Water dispenser state monitoring method and system based on data analysis

Abstract

The application discloses a water dispenser state monitoring method and system based on data analysis, and belongs to the technical field of equipment state monitoring and management, which comprises the following steps: acquiring real-time water quality parameter values, equipment operation parameter values and user water use behavior data of a water dispenser; identifying a current water use situation mode according to the user water use behavior data, and extracting corresponding dynamic water quality baseline ranges and equipment operation baseline ranges from a preset baseline library; comparing the real-time water quality parameter values and the equipment operation parameter values with the dynamic water quality baseline ranges and the equipment operation baseline ranges, and generating a parameter deviation feature set; verifying the correlation of the parameter deviation feature set based on a preset dynamic correlation rule library, and generating a comprehensive evaluation parameter; determining a risk level identifier according to the comprehensive evaluation parameter, and triggering a hierarchical response instruction. The application adopts a situation-aware dynamic baseline modeling and multi-source data cross-verification mechanism, and can realize quantitative evaluation and maintenance decision support of a water dispenser operation risk.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of equipment state monitoring and management, in particular to a water dispenser state monitoring method and system based on data analysis. BACKGROUND

[0002] As a basic equipment in public places, the water quality safety and operation stability of the water dispenser directly affect the user experience. Current monitoring technology mainly relies on sensors to collect water quality and equipment parameters, and makes state judgments through preset rules.

[0003] The existing technology generally adopts a fixed threshold alarm mechanism, for example, setting an upper limit of total dissolved solids to trigger a water quality abnormality alarm, or judging equipment failure based on a heating time threshold. Some schemes introduce simple time series analysis, such as calculating the moving average of water quality parameters.

[0004] Traditional technical means have obvious limitations. Static threshold cannot adapt to the difference in water use behavior at different times, resulting in a sharp increase in false positive rate during peak water use period. Single parameter analysis breaks the correlation between water quality changes and equipment operation, making it difficult to distinguish between real faults and normal fluctuations. The response mechanism lacks a grading strategy and cannot take differentiated measures for different risk levels. SUMMARY

[0005] To solve the above problems, the present application provides a water dispenser state dynamic monitoring method and system based on multi-dimensional behavior and water quality correlation analysis, which adopts a context-aware dynamic baseline modeling and multi-source data cross-validation mechanism to realize quantitative evaluation of water dispenser operation risk and maintenance decision support.

[0006] The above object can be achieved by the following scheme:

[0007] The water dispenser state monitoring method based on data analysis comprises: acquiring real-time water quality parameter values, equipment operation parameter values and user water use behavior data of a water dispenser, and generating a multi-source monitoring data set; performing pattern recognition on the user water use behavior data in the multi-source monitoring data set to obtain a current water use context pattern, and extracting a dynamic water quality baseline range and an equipment operation baseline range corresponding to the current water use context pattern in a preset context baseline library to generate a contextualized baseline parameter set; calculating the deviation of the real-time water quality parameter values and the equipment operation parameter values from the dynamic water quality baseline range and the equipment operation baseline range in the contextualized baseline parameter set to generate a parameter deviation feature set; verifying the correlation of the parameter deviation feature set based on a preset dynamic correlation rule library to generate a comprehensive evaluation parameter; mapping the risk level according to the comprehensive evaluation parameter to obtain a risk level identifier, and triggering a graded response instruction matched with the risk level identifier.

[0008] Optionally, the acquiring the user water consumption behavior data comprises: collecting water consumption time period distribution characteristics, single water consumption amount distribution characteristics and water consumption interval characteristics within a preset time window; and performing pattern recognition on the water consumption time period distribution characteristics, the single water consumption amount distribution characteristics and the water consumption interval characteristics to generate the user water consumption behavior data.

