors. For example, it is found that the error samples between 6 and 8 in the morning are positively correlated with the temperature increase. The average water consumption is calculated for the entire historical water use monitoring data set to obtain the average daily water consumption of the elderly. For example, statistical analysis finds that the average daily water consumption of the elderly is about 90L. For the error samples identified before, their water consumption is replaced with the average daily water consumption. For example, an error sample shows that the water consumption is 150L, ​​which is corrected to 90L. In this way, the error-corrected water use monitoring data is generated, and the real abnormal samples are removed from the original monitoring data. In this way, the water use monitoring data with the real abnormal samples removed is obtained. Combined with the error law obtained in the previous analysis, the error correction data and the abnormal removal data are locally filtered. Stronger filtering is applied to the data from 6 to 8 in the morning, while weaker filtering is applied to other time periods. Through this targeted local filtering, optimized high-quality water use monitoring data is finally obtained.

[0113] In this embodiment, refer to Figure 3 , is a flowchart of detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include:

[0114] Step S21: performing periodic water usage frequency statistics on the optimized water usage monitoring data to generate a periodic water usage frequency;

[0115] Step S22: performing multi-point water consumption peak identification on the optimized water consumption monitoring data to obtain multiple water consumption peak points;

[0116] Step S23: performing periodic water consumption characteristic analysis on multiple water consumption peak points based on periodic water consumption frequency to generate water consumption characteristic data;

[0117] Step S24: Quantitatively mine the time series water consumption distribution characteristics of the water consumption characteristic data, so as to obtain the time series water consumption distribution data within the period;

[0118] Step S25: dynamically render the distribution graph of the time-series water consumption distribution data within the period, and construct a time-series water consumption characteristic distribution graph.

[0119] In this embodiment, the optimized water use monitoring data is divided into periods, and a fixed period (such as daily or weekly) is used or an appropriate period is determined according to the characteristics of the data. The water use frequency in each period, that is, the number of times water is used in the period, is counted. According to the statistical results, periodic water use frequency data is generated, and the number of times water is used in each period is recorded. For the optimized water use monitoring data, an algorithm for peak recognition is selected. For example, based on methods such as thresholds and peak detection, peak recognition is performed on the optimized water use monitoring data to determine multiple water use peak points, that is, points with the highest water consumption. Timestamps and water consumption information of multiple water use peak points are extracted. According to the periodic water use frequency data, the water use peak points in each period are determined. Feature analysis is performed on the water use peak points in each period, such as calculating statistical indicators such as the average value, maximum value, and minimum value of the peak points, generating water consumption feature data, recording feature information of the water use peak points in each period, and collecting water consumption. Characteristic data, including characteristic information of peak water consumption points within a cycle, time series analysis of water consumption characteristic data, such as constructing a time series distribution histogram or density estimation diagram of water consumption, quantitatively mining the characteristics of time series water consumption distribution, such as peak position, peak height, distribution form, etc., generating time series water consumption distribution data within the cycle according to the analysis results, recording the characteristic information of water consumption distribution within each cycle, using the time series water consumption distribution data within the cycle, selecting a method for dynamically rendering the distribution graph, such as using data visualization tools or programming languages ​​for visualization operations, generating dynamically rendered charts or images based on the time series water consumption distribution data within the cycle, displaying the characteristics of time series water consumption distribution in the chart or image, such as peak position, peak height, etc., using different colors, marks or other visual elements to represent the characteristics, and constructing a time series water consumption characteristic distribution graph to intuitively display the distribution characteristics of time series water consumption within the cycle.

[0120] Example:

[0121] The water consumption monitoring data of the elderly for one week is as follows:

[0122] Water consumption per time (liters)

[0123] Monday 8:00 15

[0124] Monday 8:30 25

[0125] Monday 9:00 10

[0126] Monday 9:30 5

[0127] Monday 12:00 40

[0128] Monday 12:30 20

[0129] Monday 19:00 30

[0130] Monday 19:30 15

[0131] Tuesday 8:00 20

[0132] Tuesday 8:30 30

[0133] Divide the day into 48 30-minute time periods

[0134] The frequency of water use events in each time period is counted to obtain the following periodic water use frequency distribution:

[0135] Frequency of water use during a period of time (times / half hour)

[0136] 8:00-8:30 2

[0137] 8:30-9:00 1

[0138] 9:00-9:30 1

[0139] 9:30-10:00 0 ...

