Water consumption monitoring system based on cloud platform

Abstract

The invention relates to the technical field of cloud computing monitoring, in particular to a water consumption monitoring system based on a cloud platform, which comprises a data trend identification module, a water consumption transmission scheduling judgment module, a water consumption interval reconstruction module, a user behavior identification module and a scheduling feedback correction module. According to the method, by introducing a sequence matching strategy of a timestamp and a flow value, the recognition precision of continuous flow change is improved, terminal grouping and amplitude consistency screening are combined, the trend sequence stability is enhanced, short-time fluctuation interference is reduced, extreme value positioning and state segmentation processing are utilized, and multi-level sensing of water volume change is achieved; flow map features are enriched, the time sequence and behavior start-stop information are fused to identify the offset state, the response ability to water consumption mode changes is improved, a dynamic allocation mechanism is constructed through response codes and transmission features, channel resource allocation is optimized, and the data processing and resource response efficiency of the system in an abnormal state is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of cloud computing monitoring, and particularly to a water consumption monitoring system based on a cloud platform. Background Art

[0002] The technical field of cloud computing monitoring includes real-time detection and management of the operating status of distributed computing resources. The core content of this technical field lies in achieving continuous monitoring of various resources such as computing nodes, storage systems, network communications, and service loads through means such as resource scheduling, performance tracking, operating log analysis, and status perception. Cloud computing monitoring involves comprehensively controlling large-scale data centers and the virtual resources running thereon to ensure service stability and resource utilization. This field also includes summarizing, classifying, and visually managing monitoring data to assist system administrators in accurately judging the resource status and making corresponding configuration adjustments to maintain the normal operation of cloud services.

[0003] Among them, the water consumption monitoring system refers to a technical solution for data collection, transmission, and processing of water resource usage based on cloud computing monitoring technology. The technical matters targeted by this patent theme cover reading of water meter data, remote uploading of water volume data, periodic statistics of water usage status, and identification of abnormal usage behaviors. Specifically, electromagnetic sensing elements are used to obtain water flow velocity and water volume change information, an embedded processing chip is used to convert the collected signal into digital data, and a wireless communication module is used to send the data to a remote data processing platform. In the remote platform, the water usage situation is classified and marked according to data characteristics such as time period, water volume threshold, usage frequency, etc., and continuous recording and management are carried out according to set rules.

[0004] Although the prior art can read and upload water meter data, it lacks a refined recognition mechanism for extracting continuous change patterns, resulting in the judgment of flow trends being easily interfered by accidental fluctuations and unable to accurately capture regular characteristics. In the process of processing remote data, a static classification method based on thresholds is often used, ignoring the dynamic evolution trend of water usage behavior in the time sequence structure and making it difficult to reveal the delay or jump characteristics of user behavior, thus affecting the comprehensive recognition of abnormal behaviors. Since the uploaded data is not refined and grouped and the flow state is not evaluated, the system has a weak ability to recognize sudden data growth and may have problems of unbalanced resource allocation in high-concurrency scenarios. There is a lack of in-depth analysis of the response status during the transmission process, and a dynamic scheduling mechanism based on transmission integrity and response codes cannot be constructed, restricting the elastic processing ability of the system under conditions of tight bandwidth resources. For example, when multiple terminals concurrently upload a large amount of data, the system fails to timely identify high-priority data streams, resulting in delays or losses of key data, thus affecting the overall monitoring accuracy and feedback timeliness. The above deficiencies are likely to cause a slow response to behavior variations and a decrease in resource utilization efficiency in continuous water consumption monitoring scenarios. Summary of the Invention

[0005] The object of the present invention is to solve the disadvantages existing in the prior art, and a water consumption monitoring system based on a cloud platform is proposed.

[0006] To achieve the above object, the present invention adopts the following technical solutions: The water consumption monitoring system based on a cloud platform includes: The data trend recognition module filters data groups with consistent water flow value changes based on the timestamp item, flow value item, and terminal identification item in the terminal water consumption data stream, and makes a judgment in combination with the coherence of the timestamp item to generate a co-directional change sequence record; The water consumption transmission scheduling determination module compares the terminal bandwidth utilization rate value corresponding to the timestamp item with the current data packet size value based on the terminal identification item and timestamp item corresponding to the co-directional change sequence record, determines whether it is within the channel capacity threshold range, and generates a traffic scheduling sorting label; The water volume interval reconstruction module performs sampling grouping based on the flow value item and terminal identification item within the time period corresponding to the traffic scheduling sorting label, identifies the flow items in each group, marks the time period as a change status group, and extracts the flow value within the change status group to obtain a water consumption flow fluctuation distribution map; The user behavior recognition module calls the water consumption flow fluctuation distribution map, and based on the daily timestamp sequence and behavior start and end time items of the terminal identification item, identifies the change amplitude and continuous position of the behavior start and end time items, and determines whether there are phenomena of advance, delay, and jump segments in the behavior state chain, and generates a user water consumption behavior deviation trend monitoring data set.

[0007] As a further solution of the present invention, the co-directional change sequence record includes a flow change direction feature, a continuous time feature, and a fluctuation amplitude threshold, the traffic scheduling sorting label includes a priority factor, a channel capacity identifier, and a data packet adaptation coefficient, the water consumption flow fluctuation distribution map includes a flow fluctuation interval, an extreme point distribution, and an intra-group change feature, and the user water consumption behavior deviation trend monitoring data set includes a behavior start and end deviation value, a time period continuity index, and a behavior chain stability feature.