[0009] Optionally, the method further comprises: acquiring a multi-dimensional monitoring data set in a historical period, the multi-dimensional monitoring data set comprising historical water quality parameter values, historical equipment operation parameter values and historical user water consumption behavior data; dividing typical situation mode categories according to the historical user water consumption behavior data; calculating water quality parameter fluctuation threshold intervals and equipment operation parameter fluctuation threshold intervals under each typical situation mode category to form dynamic water quality baseline ranges and equipment operation baseline ranges, and constructing a situation baseline library; and analyzing the correlation between water quality parameter change trends and equipment operation states based on the dynamic water quality baseline ranges and the equipment operation baseline ranges to generate a dynamic correlation rule library.

[0010] Optionally, the correlation verification based on the preset dynamic correlation rule library comprises: extracting a correlation verification rule matched with a current water consumption situation mode from the dynamic correlation rule library; inputting the parameter deviation characteristic set into the correlation verification rule for logical verification to output an abnormality credibility score and a potential root cause type identifier; and combining the abnormality credibility score, the potential root cause type identifier and the parameter deviation characteristic set to generate the comprehensive evaluation parameter.

[0011] Optionally, the correlation verification rule comprises: a low water consumption situation rule, a peak water consumption situation rule and a cross verification rule; wherein the low water consumption situation rule is to determine normal concentration characteristics when a long standby period is detected and water quality parameter values show a slow rising trend; the peak water consumption situation rule is to trigger a device fault suspicion identifier when a concentrated water consumption period is detected and water quality parameter values abnormally rise; and the cross verification rule is to improve an abnormality credibility score level when water quality parameter deviation characteristics and equipment operation abnormality characteristics exist simultaneously.

[0012] Optionally, the trigger hierarchical response instruction comprises: generating an event log record containing deviation characteristics when a risk level identifier is an observation level; generating a water quality abnormality report and sending a detection reminder notification when the risk level identifier is a detection suggestion level; generating a device maintenance suggestion and activating a preset maintenance work order interface when the risk level identifier is a maintenance warning level; and performing an outlet locking operation and pushing emergency alarm information containing a comprehensive evaluation parameter when the risk level identifier is an emergency alarm level.

[0013] Optionally, the method further comprises: recording the triggered hierarchical response instruction and subsequent processing feedback data; and correcting the dynamic water quality baseline range, the equipment operation baseline range, and the dynamic correlation rule base according to the subsequent processing feedback data.

[0014] Optionally, the generating of the water quality anomaly report comprises: correlating dynamic water quality baseline historical data in the current water use context mode; comparing deviation amplitudes of the real-time water quality parameter value and the dynamic water quality baseline historical data; and generating a readability report containing context correlation analysis based on the deviation amplitudes and potential root cause type identification.

[0015] Optionally, the equipment operation parameter value comprises at least two combined parameters of heating start frequency, single heating duration, cumulative working period, and standby duration.

[0016] Based on the same inventive concept, the application also provides a data analysis-based water dispenser state monitoring system, which comprises: a data acquisition module, configured to acquire real-time water quality parameter values, equipment operation parameter values, and user water use behavior data of a water dispenser, and generate a multi-source monitoring data set; a context identification module, configured to perform mode identification on the user water use behavior data in the multi-source monitoring data set to obtain a current water use context mode, extract dynamic water quality baseline ranges and equipment operation baseline ranges corresponding to the current water use context mode from a preset context baseline library, and generate a contextualized baseline parameter set; a dynamic analysis module, configured to perform deviation degree calculation on the real-time water quality parameter values and the equipment operation parameter values and the dynamic water quality baseline ranges and the equipment operation baseline ranges in the contextualized baseline parameter set, and generate a parameter deviation feature set; and a comprehensive evaluation module, configured to perform correlation verification on the parameter deviation feature set based on a preset dynamic correlation rule base, and generate a comprehensive evaluation parameter.

[0017] A hierarchical response module is configured to perform risk level mapping according to the comprehensive evaluation parameter, obtain a risk level identification, and trigger a hierarchical response instruction matched with the risk level identification.

[0018] Compared with the prior art, the application has the following advantages:

[0019] 1. The application realizes precision and efficiency improvement of water quality monitoring of a water dispenser, effectively distinguishes normal behavior fluctuations from real equipment anomalies by establishing a mapping relationship between a water use context mode and dynamic baseline ranges, and solves the false alarm problem caused by centralized water use, periodic equipment work, and the like;

[0020] 2. An association verification mechanism of water quality parameters, equipment operation states, and user behaviors is adopted to realize accurate positioning of anomaly root causes; by strictly matching maintenance measures and risk levels, the problem of resource mismatch in a traditional maintenance scheme is overcome, and the accuracy of maintenance decisions is significantly improved.