[0140] 11:30-12:00 1

[0141] 12:00-12:30 2

[0142] 12:30-13:00 1 ...

[0143] 18:30-19:00 1

[0144] 19:00-19:30 1

[0145] 19:30-20:00 1

[0146] Traverse the water consumption data and identify the local maximum as the peak point of water consumption

[0147] From Monday’s data, we can identify three peak points in water usage:

[0148] 8:30 Water consumption 25 liters

[0149] 12:00 Water consumption 40 liters

[0150] 19:00 Water consumption 30 liters

[0151] For each water consumption peak point, count its water consumption and duration:

[0152] 8:30 Peak point: 25 liters of water, lasting 10 minutes

[0153] 12:00 Peak point: 40 liters of water used for 15 minutes

[0154] 19:00 Peak point: 30 liters of water used for 10 minutes

[0155] Statistics of total water consumption and water use frequency in each time period:

[0156] 8:00-9:00 Total 50 liters, 4 times

[0157] 9:00-10:00 15 liters, 1 time

[0158] 11:00-12:00 Total 10 liters, 1 time

[0159] 12:00-13:00 Total 60 liters, 3 times

[0160] 18:00-19:00 30 liters, 1 time

[0161] 19:00-20:00 15 liters, 1 time

[0162] Dynamically render the distribution graph of the time-series water consumption distribution data within the cycle and construct a time-series water consumption characteristic distribution graph

[0163] The above time series water usage distribution data is visualized as a dynamic heat map.

[0164] In this embodiment, reference Figure 4 The above is a schematic flow chart of the detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include:

[0165] Step S31: Perform dynamic water use behavior analysis on the time series water use characteristic distribution diagram to obtain dynamic water use behavior characteristics;

[0166] Step S32: based on the time series water use characteristic distribution diagram, start and stop calculation of water use behavior point by point is performed to obtain the duration of each water use behavior;

[0167] Step S33: classifying and identifying the water use behavior pattern of the dynamic water use behavior characteristics based on the duration of each water use behavior, thereby generating water use behavior pattern characteristic data;

[0168] Step S34: Perform personalized multi-time point evolution based on the water use behavior pattern characteristic data to construct a personalized domestic water use behavior portrait.

[0169] In this embodiment, the time-series water use characteristic distribution diagram is dynamically analyzed to observe the temporal variation pattern and trend of water use behavior, and extract dynamic water use behavior characteristics, such as the frequency, amplitude, duration, etc. of water use behavior. For example, it is found that the elderly usually have a water use peak around 7 o'clock, and also have a large water consumption at noon and in the evening. By analyzing these dynamic changes, the typical water use behavior characteristics of the elderly are extracted, such as the peak water use period, duration, water consumption distribution, etc. The time-series water use characteristic distribution diagram is further analyzed to identify the start and end time points of each water use behavior. For example, it is found that the elderly have a water use behavior with a duration of about 30 minutes between 7:00 and 7:30, and a water use behavior with a duration of about 45 minutes between 12:00 and 12:45. The duration data of each water use behavior is obtained, and combined with the obtained dynamic water use Behavioral characteristics and water use behavior are durations. Pattern recognition is performed on the water use behavior of the elderly, and water use behavior is divided into "short-term water use", "medium water use", "long-term water use" and other modes. The corresponding characteristic data are the number of times each water use mode occurs in a day, the duration distribution, etc. The water use behavior pattern characteristic data is dynamically modeled and evolved over time. It is found that the water use behavior of the elderly is different on weekdays and weekends, and will change with the seasons. Through long-term analysis, a personalized life water use behavior portrait of the elderly is constructed to reflect their water use characteristics at different time nodes. According to the individual water use behavior pattern characteristic data, personalized multi-time point evolution is performed. At different time points, according to the individual water use behavior pattern characteristic data, the water use behavior portrait of a specific period of time is displayed, and it is visualized through charts, images or other forms.

[0170] Example:

[0171] Divide a day into five periods: early morning (0:00-6:00), morning (6:00-12:00), noon (12:00-14:00), afternoon (14:00-18:00), and evening (18:00-24:00).

[0172] The water consumption and water usage interval characteristics in each time period are counted to obtain the water usage characteristic values ​​of each time period.