[0008] As a further solution of the present invention, the data trend recognition module includes: The terminal sequence reconstruction sub-module classifies the water consumption data stream based on the timestamp item, flow value item, and terminal identification item in the terminal water consumption data stream, extracts the corresponding timestamp item and combines the corresponding flow value item and timestamp item in order to obtain a terminal time sequence; The amplitude change extraction sub-module calls the terminal time sequence, identifies the flow change values in adjacent time periods, compares the change amplitudes, and filters pairs of flow values with consistent amplitudes to obtain an equal-amplitude change interval sequence; The same - direction sequence judgment sub - module determines whether it is continuously in the same direction according to the sign of the flow difference in the equal - amplitude change interval sequence, using the formula: ; Calculate the value of the continuous same - direction change degree, determine whether it is classified into the sequence, and obtain the record of the same - direction change sequence; Among them, represents the value of the continuous same - direction change degree, represents the flow change difference in the th time period, represents the time - stamp increment corresponding to the th time period, represents the flow value in the th time period, represents the flow value in the th time period, represents the total number of time periods.

[0009] As a further solution of the present invention, the water - use transmission scheduling determination module includes: The terminal data calling sub - module extracts the terminal data transmission records matching the time range based on the corresponding terminal identification item and time - stamp item in the same - direction change sequence record, identifies the water - use data packet size value and the real - time terminal bandwidth usage value corresponding to each time - stamp, and generates the terminal data synchronization result; The channel state evaluation sub - module calls the terminal data synchronization result, extracts the bandwidth usage value and the data packet size value, and identifies the unit - bandwidth load index according to the set channel capacity threshold, using the formula: ; Calculate the channel load intensity value, compare the load intensity value with the set threshold value range, determine whether it is within the channel capacity range, and obtain the channel load state classification value; Among them, represents the channel load intensity value, represents the bandwidth usage at the current time point, represents the current data packet size value, represents the absolute value of the bandwidth difference between the previous two time points within the time period, represents the cumulative size of data packet transmission in the recent three segments; The scheduling order analysis sub - module determines whether the current channel meets the bandwidth scheduling standard according to the channel load state classification value, filters the data records corresponding to the time - stamps that meet the conditions, and sorts them in ascending order according to the channel load state to obtain the flow scheduling sorting label.

[0010] As a further solution of the present invention, the water - volume interval reconstruction module includes: The section sampling grouping sub-module collects the traffic data of the terminals based on the traffic value items and terminal identification items within the corresponding time period of the traffic scheduling sorting tag, classifies the data by terminal, and arranges the data in chronological order to obtain a grouped traffic sampling structure; The status change recognition sub-module calls the traffic values of each group in the grouped traffic sampling structure, detects the extreme values of the traffic within each group, and if the difference exceeds the set threshold, marks the time period as a change status group to obtain a traffic status classification list; The traffic distribution calculation sub-module extracts the corresponding traffic values according to the time periods marked as change status in the traffic status classification list, sorts them, and uses the formula: ; Calculate the traffic fluctuation degree value within the interval, and map the section traffic fluctuation degree value to the column height to obtain a water consumption traffic fluctuation distribution map; Wherein, is the traffic fluctuation degree value within the interval, is the traffic value of the th traffic item, is the average value of the traffic values within the th traffic item group, is the same-terminal traffic variance value of the th traffic item, is the active level quantization value of the terminal to which the th traffic item belongs, is the scheduling sorting level value of the th traffic item, is the number of terminals of the th traffic item, represents the serial number index in the change status time period, represents the total number of traffic items in the change status group.

[0011] As a further solution of the present invention, the user behavior recognition module includes: The behavior time fluctuation extraction sub-module calls the water consumption traffic fluctuation distribution map, identifies the time position difference between the daily time stamp sequence and the behavior start and end time items, and extracts the fluctuation amplitude of the start and end times to obtain the start and end time fluctuation amplitude value; The behavior change position positioning sub-module calls the start and end time fluctuation amplitude value, and based on the time stamp sequence, identifies the start and end time change jump points and continuous missing points to generate a set of behavior start and end time change positions; The offset trend monitoring and calculation sub-module, according to the set of behavior start and end time change positions, identifies the change amplitude and continuous offset quantity of the behavior start and end time under the terminal, analyzes the distribution trend of continuous offset and mutation offset, and uses the formula: ; Calculate the behavior offset trend value under the computing terminal, sort the trend offset direction and intensity distribution, and generate a monitoring dataset for the user's water usage behavior offset trend; Among them, represents the position change value of the start time, represents the position change value of the end time, represents the number of jump points in the set of position changes of the behavior start and end times, represents the number of continuous jump segments, represents the number of consecutive days of records under the terminal in the timestamp sequence, represents the behavior offset trend value under the terminal.

[0012] As a further solution of the present invention, the system further includes a scheduling feedback correction module: The scheduling feedback correction module extracts the transmission duration value, integrity rate value, and response code item of the corresponding terminal in the previous cycle according to the monitoring dataset of the user's water usage behavior offset trend, determines whether there are consecutive abnormal response code items, performs segmented processing on the data transmission, divides the application data traffic item, and allocates the corresponding number of channels according to the current bandwidth usage rate to obtain a water usage parallel channel allocation list; The water usage parallel channel allocation list includes channel allocation parameters, data stream segment numbers, and bandwidth allocation factors.