[0021] 3. By using dynamic baseline and continuous correction of association rules, the system can adapt to long-term factors such as equipment aging and environmental changes, forming a closed-loop evolution capability that gets more accurate with use.

[0022] 4. Based on three-dimensional dynamic coupling analysis of water quality parameters, equipment operation and user behavior, a technical effect beyond single data dimension stacking is generated, which reduces the false alarm rate while significantly improving the fault prediction capability.

[0023] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art upon examination of the following or can be learned by practice of the present application. The objects and other advantages of the present application can be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings. BRIEF DESCRIPTION OF DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0025] Figure 1 is a flowchart of the water dispenser state monitoring method based on data analysis of the embodiments of the present application.

[0026] Figure 2 is a water quality parameter and equipment operation state correlation analysis diagram of the embodiments of the present application.

[0027] Figure 3 is a three-dimensional graph of association rule matching and scoring of the embodiments of the present application.

[0028] Figure 4 is a cross-validation rule contribution degree heat map of the embodiments of the present application.

[0029] Figure 5 is a structure diagram of the water dispenser state monitoring system based on data analysis of the embodiments of the present application. DETAILED DESCRIPTION

[0030] In order to make the objects, technical solutions and advantages of the embodiments of the present application more clear, the following will combine the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0031] Referring Figure 1 , one embodiment of the present application proposes a water dispenser state monitoring method based on data analysis, which adopts a context-aware dynamic baseline modeling and multi-source data cross-validation mechanism to achieve accurate anomaly diagnosis, risk classification response and model self-optimization.

[0032] The method of the embodiment specifically includes:

[0033] Obtaining real-time water quality parameter values, equipment operation parameter values and user water use behavior data of the water dispenser, and generating a multi-source monitoring data set;

[0034] Specifically, by real-time collection of water quality parameters, equipment operation parameters and user water use behavior data of the water dispenser, a mapping relationship between water use context mode and dynamic baseline range is established. The user water use behavior data is obtained by a flow sensor and a timer, including water use time period distribution, single water use amount and water use interval, etc. After time sequence alignment, all data form a multi-source monitoring data set.

[0035] Mode recognition is performed on the user water use behavior data in the multi-source monitoring data set to obtain a current water use context mode, and a dynamic water quality baseline range and an equipment operation baseline range corresponding to the current water use context mode in a preset context baseline library are extracted to generate a contextualized baseline parameter set;

[0036] Specifically, the water use frequency is counted in multiple time periods within 24 hours, and a 30-dimensional behavior feature vector is generated by combining single water use amount standard deviation and interval time regularity, matching a preset peak / conventional / low peak water use mode, and extracting the corresponding dynamic baseline range from the context baseline library.

[0037] The real-time water quality parameter values and the equipment operation parameter values are subjected to deviation calculation with the dynamic water quality baseline range and the equipment operation baseline range in the contextualized baseline parameter set to generate a parameter deviation feature set;

[0038] Specifically, the real-time detected water quality parameters such as total dissolved solids (TDS) and pH are compared with the dynamic baseline range (a reasonable fluctuation interval based on historical statistics) to calculate the percentage deviation of the measured value beyond the upper or lower limit of the baseline, and the real-time operation data such as the number of heater start-stop times and single heating time are compared with the equipment operation baseline range to quantify the difference proportion between the current value and the standard value. The deviation data of water quality and equipment are aligned by time stamp to form a structured data set containing parameter deviation values for subsequent correlation analysis. As shown in Figure 2 The equipment operation anomaly and water quality anomaly under the typical context mode category show a strong correlation.