[0173] The water consumption characteristics of an elderly person living alone at different time periods are as follows

[0174] Early morning: average water consumption 10L, average interval 90 minutes

[0175] Morning: Average water consumption 35L, average interval 40 minutes

[0176] Noon: Average water consumption 20L, average interval 50 minutes

[0177] Afternoon: Average water consumption 30L, average interval 45 minutes

[0178] Night: Average water consumption 25L, average interval 55 minutes

[0179] Create a water usage behavior profile for the user and use his / her usual water usage characteristics as the "normal" water usage pattern.

[0180] The water usage characteristics obtained for each time period are compared with the user's "normal" water usage pattern.

[0181] Identify time periods when there are significant deviations from the user's normal water usage habits as "water usage difference periods".

[0182] Take a certain user as an example. His water consumption in the morning is usually about 35L, but on a certain morning his water consumption reaches 55L, which is a large deviation and can be identified as a period of water consumption difference.

[0183] The deviation between actual water consumption and expected water consumption is calculated during the different water consumption periods.

[0184] Taking the morning water consumption difference period of the above-mentioned user as an example, the actual water consumption was 55L, while the expected water consumption was 35L, with a deviation of 57%.

[0185] This fluctuation range of water use can be used as a basis for monitoring water use anomalies, reflecting abnormal changes in water use conditions.

[0186] In this embodiment, step S4 includes the following steps:

[0187] Step S41: Acquire real-time smart water meter monitoring parameters;

[0188] Step S42: Perform full-cycle water usage fluctuation monitoring on the real-time smart water meter monitoring parameters through personalized domestic water usage behavior portraits, thereby obtaining water usage fluctuation monitoring data;

[0189] Step S43: Analyze the abnormal water use behavior of the water use fluctuation monitoring data and extract the abnormal water use behavior data.

[0190] In this embodiment, a smart water meter system is used to connect to a real-time monitoring device, and real-time monitoring parameters such as current water consumption, flow, pressure and other information are obtained through the smart water meter system. The real-time smart water meter monitoring parameters and the personalized domestic water use behavior portrait are combined to perform full-cycle water use fluctuation monitoring, and the real-time monitoring parameters are compared and analyzed with the personalized domestic water use behavior portrait to identify water use fluctuations. The water use fluctuation monitoring data is recorded, including the time, amplitude, duration and other information of the water use fluctuations. The water use fluctuation monitoring data is analyzed for abnormal water use behavior to identify abnormal water use behavior. Statistical methods, machine learning or pattern recognition technology are used to perform anomaly detection and classification on the water use fluctuation data, and abnormal water use behavior data is extracted, including information such as the time period of abnormal water use, the type of anomaly, and the cause of the abnormal water use. For example, it is found that the elderly suddenly use a large amount of water at night, or the water use increases sharply during the lunch break. These abnormal water use behavior data reflect changes in the health status of the elderly, which require further attention and analysis.

[0191] In this embodiment, step S42 includes the following steps:

[0192] Extract features of real-time smart water meter monitoring parameters to obtain water usage interval features and water consumption features;

[0193] Convert the water usage interval characteristics and water consumption characteristics into time period water usage values ​​to generate water usage characteristic values ​​for each time period;

[0194] Through personalized domestic water use behavior portraits, the water use characteristic values ​​of each period are compared and processed throughout the entire cycle to identify the water use difference periods;

[0195] Water use fluctuation deviation analysis is performed based on water use difference periods to obtain water use fluctuation monitoring data.

[0196] In this embodiment, real-time smart water meter monitoring parameters are obtained, such as the timestamp and water consumption of each record, and the water use interval feature, that is, the time interval between two consecutive water use records, is calculated according to the timestamp, and water consumption features are extracted, such as the total water use and average water use of each water use record. Time periods are defined, such as divided by hours, days, or specific time periods. According to the defined time periods, the water use interval features and water consumption features are classified into each time period, and the water use interval features in each time period are numerically converted, such as calculating the average interval time, the maximum interval time, etc. The water consumption features in each time period are numerically converted, such as calculating the total water use, the average water use, etc., and the water use of each time period is converted. Compare and analyze the characteristic values ​​with the personalized domestic water use behavior portrait, compare the water use characteristic values ​​of each time period with the typical value or average value of the corresponding time period in the personalized domestic water use behavior portrait, identify the water use difference time period, that is, the time period that is significantly different from the typical behavior in the personalized domestic water use behavior portrait, and perform fluctuation deviation analysis on the data of the water use difference time period to detect whether there is abnormal water use behavior. Use statistical methods, machine learning or pattern recognition technology, etc., according to the individual's water use behavior portrait and historical data, to determine whether the water use fluctuation deviates from the normal range, generate water use fluctuation monitoring data based on the analysis results, and record information such as the time period, degree of fluctuation, and cause of the abnormal water use behavior.