[0013] As a further solution of the present invention, the scheduling feedback correction module includes: The abnormal response recognition sub-module extracts the transmission duration value, integrity rate value, and response code item according to the monitoring dataset of the user's water usage behavior offset trend, determines whether there are consecutive abnormal response code items, and obtains a continuous abnormal recognition coefficient; The data traffic segmentation sub-module calls the continuous abnormal recognition coefficient, performs segmented processing on the terminal transmission data, and statistically calculates the segmented traffic value and average interval according to the traffic density to obtain a segmented traffic distribution value; The parallel channel allocation sub-module matches the segmented traffic distribution value with the current bandwidth usage rate, extracts the bandwidth allocation trend of each segment, and uses the formula: ; Calculate the allocated channel value and map it to the segmented data to obtain a water usage parallel channel allocation list; Among them, represents the average transmission interval time of the segmented data, represents the segmented traffic value of the segmented data, represents the bandwidth usage rate of the segmented data, represents the remaining available number of the current channel, represents the cumulative value of the number of abnormal responses, is the allocated channel value.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In the present invention, by introducing a sequence matching strategy for timestamp items and flow rate value items in the terminal water usage data, the continuous flow rate change pattern can be effectively mined, thereby enhancing the recognition accuracy of the water flow trend. The grouping discrimination based on the terminal identifier combined with the flow rate amplitude consistency screening mechanism helps to construct a more stable change trend sequence, which can significantly reduce the interference of short-term fluctuations on the judgment accuracy during the actual recognition process. By performing the maximum and minimum value positioning operations on the flow rate data within a time period and introducing a state segmentation processing strategy, a multi-level perception of the dynamic water volume change process can be formed, so as to reflect richer periodicity and mutation characteristics in the flow rate distribution map. By means of the joint analysis method of the daily time series and the start and end information of the behavior, the accurate recognition of various offset states in the behavior chain can be realized, and the behavior offset type is used as the classification basis to improve the response sensitivity to the change of the water usage pattern. During the transmission and allocation process, relying on the monitoring results of the previous behavior offset state, a dynamic segmentation mechanism is constructed through the analysis of the response code and the transmission characteristics, which can realize the flexible allocation of the data channel resources, so that in the scenario of network bandwidth load fluctuation, the smooth transmission of high-priority data streams can still be guaranteed. By introducing time series comparison, segmented data matching and behavior anomaly classification strategies, a closed-loop processing flow covering data perception, feature extraction, behavior tracking and resource allocation is constructed, which significantly improves the processing ability of the water volume monitoring system for dynamic data changes and the resource response efficiency under abnormal states. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is the system flow chart of the present invention; Figure 2 is the flow chart of the data trend recognition module in the present invention; Figure 3 is the flow chart of the water usage transmission scheduling determination module in the present invention; Figure 4 is the flow chart of the water volume interval reconstruction module in the present invention; Figure 5 is the flow chart of the user behavior recognition module in the present invention; Figure 6 is the flow chart of the scheduling feedback correction module in the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. 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.

[0017] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by terms such as "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation to the present invention. In addition, in the description of the present invention, the meaning of "a plurality of" is two or more unless otherwise specifically defined.

[0018] Please refer to Figure 1 , the water consumption monitoring system based on the cloud platform includes: Based on the timestamp item, flow value item and terminal identification item in the terminal water use data stream, the data trend recognition module matches the continuous order of the timestamp item according to the terminal identification item, filters out the data groups with the same variation range of water flow values, and makes a judgment in combination with the coherence of the timestamp item to generate a co-directional change sequence record; Based on the terminal identification item and timestamp item corresponding to the co-directional change sequence record, the water use transmission scheduling determination module compares the terminal bandwidth utilization rate value corresponding to the timestamp item with the current data packet size value, judges whether it is within the channel capacity threshold range, and generates a flow scheduling sorting label; Based on the flow value item and terminal identification item within the time period corresponding to the flow scheduling sorting label, the water volume interval reconstruction module performs sampling grouping, identifies the maximum flow item and the minimum flow item in each group, marks the time period as a change status group, extracts the flow values within the change status group, and performs a sorting operation to obtain a water use flow fluctuation distribution map; The user behavior recognition module calls the water use flow fluctuation distribution map, and according to the daily timestamp sequence and behavior start and end time items of the terminal identification item, identifies the change amplitude and continuous position of the behavior start and end time items, and judges whether there are phenomena such as advance, delay and jump segments in the behavior status chain, and generates a user water use behavior deviation trend monitoring data set; Based on the user water use behavior deviation trend monitoring data set, the scheduling feedback correction module extracts the transmission duration value, completion rate value and response code item of the corresponding terminal in the previous cycle, judges whether there are continuous abnormal response code items, performs segmented processing on the data transmission, splits the application data flow item, and allocates the corresponding number of channels according to the current bandwidth utilization rate to obtain a water use parallel channel allocation list.

[0019] The co - varying sequence record includes the flow change direction feature, the continuous time feature, and the fluctuation amplitude threshold. The flow scheduling sorting tag includes the priority factor, the channel capacity identifier, and the data packet adaptation coefficient. The water flow fluctuation distribution map includes the flow fluctuation interval, the extreme point distribution, and the variation characteristics within the group. The user water usage behavior deviation trend monitoring data set includes the behavior start - end deviation value, the time - period continuity index, and the behavior chain stability feature. The water - using parallel channel allocation list includes the channel allocation parameter, the data stream segmentation number, and the bandwidth allocation factor.

[0020] Please refer to Figure 2 , the data trend recognition module includes: The terminal sequence reconstruction sub - module classifies the water - using data stream based on the timestamp item, the flow value item, and the terminal identifier item in the terminal water - using data stream, extracts the corresponding timestamp item, and combines the corresponding flow value item and timestamp item in order to obtain the terminal time - order sequence. Collect the water - using data of each terminal through sensors, including the timestamp item and the flow value item. The data is first grouped by the terminal identifier item, and all data of the same terminal is classified as one category. Each group of data is sorted according to the order of the timestamp item to ensure the time continuity of the data. This step is crucial for subsequent data analysis because only correctly sorted data can correctly reflect the change trend of water flow. For example, in practical applications, for a commercial building, the water usage of each floor can be monitored in real - time and recorded in chronological order to facilitate the subsequent analysis of the overall water - using pattern of the building. Re - combine the corresponding flow value item and timestamp item to obtain the terminal time - order sequence.

[0021] The amplitude change extraction sub - module calls the terminal time - order sequence, identifies the flow change values within adjacent time periods, compares the change amplitudes, and filters the flow value pairs with consistent amplitudes to obtain the equal - amplitude change interval sequence. Extract the flow value within each time period and calculate the flow change amount between two adjacent timestamps. The calculation process involves simple difference calculation, that is, subtracting the flow value of the previous time period from the flow value of the current time period. In this way, data with consistent flow change trends in different time periods can be filtered, which is very useful for identifying water usage anomalies or equipment leaks. For example, in the monitoring of a water plant, by continuously tracking the flow change, once a sudden increase or decrease in flow is detected within consecutive time periods, an automatic alarm can be issued to indicate a leak or equipment failure. Compare the change amplitudes and filter the flow value pairs with consistent amplitudes to obtain the equal - amplitude change interval sequence.