[0039] Based on a preset dynamic correlation rule library, the parameter deviation feature set is subjected to correlation verification to generate a comprehensive evaluation parameter;

[0040] Specifically, the logical consistency of the parameter deviation feature is verified based on a preset association rule library, multi-dimensional data anomaly association is converted into a quantifiable risk assessment index, a differentiated response mechanism is triggered according to the risk level, and a closed-loop monitoring system of “data collection-situation recognition-association verification-dynamic response” is formed.

[0041] Through real-time matching of dynamic baseline and water use situation, normal fluctuations and real anomalies can be effectively distinguished, false alarms caused by concentrated water use, periodic equipment operation and other scenarios can be reduced, and sensitivity to slowly accumulated anomalies can be enhanced. The risk grading mechanism based on multi-dimensional parameter association verification closely matches the response measures with the severity of the anomaly, avoiding the mismatch of maintenance resources. The dynamic association analysis of three aspects of quality parameters, equipment operation and user behavior produces a technical effect beyond the superposition of a single dimension: the behavior pattern provides a situational basis for water quality assessment, the equipment state provides a root cause clue for water quality anomaly, and the three work together to achieve a coordinated leap in monitoring accuracy and decision-making efficiency.

[0042] Optionally, the acquisition of the user water use behavior data comprises:

[0043] Collecting water use time period distribution characteristics, single water use amount distribution characteristics and water use interval characteristics within a preset time window;

[0044] Specifically, collecting water use time period distribution characteristics within a preset time window comprises recording the occurrence time and duration of each water use event within the preset time window through a flow sensor and a timer of a water dispenser, dividing the preset time window into a plurality of continuous time periods, counting the total number of water use events in each time period, and forming a water use time period distribution characteristic vector; collecting single water use amount distribution characteristics comprises counting the water amount value of each water use event within the preset time window, calculating the average value and standard deviation of the single water use amount, and forming a single water use amount distribution characteristic vector, wherein the average value of the single water use amount reflects the typical water use amount level, and the standard deviation represents the water use amount fluctuation degree; collecting water use interval characteristics comprises calculating the time interval between adjacent two water use events within the preset time window, calculating the average value and standard deviation of all time intervals, and forming a water use interval characteristic vector, wherein the average value of the time interval reflects the average waiting time, and the standard deviation represents the water use regularity.

[0045] Performing pattern recognition on the water use time period distribution characteristics, the single water use amount distribution characteristics and the water use interval characteristics to generate the user water use behavior data.

[0046] Specifically, the mode recognition on the water use time period distribution characteristics, the single water use amount distribution characteristics and the water use interval characteristics includes splicing the water use time period distribution characteristic vector, the single water use amount distribution characteristic vector and the water use interval characteristic vector to generate a user water use behavior data vector, which contains the joint representation of the time period distribution characteristic dimension, the water use amount statistical characteristic dimension and the interval characteristic dimension.

[0047] Exemplarily, taking an office drinking water machine as an example, 56 water use events are recorded within a 24-hour time window, the number of water use per hour is counted to form a 24-dimensional time period distribution characteristic vector, the average value of the single water use amount in the 56 water use events is 230 milliliters, and the standard deviation is 85 milliliters, forming a single water use amount distribution characteristic vector, the average value of the 55 adjacent water use intervals is 26 minutes, and the standard deviation is 18 minutes, forming a water use interval characteristic vector, and the three characteristic vectors are spliced to generate a 30-dimensional user water use behavior data vector. Through the joint representation of multiple dimensions, this method can accurately capture the concentrated water use characteristics during the peak working hours and the intermittent water use characteristics during the lunch break, avoid mode misjudgment caused by single dimension statistics, provide comprehensive and reliable behavior characteristic input for subsequent situation recognition, and significantly improve the accuracy of water use mode judgment.

[0048] Optionally, the method further comprises:

[0049] acquiring a multi-dimensional monitoring data set in a historical period, the multi-dimensional monitoring data set containing historical water quality parameter values, historical equipment operation parameter values and historical user water use behavior data;

[0050] Specifically, acquiring the multi-dimensional monitoring data set in the historical period includes extracting historical records of the past 30 consecutive days from the drinking water machine memory, and the data set contains historical water quality parameter values sampled every hour, historical equipment operation parameter values triggered by each water use event and corresponding historical user water use behavior data.