[0197] Example:

[0198] Morning (6:00-9:00): water use interval 15min, water consumption 50L

[0199] Morning (9:00-12:00): water use interval 25min, water use 20L

[0200] Noon (12:00-14:00): water use interval 50min, water use 10L

[0201] Afternoon (14:00-17:00): water use interval 20min, water consumption 30L

[0202] Evening (17:00-21:00): water use interval 10min, water consumption 40L

[0203] Nighttime (21:00-6:00): water use interval 5min, water consumption 15L

[0204] The analysis found that the water usage characteristics of the elderly at night (21:00-6:00) were obviously inconsistent with their normal life portrait, and the frequency and total water consumption were beyond the normal range.

[0205] At the same time, the fluctuation range of water consumption during this period has increased significantly compared to usual times, and there may be abnormal situations that require urgent assistance.

[0206] In this embodiment, step S5 includes the following steps:

[0207] Step S51: numerically process the abnormal water use behavior data to generate a quantitative value of the abnormal result;

[0208] Step S52: Perform safety status assessment on the quantified value of the abnormal result based on a preset safety assessment model to generate safety status assessment data for the elderly living alone;

[0209] Step S53: Perform safety risk prediction on the safety status assessment data of the elderly living alone to generate safety status risk prediction data.

[0210] In this embodiment, the extracted abnormal water use behavior data, such as large amounts of water use at night and a sudden increase in water use during lunch break, are classified and quantified. Different quantitative indicators are set according to the severity of the abnormal water use behavior. If the water use at night exceeds the normal value by 20%, the abnormal result quantification value is 3 points, if the water use during lunch break increases by 50%, the abnormal result quantification value is 4 points, and if the water use interval is extended by more than 1 hour, the abnormal result quantification value is 5 points. Based on the previous survey and test data, a safety assessment model that comprehensively considers abnormal water use behavior, the health status of the elderly, and their ability to take care of themselves is constructed. Based on the calculated abnormal result quantity, the abnormal water use behavior, the health status of the elderly, and their ability to take care of themselves are evaluated. The quantified value is input into the safety assessment model. The model will evaluate the current safety status of the elderly according to the size of the quantified value. If the quantified value of the abnormal result is between 5 and 10 points, it is assessed as a good safety status. If the quantified value of the abnormal result is between 10 and 15 points, it is assessed as a general safety status. If the quantified value of the abnormal result is above 15 points, it is assessed as a poor safety status. Based on historical data and safety assessment results, a safety risk prediction model is established. The safety risk prediction model is used to predict the safety status of elderly people living alone, identify potential safety risks, generate safety status risk prediction data, and record information such as the time, risk level or probability of safety risks.

[0211] In this embodiment, step S6 includes the following steps:

[0212] Step S61: Make safety warning decisions based on safety status risk prediction data and build a safety status warning strategy for elderly people living alone;

[0213] Step S62: dynamically optimizing and modeling the personalized domestic water use behavior portrait based on the real-time smart water meter monitoring parameters to build a water use behavior safety monitoring model;

[0214] Step S63: Execute safety monitoring and early warning operations for elderly people living alone based on the safety status early warning strategy for elderly people living alone and the water use behavior safety monitoring model.