[0022] The co - direction sequence judgment sub - module judges whether it is co - directionally continuous according to the flow difference sign in the equal - amplitude change interval sequence, using the formula: ; Calculate the value of the continuous co-directional change degree, determine whether it belongs to the sequence, and obtain the record of the co-directional change sequence; Among them, represents the value of the continuous co-directional change degree, represents the flow change difference in the th time period, represents the time stamp increment corresponding to the th time period, represents the flow value in the th time period, represents the flow value in the th time period, represents the total number of time periods; First, call the obtained equal-amplitude change interval sequence, which is composed of multiple adjacent time stamp items and their corresponding flow value items. The system calculates the difference between each pair of adjacent flow values in the sequence and extracts its sign feature. Specifically, the flow value in the latter time period is subtracted from the flow value in the former time period. If the result is positive, it is recorded as +1; if it is negative, it is recorded as -1; and if it is zero, it is ignored. By comparing the signs of all differences item by item, it is judged whether the same direction is maintained in multiple consecutive time periods. When the number of consecutive identical signs in the sign sequence is not less than 3, it can be defined as a co-directional change segment. On this basis, it is further judged whether the change intensity in the continuous segment is within a reasonable interval range. The minimum continuous change degree value interval is set to 20 - 150 units. If the cumulative change degree in the continuous segment is less than 20, it is excluded; Taking a section of actual water use monitoring data as an example: The time stamp sequence is: 9:00, 9:05, 9:10, 9:15, 9:20, with the unit of minutes, and the flow value sequence is: 110, 125, 140, 155, 170, with the unit of L. First, calculate the flow difference in each section: = 125 - 110 = 15, = 140 - 125 = 15, = 155 - 140 = 15, = 170 - 155 = 15; It can be obtained that the time increment in each section minutes, the difference between the values is consistent with , so the calculation of each section is as follows: The first item: ; The second item: ; The third item: Similarly, it is 18.75; The fourth item: Similarly, it is 18.75; Substitute into the formula: ; According to the preset judgment conditions, the S value of this continuous segment is 75, which is within the set reasonable continuous change degree value range of 20-150. It is judged that its change direction is continuous and the intensity is sufficient. Finally, it is classified into the same-direction change sequence to obtain the same-direction change sequence record. The formula calculation value S represents the continuous same-direction change degree value of this equal-amplitude segment. By introducing the normalization processing of the flow change difference, time increment, and difference denominator, the stable characterization and normalization evaluation of the same-direction change sequence in terms of trend intensity are realized. The result shows that this data sequence segment is a typical water use trend segment with consistent direction, amplitude, and time interval. The numerical result S = 75, which belongs to the judgment interval, and it is confirmed that it belongs to the same-direction change sequence record.

[0023] Please refer to Figure 3 , the water transfer and scheduling judgment module includes: The terminal data call sub-module extracts the terminal data transmission records matching the time range based on the corresponding terminal identification item and time stamp item in the same-direction change sequence record, identifies the water use data packet size value and real-time terminal bandwidth usage value corresponding to each time stamp, and generates the terminal data synchronization result; Extracting the terminal data transmission records matching the time stamp involves screening the data packets matching the specific terminal identification item and time stamp item from a large amount of data. For example, assuming that within a specific time period, it is monitored that the water use of a terminal increases abnormally, all relevant data transmission records within the time period are automatically identified, including flow data and time stamps, and the data packet size and bandwidth usage in the records are compared, which is completed through the analysis of the data transmission log. For example, in a specific instance, the total data packet size of a terminal between 9:00 and 10:00 is 500MB, and the bandwidth usage is 2GB. According to the comparison result, data synchronization is performed, the information is integrated, and the terminal data synchronization result is generated, providing an accurate input data source for the subsequent module and ensuring the real-time and accuracy of the data.

[0024] The channel status evaluation sub-module calls the terminal data synchronization result, extracts the bandwidth usage value and data packet size value, and according to the set channel capacity threshold, identifies the unit bandwidth load index, using the formula: ; Calculate the channel load intensity value, compare the load intensity value with the set threshold value range, judge whether it is within the channel capacity range, and obtain the channel load status classification value; Among them, represents the channel load intensity value, represents the bandwidth usage at the current time point, represents the current data packet size value, represents the absolute value of the bandwidth difference between the previous two time points within the time period, Represents the cumulative size of data packet transmission in the recent three segments; First, process the acquisition of the bandwidth usage value, which is sourced from the real-time bandwidth monitoring log of the cloud platform for the terminal access device, with the unit of Mbps. For example, at a certain time point, the collected bandwidth usage value is 92 Mbps. Meanwhile, extract the corresponding data packet size value at this time point, which is sourced from the terminal upload data record log, with the unit of MB, and this value is set to 340 MB. Next, it is necessary to calculate the change value of the bandwidth usage in this time period, that is, select the absolute value of the difference between the bandwidth usage at the current time point and the previous time point. Assume the previous time point is 80 Mbps, then the change value is |92 - 80| = 12 Mbps. For unit consistency processing, all units need to be converted to a unified dimension. The bandwidth value of 92 Mbps can be converted to the transmission capacity per second: 92 Mbps = 11.5 MB / s, 12 Mbps = 1.5 MB / s. At this time, the bandwidth change value L = 1.5; Meanwhile, select the cumulative size value of the data packets within the recent three time periods. Assume the values of the previous two periods are 310 MB and 290 MB respectively, then the three-segment accumulation is 340 + 310 + 290 = 940 MB. Since there is an amplification factor in the square root calculation, to avoid affecting the ratio, it is necessary to perform normalization processing. All data is unified in units before input, where L is in MB / s, B is in MB / s, D is in MB, and U is in MB; Substitute the above values into the formula for calculation: ; ; Obtain the channel load intensity value of 127.5. Next, compare it with the set channel capacity threshold value. This threshold value is set as a range threshold, with a reasonable upper limit of 160 and a lower limit of 80. When the load intensity value is within this range, it is determined as "adapted", otherwise as "abnormal". The current value of 127.5 is between the upper and lower thresholds, and it is judged that the current terminal is within the capacity range, obtaining the channel load status classification value for subsequent scheduling and sorting processing; Among them, represents the channel load intensity value, represents the terminal bandwidth usage under the unified dimension, represents the data packet size at this time point, represents the absolute difference between the current bandwidth usage value and the previous bandwidth usage value, is the cumulative transmission volume of data packets at the current time point and the previous two time points. Adding 1 in the denominator is a constant adjustment term to avoid numerical fluctuations when the denominator approaches zero. The benefit of this formula is that by introducing the combined structure of the bandwidth change term L and the data transmission product term B·D, and at the same time performing a square root normalization process on the historical cumulative data term U, it realizes the comprehensive evaluation of the instantaneous bandwidth pressure and the historical data transmission volume, effectively depicts the instantaneous transmission intensity and dynamically balances the difference in transmission trends. The result shows that the current channel is in a load adaptation state and can enter the scheduling priority evaluation process.