[0051] dividing typical situation mode categories according to the historical user water use behavior data;

[0052] Specifically, dividing the typical situation mode categories according to the historical user water use behavior data includes dividing the 24 hours of each day into three fixed categories of peak water use period, regular water use period and low peak water use period, wherein the peak water use period is defined as 8:00-10:00 and 14:00-16:00 every day, the regular water use period is 10:00-14:00, and the low peak water use period is the remaining period.

[0053] calculating water quality parameter fluctuation threshold intervals and equipment operation parameter fluctuation threshold intervals under each typical situation mode category to form dynamic water quality baseline ranges and equipment operation baseline ranges, and constructing a situation baseline library;

[0054] Specifically, the water quality parameter fluctuation threshold interval under each typical situation mode category is calculated using the following formula:

[0055]

[0056] wherein is the upper threshold value, is the lower threshold value, is the arithmetic mean of all historical water quality parameter values under the situation, is the standard deviation, forming a dynamic water quality baseline range; the device operation baseline range is calculated using the same method to obtain the device parameter fluctuation threshold interval.

[0057] Based on the dynamic water quality baseline range and the device operation baseline range, the correlation between the water quality parameter change trend and the device operation state is analyzed, and a dynamic correlation rule library is generated.

[0058] Specifically, analyzing the correlation between the water quality parameter change trend and the device operation state includes counting the proportion of synchronous water quality parameter abnormalities when the device operation parameter exceeds the baseline range under each situation mode. When the proportion exceeds the preset threshold of 75%, the correlation rule of "device abnormality leading to water quality abnormality" is generated, and when the proportion is less than 25%, the correlation rule of "water quality abnormality occurring independently" is generated, finally forming a dynamic correlation rule library.

[0059] Exemplarily, taking the historical data of a school water dispenser as an example, the of TDS value in the low peak water period is 205 ppm, is 15 ppm, and the dynamic water quality baseline range is 160-250 ppm; the of heating start frequency in the same period is 0.8 times / hour, is 0.3 times, and the device operation baseline range is 0-1.7 times / hour; analysis shows that when the heating frequency is abnormally high, 93% is accompanied by an increase in TDS value, so the strong correlation rule of "abnormal heating frequency in low peak period → abnormal TDS" is generated. The causal relationship model between device operation and water quality change is established by quantitative statistics, which can distinguish between water quality abnormalities caused by device failure and water quality abnormalities caused by external pollution, significantly improve the accuracy of root cause judgment, and provide decision basis for hierarchical response.

[0060] Optionally, the correlation verification based on the preset dynamic correlation rule library comprises:

[0061] extracting the correlation verification rule matched with the current water use situation mode from the dynamic correlation rule library;

[0062] ​​Specifically, the extraction of the associated verification rule in the dynamic association rule base matched with the current water use context mode includes retrieving all associated rule entries in the rule base marked with the context identifier according to the current water use context mode identifier output by the context recognition module.

[0063] The parameter deviation feature set is input into the associated verification rule for logical verification, and an abnormality credibility score and a potential root cause type identifier are output.

[0064] Specifically, inputting the parameter deviation feature set into the associated verification rule for logical verification includes traversing the condition expression of each associated rule. For example, when it is detected that the parameter deviation feature set contains "TDS value single jump amplitude exceeds 50 ppm during low peak period" and there is also "standby time greater than 4 hours", the "abnormal concentration abnormality" judgment of rule number R001 is activated.

[0065] The abnormality credibility score, the potential root cause type identifier, and the parameter deviation feature set are combined to generate the comprehensive evaluation parameter.

[0066] Specifically, the output abnormality credibility score uses a rule weight accumulation model:

[0067] ,

[0068] wherein is the abnormality credibility score, is the rule importance weight predefined by the rule base, is the current rule condition matching degree, and the matching degree is linearly valued between 0 and 1 according to the ratio of the actual parameter deviation value to the rule threshold value, as shown in Figure 3 The weight and condition matching degree of different rules determine the current rule score; the potential root cause type identifier is directly output according to the root cause category field predefined in the completely matched rule entry;

[0069] The combination of the abnormality credibility score, the potential root cause type identifier, and the parameter deviation feature set generates the comprehensive evaluation parameter, which is specifically implemented by constructing a triple data structure wherein is the abnormality credibility score, is the potential root cause type identifier, represents the parameter deviation feature set.