[0215] In this embodiment, a safety warning decision strategy is formulated based on preset rules, thresholds or machine learning models, and whether to trigger a safety warning is determined according to the risk level or probability in the safety status risk prediction data. If the safety warning is triggered, the corresponding warning operation is performed, such as sending an alarm notification, contacting an emergency contact, etc. The safety warning decision strategy is continuously optimized, updated and improved according to actual feedback and adjustment needs, and personalized domestic water use behavior portrait data, including historical water use patterns, water use habits and other information, is collected. Dynamic optimization modeling is performed in combination with real-time smart water meter monitoring parameters and personalized domestic water use behavior portrait data. Machine learning, statistical modeling and other methods are used to construct a water use behavior safety monitoring model. The water use behavior safety monitoring model determines the current water use behavior based on the real-time smart water meter monitoring parameters. To determine whether it conforms to the individual's normal pattern, the water use behavior safety monitoring model is continuously updated and optimized, and improvements are made based on new monitoring data and user feedback. In combination with the safety status early warning strategy for elderly people living alone and the water use behavior safety monitoring model, a safety monitoring and early warning operation plan for elderly people living alone is formulated, and safety monitoring operations are implemented. Real-time smart water meter monitoring parameters and water use behavior safety monitoring models are used for monitoring. During the monitoring process, based on the results of the safety early warning decision-making strategy and the water use behavior safety monitoring model, it is determined whether there are safety risks or abnormal water use behaviors. If safety risks or abnormal water use behaviors are found, appropriate measures are taken in a timely manner, such as sending alarms, notifying relevant personnel, or providing emergency assistance. The safety monitoring and early warning operations for elderly people living alone are regularly evaluated and optimized, and improvements and adjustments are made based on actual conditions and user feedback.

[0216] In this embodiment, a safety monitoring and early warning device based on water consumption characteristics of a smart water meter is provided, which is used to execute the safety monitoring and early warning method based on water consumption characteristics of a smart water meter as described above, including:

[0217] The extreme value filtering module is used to obtain the historical water consumption monitoring data of the elderly living alone based on the smart water meter; the historical water consumption monitoring data of the elderly living alone is subjected to extreme value filtering optimization processing to generate optimized water consumption monitoring data;

[0218] The time series water consumption characteristic module is used to analyze the periodic water consumption characteristics of the optimized water consumption monitoring data to generate water consumption characteristic data; to quantitatively mine the time series water consumption distribution characteristics of the water consumption characteristic data to construct a time series water consumption characteristic distribution map;

[0219] The water use behavior portrait module is used to obtain dynamic water use behavior characteristics and the duration of each water use behavior based on the time series water use characteristic distribution map; based on the duration of each water use behavior, the dynamic water use behavior characteristics are personalized and evolved at multiple time points to construct a personalized domestic water use behavior portrait;

[0220] The water consumption fluctuation monitoring module is used to obtain real-time smart water meter monitoring parameters; the real-time smart water meter monitoring parameters are monitored for full-cycle water consumption fluctuations through personalized domestic water consumption behavior portraits, and abnormal water consumption behavior data is extracted;

[0221] A safety status assessment module is used to generate abnormal result quantification values ​​based on abnormal water use behavior data; perform safety status assessment on the abnormal result quantification values ​​based on a preset safety assessment model to generate safety status risk prediction data;

[0222] The safety warning decision-making module is used to make safety warning decisions based on safety status risk prediction data and to build a safety status warning strategy for elderly people living alone; based on the safety status warning strategy for elderly people living alone, the safety monitoring and warning operations for elderly people living alone are performed.

[0223] The present invention removes abnormal extreme values ​​or noise in historical water use monitoring data through extreme value filtering optimization processing, thereby improving the accuracy and credibility of the data. The generated optimized water use monitoring data can better reflect the actual water use situation of elderly people living alone, and provide a more reliable data basis for subsequent analysis and modeling. Based on the periodic water consumption feature analysis, the periodic law of water use of elderly people living alone, such as daily average water consumption, weekly average water consumption, etc., is extracted to provide more detailed feature data for subsequent water use behavior analysis. The constructed time series water use feature distribution diagram shows the water use distribution situation in different time periods, which helps to observe and analyze the water use habits and behavior patterns of elderly people living alone. Through the time series water use feature distribution diagram, dynamic water use behavior features, such as water use frequency, water use duration, etc., are extracted to reveal the water use behavior law of elderly people living alone. Based on the personalized multi-time point evolution of the duration of each water use behavior, the changing trend of water use behavior of elderly people living alone in different time periods is analyzed to construct a more detailed personalized life water use behavior portrait. The real-time intelligent water meter monitoring parameters provide the current water use situation of elderly people living alone, including water consumption, water use time and other information. The domestic water use behavior portrait monitors the full-cycle water use fluctuations of the real-time smart water meter monitoring parameters, identifies abnormal water use behaviors, such as sudden increases or decreases in water consumption, and helps to timely discover potential safety risks. By generating quantitative values ​​of abnormal results, the degree or severity of abnormal water use behaviors is quantitatively evaluated, and objective evaluation indicators for the safety status of elderly people living alone are provided. The safety status of the quantitative values ​​of abnormal results is evaluated based on the preset safety assessment model, and the abnormal water use behavior is compared with the preset safety standards to predict the safety status risk level of the elderly living alone. Based on the safety status risk prediction data, safety warning decisions are made to determine whether there are safety risks, and corresponding warning strategies are formulated. The safety status warning strategy for the elderly living alone is constructed to help monitoring personnel or relevant institutions take timely measures when safety risks are discovered to ensure the safety of the elderly living alone. The execution of safety monitoring and early warning operations for the elderly living alone includes monitoring and early warning of the safety status according to the early warning strategy, identifying and processing abnormal water use behaviors, timely discovering potential safety risks, and taking necessary intervention measures to ensure the safety of the elderly living alone.