[0025] The scheduling order analysis sub-module determines whether the current channel meets the bandwidth scheduling standard according to the channel load status classification value, filters the data records corresponding to the eligible timestamps, and sorts them in ascending order according to the channel load status to obtain the traffic scheduling sorting label. Determine whether the current channel meets the bandwidth scheduling standard. This process involves sorting all eligible terminal data according to the channel load status. For example, if the channel load status classification value at a certain time point is lower than the set threshold, it indicates that the bandwidth usage of the terminal during this period is in an ideal state and is suitable for data transmission scheduling. Sort the information and give priority to processing terminals with low bandwidth usage rates to optimize the overall network data stream, ensure the effective utilization of network bandwidth, and dynamically adjust the data transmission priority according to the actual bandwidth usage situation to generate a traffic scheduling sorting label. The label will be used for network traffic management to ensure the effective and priority transmission of data packets. The final result is to improve the overall network efficiency and reduce data transmission latency.

[0026] Please refer to Figure 4 , the water volume interval reconstruction module includes: The section sampling grouping sub-module collects the traffic data of the terminals based on the traffic value items and terminal identification items within the time period corresponding to the traffic scheduling sorting label, classifies them by terminal, and arranges the data in chronological order to obtain the grouped traffic sampling structure. For the time period indicated by the traffic scheduling and sorting tags, automatically extract the traffic data within the corresponding time from the cloud database. The data includes the traffic values of each terminal within the specified time period. For example, from 9:00 to 9:30, for terminal A, the recorded traffic data is 100L, 120L, 130L, etc. The steps involve not only data extraction but also a comprehensive data processing process of classifying by terminal identifier and sorting by time. Group the data into units of every 30 minutes, ensuring that each group contains data of at least three time points to analyze the traffic change trend. This process is completed through data mining techniques, ensuring the timeliness and integrity of the data. Sort the data within each group in the order of traffic values over time, which facilitates subsequent change state analysis and generates a grouped traffic sampling structure. This structure is crucial for understanding the water usage behavior patterns of each terminal at different time periods and provides basic data for water volume management.

[0027] The state change recognition sub-module calls the traffic values of each group in the grouped traffic sampling structure, detects the extreme values of the traffic within each group. If the difference exceeds the set threshold, mark the time period as a change state group to obtain a traffic state classification list. Identify significant traffic changes from the sampled traffic data. This process starts with the analysis of each group of data. First, calculate the maximum traffic value and the minimum traffic value of each group. For example, if the traffic values recorded for a certain terminal from 9:00 to 9:30 are 100L, 120L, 130L, then the maximum value is 130L and the minimum value is 100L. Calculate the difference between them. If the difference value exceeds the preset threshold of 20L, it is considered that the traffic change within the time period is significant. Such analysis helps to accurately identify water usage peaks or abnormal water usage behaviors. Mark all time periods that meet the significant change conditions as change state groups. The process is automated to ensure the accuracy and efficiency of the operation and establish a traffic state classification list. This list details the traffic change states within each time period and is an effective characterization of the terminal water usage behavior.

[0028] The traffic distribution calculation sub-module extracts the corresponding traffic values for sorting according to the time periods marked as change states in the traffic state classification list, using the formula: ; Calculate the traffic fluctuation degree value within the interval and map the section traffic fluctuation degree value to the column height to obtain a water usage traffic fluctuation distribution map. Among them, is the traffic fluctuation degree value within the interval, is the traffic value of the th traffic item, is the average value of the traffic values within the th traffic item group, is the variance value of the same-terminal traffic for the th traffic item, is the quantization value of the activity level of the terminal to which the th traffic item belongs, is the scheduling and sorting level value of the th traffic item, is the number of terminals of the th traffic item, represents the sequence index in the change status time period, represents the total number of traffic items in the change status group; All traffic data marked as in the change status are subjected to detailed analysis and sorting processing. In order to deeply analyze the volatility of the traffic, the following complex mathematical model is adopted, and the model analysis and parameter concretization are as follows: : refers to the traffic value at a certain time point. For example, the traffic measurement within a certain time period is 120L, 130L, 140L; : is the average value of the traffic within this time period, which is for the above data; : is the variance of the traffic within this time period, and the calculation formula is ; : is the quantization value of the activity level of the terminal to which this time period belongs. Here, it is assumed to be 2 (this value is obtained by analyzing the terminal activity log); : is the scheduling and sorting level value within this time period. For example, it is 1, which is determined by the sorting logic of the traffic scheduling system; : is the number of terminals within this time period. It is assumed that there are 3 active terminals in this time period; All traffic values ( , ) are in liters (L). No further conversion is required here. After the variance calculation is completed, the result is expressed in liters squared ( ). The square root of the variance plus 1 is used for stable numerical processing to avoid the situation of division by zero. The activity level and scheduling level are dimensionless scores or levels; Substitute specific values for calculation to obtain the contribution degree of each item: For , traffic : ; For , traffic : ; For , traffic : ; Sum up the above calculation results to obtain the overall flow fluctuation value. : ; This calculation process provides a quantified value of flow fluctuation. Finally, the value is converted into a graphical display, such as being shown by a bar chart, to generate a water consumption flow fluctuation distribution map. This map effectively shows the water consumption fluctuations in different time periods, providing a powerful analysis tool for water resource management and scheduling.

[0029] Please refer to Figure 5 , the user behavior recognition module includes: The behavior time fluctuation extraction sub-module calls the water consumption flow fluctuation distribution map, identifies the time position difference between the daily timestamp sequence and the start and end time items of the behavior, extracts the fluctuation amplitude of the start and end times, and obtains the start and end time fluctuation amplitude value; Identify the matching points between the daily timestamp sequence and the start and end time items of the behavior through a specific algorithm. For example, in practical applications, the water consumption per hour can be determined by the water meter reading, and then the data is matched with the user's activity records to find the start and end times of the peak water consumption period and daily activities, and calculate the set of position difference values of the time points. For example, in a day, the user's morning water consumption peak starts at 7 o'clock and ends at 7:30, and the position difference of this time period is 30 minutes. According to the set of position difference values, the set of fluctuation amplitudes in the daily sequence can be further calculated. This fluctuation amplitude shows the instability of the user's water consumption behavior in time, and finally generates the start and end time fluctuation amplitude value. Such calculation reflects the influence of the user's living habits on the water consumption time and can better understand the user's water consumption behavior pattern.

[0030] The behavior change position positioning sub-module calls the start and end time fluctuation amplitude value, and based on the timestamp sequence, identifies the sudden jump points and continuous missing points of the start and end time changes, and generates a set of behavior start and end time change positions; Compare and calculate according to the start and end time items of the behavior in the daily timestamp sequence, analyze the distribution of the start and end time of the behavior on the time axis, and then identify the continuous missing positions and the sudden jump points of the start and end time changes. For example, if the user's water consumption behavior suddenly changes from 7 o'clock in the morning to 6:30 in the morning one day, this sudden jump will be recorded. The sudden jump points and continuous missing points are grouped, and then the position differences are calculated. This can be achieved by comparing the data of several consecutive days, identifying the patterns and anomalies of time changes, and finally generating a set of behavior start and end time change positions. The set can help the water supply company optimize the water supply plan and reduce the water supply pressure caused by sudden changes in user behavior.

[0031] The offset trend monitoring calculation sub-module identifies the change amplitude and continuous offset quantity of the start and end times of the behavior under the terminal based on the set of changed positions of the start and end times of the behavior, analyzes the distribution trends of continuous offsets and mutation offsets, and uses the formula: ; Calculate the behavior offset trend value under the terminal, sort the trend offset direction and intensity distribution, and generate a monitoring data set for the user's water use behavior offset trend; Among them, represents the position change value of the start time, represents the position change value of the end time, represents the number of jump points in the set of changed positions of the start and end times of the behavior, represents the number of continuous jump segments, represents the number of consecutive days of continuous records under the terminal in the time stamp sequence, represents the behavior offset trend value under the terminal; Call the daily time stamp sequence under each terminal identifier, count the time position change values of its start time and end time within the number of consecutive record days, and set to represent the change value of the start time, to represent the change value of the end time. Both are quantified in minutes, and the time change amount is obtained by subtracting the start and end time sequences of the same user behavior for consecutive days. For example, in terminal A, the start times of the user's water use behavior in five consecutive days are 7:10, 7:15, 7:20, 7:30, 7:35 in sequence, and the corresponding adjacent time differences are 5, 5, 10, 5 minutes. The average change value is 6.25 minutes. The end times are 7:40, 7:45, 7:50, 8:00, 8:05 in sequence, then is 6.25 minutes; Subsequently, detect the number of jump points η and the number of continuous jump segments γ in the set of changed positions of the start and end times of the behavior. The definition of a jump point is a time point where the start or end time change between a certain day and the previous day exceeds 20 minutes. Suppose the start time on the sixth day in the above sequence becomes 6:30, the change is 65 minutes, and it is recorded as 1 jump point, then η = 1; Continuous jump segments refer to jump behaviors for two or more consecutive days. For example, if the seventh day continues to be 6:25, it is defined as a group of continuous jump segments, γ = 1; The number of consecutive record days ξ is calculated as 7 days according to the time stamp sequence record. To unify the dimension, , , η, and γ are all transformed into minute values or standardized coefficients for processing, and the normalization method is used to process each variable. For example The maximum change value of is 60 and the minimum is 0, then =(6.25 - 0) / 60 = 0.104. After variable normalization, they are respectively = 0.104, η = 0.05 (if the maximum jump points are 20), γ = 0.1 (the maximum is 10), and ξ is directly substituted with 7; Substitute the values: ; The result shows that the user behavior deviation trend value of terminal A is 0.0140, indicating a weak deviation trend in the current period. It is positively correlated with the degree of behavior mutation and continuous change, and is applicable to subsequent hierarchical screening and classification modeling in the deviation trend distribution. Further, a user water consumption behavior deviation trend monitoring data set is obtained. By synthesizing and calculating the change values of the behavior period and the sudden and continuous change signals, a quantifiable index of the water consumption behavior stability trend is established, which is convenient for subsequent systematic classification and monitoring.