[0224] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0225] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A safety monitoring and early warning method based on water consumption characteristics of smart water meters, characterized in that: The following steps are involved: Step S1: Based on the smart water meter, historical water consumption monitoring data of the elderly living alone is obtained; the historical water consumption monitoring data of the elderly living alone is subjected to extreme value filtering and optimization processing, thereby generating optimized water consumption monitoring data; the specific steps of step S1 are: Step S11: Acquire historical water consumption monitoring data based on the smart water meter; Step S12: Performing a distribution analysis of the value range of the historical water monitoring data to obtain a normalized value range; Step S13: Identify abnormal samples of historical water consumption monitoring data based on the normalized value range and extract abnormal monitoring samples; Step S14: Perform extreme value filtering optimization processing according to the abnormal monitoring samples to generate optimized water use monitoring data; wherein the specific steps of step S14 are: Performing an abnormal type analysis on the monitored abnormal samples to obtain abnormal sample types; the abnormal sample types include true abnormal samples and error samples; The timestamp of the error sample is extracted; Performing error sample timing analysis based on the timestamp to generate error sample timing data; Perform multivariate analysis on error sample time series data to identify error patterns; Calculate the average value of historical water use monitoring data to obtain the mean water use value; Perform mean filling correction on the error samples based on the water mean to generate error correction data; Perform anomaly elimination processing based on real abnormal samples to obtain abnormal sample elimination data; According to the error law, local filtering optimization is performed on the error correction data and abnormal sample elimination data to generate optimized water use monitoring data; Step S2: Perform periodic water consumption characteristic analysis on the optimized water consumption monitoring data to generate water consumption characteristic data; perform quantitative mining on the time series water consumption distribution characteristics of the water consumption characteristic data to construct a time series water consumption characteristic distribution map; the specific steps of step S2 are: Step S21: performing periodic water usage frequency statistics on the optimized water usage monitoring data to generate a periodic water usage frequency; Step S22: performing multi-point water consumption peak identification on the optimized water consumption monitoring data to obtain multiple water consumption peak points; Step S23: performing periodic water consumption characteristic analysis on multiple water consumption peak points based on periodic water consumption frequency to generate water consumption characteristic data; Step S24: Quantitatively mine the time series water consumption distribution characteristics of the water consumption characteristic data, so as to obtain the time series water consumption distribution data within the period; Step S25: dynamically render the distribution graph of the time-series water consumption distribution data within the period to construct a time-series water consumption characteristic distribution graph; Step S3: Based on the time series water use characteristic distribution map, the dynamic water use behavior characteristics and the duration of each water use behavior are obtained; based on the duration of each water use behavior, the dynamic water use behavior characteristics are personalized and evolved at multiple time points to construct a personalized domestic water use behavior portrait; the specific steps of step S3 are: Step S31: Perform dynamic water use behavior analysis on the time series water use characteristic distribution diagram to obtain dynamic water use behavior characteristics; Step S32: based on the time series water use characteristic distribution diagram, start and stop calculation of water use behavior point by point is performed to obtain the duration of each water use behavior; Step S33: classifying and identifying the water use behavior pattern of the dynamic water use behavior characteristics based on the duration of each water use behavior, thereby generating water use behavior pattern characteristic data; Step S34: Perform personalized multi-time point evolution based on the water use behavior pattern characteristic data to construct a personalized domestic water use behavior portrait; Step S4: Acquire real-time smart water meter monitoring parameters; monitor the full-cycle water use fluctuation of the real-time smart water meter monitoring parameters through personalized domestic water use behavior portraits, and extract abnormal water use behavior data; the specific steps of step S4 are: Step S41: Acquire real-time smart water meter monitoring parameters; Step S42: Perform full-cycle water usage fluctuation monitoring on the real-time smart water meter monitoring parameters through personalized domestic water usage behavior portraits, thereby obtaining water usage fluctuation monitoring data; Step S43: Analyze the abnormal water use behavior of the water use fluctuation monitoring data and extract the abnormal water use behavior data; wherein the specific steps of step S42 are: Extract features of real-time smart water meter monitoring parameters to obtain water usage interval features and water consumption features; Convert the water usage interval characteristics and water consumption characteristics into time period water usage values ​​to generate water usage characteristic values ​​for each time period; Through personalized domestic water use behavior portraits, the water use characteristic values ​​of each period are compared and processed throughout the entire cycle to identify the water use difference periods; Conduct water use fluctuation deviation analysis based on water use difference periods to obtain water use fluctuation monitoring data; Step S5: Generate an abnormal result quantification value based on the abnormal water use behavior data; perform a safety status assessment on the abnormal result quantification value based on a preset safety assessment model to generate safety status risk prediction data; the specific steps of step S5 are: Step S51: numerically process the abnormal water use behavior data to generate a quantitative value of the abnormal result; Step S52: Perform safety status assessment on the quantified value of the abnormal result based on a preset safety assessment model to generate safety status assessment data for the elderly living alone; Step S53: Performing safety risk prediction on the safety status assessment data of the elderly living alone to generate safety status risk prediction data; Step S6: Make safety warning decisions based on the safety status risk prediction data and construct a safety status warning strategy for elderly people living alone; perform safety monitoring and warning operations for elderly people living alone based on the safety status warning strategy for elderly people living alone.