[0032] Please refer to Figure 6 , the scheduling feedback correction module includes: The abnormal response recognition sub-module extracts the transmission duration value, integrity rate value, and response code item according to the user water consumption behavior deviation trend monitoring data set, judges whether there are continuous abnormal response code items, and obtains the continuous abnormal recognition coefficient; First, monitor the data upload frequency and integrity of each terminal. For example, in an actual water service management, if the data upload frequency of a sensor significantly decreases or the data integrity greatly reduces within several consecutive hours, mark the time point as abnormal. Then, check the response codes near the abnormal point. If the continuously received response codes are errors or failures, calculate the abnormal frequency within this period. By comparing the frequency with the historical data of the same period, judge whether there is really a device failure or data transmission problem. The refined process ensures that each detection point is accurately analyzed, and each abnormality is recorded and evaluated, obtaining the continuous abnormal recognition coefficient.

[0033] The data flow segmentation sub-module calls the continuous abnormal recognition coefficient, segments the terminal transmission data, and statistically calculates the segmented flow value and average interval according to the flow density, obtaining the segmented flow distribution value; Statistically calculate the segmented flow value and average interval according to the flow density. For example, in urban water supply, the data volume transmitted by terminal devices such as water meters will vary greatly between peak and off-peak hours. By real-time monitoring the data flow, according to the set flow density threshold, automatically distinguish the high-flow and low-flow periods of the data flow, record and analyze the data flow of each period, calculate the data transmission characteristics of each segment, including the size of each data packet and the transmission interval time on average. In this way, the water usage pattern and changes can be understood more carefully, providing a basis for subsequent data management and analysis, and obtaining the segmented flow distribution value.

[0034] The parallel channel allocation sub-module matches according to the segmented traffic distribution value and the current bandwidth utilization rate, extracts the bandwidth allocation trend for each segment, and uses the formula: ; Calculate the allocated channel value and map it to the segmented data to obtain the water use parallel channel allocation list; Among them, represents the average transmission interval time of the segmented data, represents the segmented traffic value of the segmented data, represents the bandwidth utilization rate of the segmented data, represents the remaining available quantity of the current channel, represents the cumulative value of the abnormal response times, is the allocated channel value; Extract the bandwidth allocation trend for each segment and calculate the number of channels. If the data traffic suddenly increases within a certain period of time, first check the bandwidth utilization rate of the current network. For example, if the bandwidth utilization rate is 70% and the data traffic in this period has increased by 30% compared with the previous period, then adjust the number of channels allocated to this period according to the following formula to ensure that the data can be uploaded smoothly. Adjust the priority of channel allocation according to the continuously abnormal recognition coefficient calculated above. The period with extremely high abnormality will be given priority to ensure channel resources to prevent key data from being lost due to bandwidth problems; represents the average transmission interval time of the segmented data, assumed to be 200 seconds; represents the segmented traffic value of this segment of data, assumed to be 500 MB; represents the bandwidth utilization rate of this segment of data, and the normalized value here is 0.7; represents the remaining available quantity of the current channel, set to 100; represents the cumulative value of the abnormal response times, set to 10 times; Adjust each parameter to keep the dimension consistent: (MB) needs to be converted to GB, so GB, (seconds) needs to be converted to minutes, so minutes; Now, all parameters in the formula have been unified with appropriate dimensions: ; The calculation result indicates that for a specific period, approximately 0.036 channels need to be added. The value is a theoretical value, and in actual operation, an integer needs to be taken, that is, at least 1 additional channel needs to be allocated to cope with the increased data traffic during this period. This calculation method ensures the scientificity and accuracy of channel allocation, avoiding data loss and transmission delay.

[0035] The above is only a preferred embodiment of the present invention and does not limit the present invention in other forms. Any person skilled in the art may use the technical content disclosed above to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A water consumption monitoring system based on a cloud platform, characterized in that, The system includes: Based on the timestamp item, flow value item and terminal identification item in the terminal water usage data stream, the data trend recognition module filters data groups with consistent water flow value change amplitudes, combines the coherence of the timestamp item for judgment, and generates a co-directional change sequence record; Based on the terminal identification item and timestamp item corresponding to the co-directional change sequence record, the water usage transmission scheduling determination module compares the terminal bandwidth usage value corresponding to the timestamp item with the current data packet size value, determines whether it is within the channel capacity threshold range, and generates a flow scheduling sorting label; Based on the flow value item and terminal identification item within the time period corresponding to the flow scheduling sorting label, the water volume interval reconstruction module performs sampling grouping, identifies the flow items in each group, marks the time period as a change status group, extracts the flow value within the change status group, and obtains a water usage flow fluctuation distribution map; The user behavior recognition module calls the water usage flow fluctuation distribution map, and based on the daily timestamp sequence of the terminal identification item and the behavior start and end time items, identifies the change amplitude and continuous position of the behavior start and end time items, and determines whether there are phenomena of advance, delay and jump segments in the behavior status chain, and generates a user water usage behavior deviation trend monitoring data set.

2. The water consumption monitoring system based on the cloud platform according to claim 1, wherein, The co-directional change sequence record includes a flow change direction feature, a continuous time feature, and a fluctuation amplitude threshold. The flow scheduling sorting label includes a priority factor, a channel capacity identifier, and a data packet adaptation coefficient. The water usage flow fluctuation distribution map includes a flow fluctuation interval, an extreme point distribution, and an intra-group change feature. The user water usage behavior deviation trend monitoring data set includes a behavior start and end deviation value, a time period continuity index, and a behavior chain stability feature.

3. The water consumption monitoring system based on a cloud platform according to claim 1, characterized in that, The data trend recognition module includes: Based on the timestamp item, flow value item and terminal identification item in the terminal water usage data stream, the terminal sequence reconstruction sub-module classifies the water usage data stream, extracts the corresponding timestamp item and combines the corresponding flow value item and timestamp item in order to obtain a terminal time sequence; The amplitude change extraction sub-module calls the terminal time sequence, identifies the flow change values in adjacent time periods, compares the change amplitudes, filters the flow value pairs with consistent amplitudes, and obtains an equal-amplitude change interval sequence; According to the flow difference sign in the equal-amplitude change interval sequence, the co-directional sequence judgment sub-module determines whether it is co-directionally continuous, and uses the formula: ； Calculates the continuous co-directional change degree value, determines whether to classify it into the sequence, and obtains a co-directional change sequence record; Among them, represents the value of the degree of continuous co-directional change, represents the flow change difference in the th time period, represents the time stamp increment corresponding to the th time period, represents the flow value in the th time period, represents the flow value in the th time period, represents the total number of time periods.

4. The water consumption monitoring system based on a cloud platform according to claim 3, wherein, The water usage transmission scheduling determination module includes: Based on the terminal identification item and timestamp item corresponding to the co-directional change sequence record, the terminal data call sub-module extracts the terminal data transmission record matching the time range, identifies the water usage data packet size value and the real-time terminal bandwidth usage value corresponding to each timestamp, and generates a terminal data synchronization result; The channel state evaluation sub-module calls the terminal data synchronization result, extracts the bandwidth usage value and the data packet size value, and according to the set channel capacity threshold, identifies the unit bandwidth load index, and uses the formula: ； Calculates the channel load intensity value, compares the load intensity value with the set threshold value range, determines whether it is within the channel capacity range, and obtains the channel load state classification value; Among them, represents the channel load intensity value, represents the bandwidth usage at the current time point, represents the current data packet size value, represents the absolute value of the bandwidth difference between the previous two time points within a time period, represents the cumulative size of data packet transmissions within the last three segments; The scheduling order analysis sub-module determines whether the current channel meets the bandwidth scheduling criteria according to the channel load status classification value, filters the data records corresponding to the timestamps that meet the conditions, and sorts them in ascending order according to the channel load status to obtain the traffic scheduling sorting label.

5. The water consumption monitoring system based on the cloud platform according to claim 4, characterized in that, The water volume interval reconstruction module includes: The section sampling grouping sub-module collects the traffic data of the terminal based on the traffic value items and terminal identification items within the time period corresponding to the traffic scheduling sorting label, classifies them by terminal, and arranges the data in chronological order to obtain the grouped traffic sampling structure; The state change identification sub-module calls each group of traffic values in the grouped traffic sampling structure, detects the extreme values of the traffic within each group, and if the difference exceeds the set threshold, marks the time period as a change status group to obtain the traffic state classification list; The traffic distribution calculation sub-module extracts the corresponding traffic values for sorting according to the time periods marked as change status in the traffic state classification list, and uses the formula: ； Calculates the traffic fluctuation degree value within the interval, and maps the section traffic fluctuation degree value to the column height to obtain the water consumption traffic fluctuation distribution map; Among them, is the traffic fluctuation degree value within the interval, is the traffic value of the th traffic item, is the average value of the traffic values within the th traffic item group, is the same-terminal traffic variance value of the th traffic item, is the active level quantization value of the terminal to which the th traffic item belongs, is the scheduling sorting level value of the th traffic item, is the number of terminals of the th traffic item, represents the serial number index in the time period of the change state, represents the total number of traffic items in the change state group.

6. The water consumption monitoring system based on a cloud platform according to claim 5, wherein The user behavior recognition module includes: The behavior time fluctuation extraction sub-module calls the water consumption traffic fluctuation distribution map, identifies the time position difference between the daily timestamp sequence and the behavior start and end time items, and extracts the fluctuation amplitude of the start and end times to obtain the start and end time fluctuation amplitude value; The behavior change position positioning sub-module calls the start and end time fluctuation amplitude value, and based on the timestamp sequence, identifies the sudden jump points and continuous missing points of the start and end time changes to generate the behavior start and end time change position set; The offset trend monitoring calculation sub-module identifies the change amplitude and continuous offset quantity of the behavior start and end time under the terminal according to the behavior start and end time change position set, analyzes the distribution trend of continuous offset and mutation offset, and uses the formula: ； Calculates the behavior offset trend value under the terminal, sorts the trend offset direction and intensity distribution, and generates the user water consumption behavior offset trend monitoring data set; Among them, represents the position change value of the start time, represents the position change value of the end time, represents the number of jump points in the set of position changes of the start and end times of the behavior, represents the number of consecutive jump segments, represents the number of consecutive days recorded under the terminal in the timestamp sequence, represents the behavior offset trend value under the terminal.

7. The water consumption monitoring system based on a cloud platform according to claim 1, characterized in that The system also includes a scheduling feedback correction module: The scheduling feedback correction module extracts the transmission duration value, integrity rate value, and response code item of the corresponding terminal in the previous cycle according to the user water consumption behavior offset trend monitoring data set, determines whether there are continuous abnormal response code items, and performs segmented processing on the data transmission, divides the application data traffic item, and allocates the corresponding number of channels according to the current bandwidth usage rate to obtain the water consumption parallel channel allocation list; The water consumption parallel channel allocation list includes channel allocation parameters, data stream segment numbers, and bandwidth allocation factors.

8. The water consumption monitoring system based on a cloud platform according to claim 7, wherein The scheduling feedback correction module includes: The abnormal response identification sub-module extracts the transmission duration value, integrity rate value, and response code item according to the user water consumption behavior offset trend monitoring data set, determines whether there are continuous abnormal response code items, and obtains the continuous abnormal identification coefficient; The data traffic segmentation sub-module calls the continuous abnormal identification coefficient, performs segmented processing on the terminal transmission data, and statistically calculates the segmented traffic value and average interval according to the traffic density to obtain the segmented traffic distribution value; The parallel channel allocation sub-module matches the segmented traffic distribution value with the current bandwidth usage rate, extracts the bandwidth allocation trend for each segment, and uses the formula: ； Calculate the allocation channel value and map it to the segmented data to obtain the water use parallel channel allocation list; Among them, represents the average transmission interval time of segmented data, represents the split traffic value of segmented data, represents the bandwidth utilization rate of segmented data, represents the remaining available quantity of the current channel, represents the cumulative value of the number of abnormal responses, is the value of the allocated channel.

Images