2. The safety monitoring and early warning method based on the water consumption characteristics of the smart water meter according to claim 1 is characterized in that: The specific steps of step S6 are: Step S61: Make safety warning decisions based on safety status risk prediction data and build a safety status warning strategy for elderly people living alone; Step S62: dynamically optimizing and modeling the personalized domestic water use behavior portrait based on the real-time smart water meter monitoring parameters to build a water use behavior safety monitoring model; Step S63: Execute safety monitoring and early warning operations for elderly people living alone based on the safety status early warning strategy for elderly people living alone and the water use behavior safety monitoring model.

3. A safety monitoring and early warning device based on the water consumption characteristics of a smart water meter, characterized in that: The method for executing the safety monitoring and early warning method based on the water consumption characteristics of the smart water meter according to claim 1 comprises: The extreme value filtering module is used to obtain the historical water consumption monitoring data of the elderly living alone based on the smart water meter; the historical water consumption monitoring data of the elderly living alone is subjected to extreme value filtering optimization processing to generate optimized water consumption monitoring data; The time series water consumption characteristic module is used to analyze the periodic water consumption characteristics of the optimized water consumption monitoring data to generate water consumption characteristic data; to quantitatively mine the time series water consumption distribution characteristics of the water consumption characteristic data to construct a time series water consumption characteristic distribution map; The water use behavior portrait module is used to obtain dynamic water use behavior characteristics and the duration of each water use behavior based on the time series water use characteristic distribution map; based on the duration of each water use behavior, the dynamic water use behavior characteristics are personalized and evolved at multiple time points to construct a personalized domestic water use behavior portrait; The water consumption fluctuation monitoring module is used to obtain real-time smart water meter monitoring parameters; the real-time smart water meter monitoring parameters are monitored for full-cycle water consumption fluctuations through personalized domestic water consumption behavior portraits, and abnormal water consumption behavior data is extracted; A safety status assessment module is used to generate abnormal result quantification values ​​based on abnormal water use behavior data; perform safety status assessment on the abnormal result quantification values ​​based on a preset safety assessment model to generate safety status risk prediction data; The safety warning decision-making module is used to make safety warning decisions based on safety status risk prediction data and to build a safety status warning strategy for elderly people living alone; based on the safety status warning strategy for elderly people living alone, the safety monitoring and warning operations for elderly people living alone are performed.

Images