[0070] Exemplarily, the hospital night nursing station water dispenser is in a low peak water use context, and the parameter deviation feature set shows that the TDS value rises from 180 ppm to 280 ppm within 2 hours, the standby time is 3.5 hours, and the matching rule R002 trigger condition requires "TDS rise > 80 ppm and standby > 3 hours". The rule weight , actual increase 100ppm makes the matching degree , so The root cause identification is "filter failure"; meanwhile, no device operation abnormality related rule is matched;

[0071] The comprehensive evaluation parameter is generated after the generation By cross-verification of multiple rules, the possibility of device failure is excluded, the root cause of filter failure is accurately locked, and the maintenance resource mismatch caused by misjudgment of heater failure is avoided. The method significantly improves the accuracy of maintenance decision and provides a reliable basis for subsequent automatic generation of filter replacement work order.

[0072] Optionally, the correlation verification rule comprises: a low water use context rule, a peak water use context rule, and a cross-verification rule; wherein,

[0073] The low water use context rule is that when a long standby period is detected and the water quality parameter value shows a slow rising trend, it is determined as a normal concentration feature;

[0074] Specifically, the execution of the low water use context rule includes monitoring that the current water use context mode is identified as a low peak period and the standby time length characteristic value extracted from the parameter deviation feature set is greater than a preset threshold of 4 hours, and detecting that the water quality parameter value change slope is in a preset slow rising interval of 0.5-2ppm / min. Then, the normal concentration feature determination identification is triggered.

[0075] The peak water use context rule is that when a concentrated water use period is detected and the water quality parameter value abnormally rises, a device failure suspicion identification is triggered;

[0076] Specifically, the execution of the peak water use context rule includes identifying that the current is a concentrated water use period and the real-time monitoring value of the water quality parameter value exceeds the upper limit of the dynamic water quality baseline range, and detecting that the heating start frequency in the device operation parameter value exceeds the upper limit of the device operation baseline range. Then, the device failure suspicion identification is activated.

[0077] The cross-verification rule is that when the water quality parameter deviation feature and the device operation abnormality feature exist at the same time, the abnormality confidence score level is improved.

[0078] Specifically, the execution of the cross-verification rule includes that when at least one parameter in the water quality parameter deviation feature set exceeds the dynamic water quality baseline range and at least one parameter in the device operation parameter deviation feature set exceeds the device operation baseline range, a confidence improvement algorithm is used:

[0079] ,

[0080] wherein is an abnormality confidence score increment value, is a water quality parameter deviation degree, wherein is a water quality parameter actual value, is a water quality baseline upper limit value, is a device parameter deviation degree, , is a device parameter actual value, is a device baseline upper limit value, and are preset weighting coefficients respectively taking values of 0.6 and 0.4, and is added to the original abnormality credibility score value, as shown in Figure 4 the cross-verification rule is determined by the weighted calculation values of the water quality parameter deviation degree and the device parameter deviation degree.

[0081] Exemplarily, the school gymnasium drinking water machine detects that the TDS value 310ppm exceeds the dynamic baseline range upper limit 280ppm during the peak period of the match interval, and the heating start frequency 25 times / hour exceeds the baseline range upper limit 20 times / hour , triggering the cross-verification rule calculation , which makes the original credibility score increase from 0.7 to 0.864. Through the multi-dimensional abnormality cooperative verification mechanism, the detection confidence of the complex fault is significantly improved, the single parameter accidental fluctuation and the real device fault are effectively distinguished, the TDS dilution phenomenon caused by a large number of personnel drinking water during the match interval is avoided from being misjudged as water quality abnormality, and the overload operation hidden danger of the heater is accurately captured, thereby providing strong decision support for early maintenance.

[0082] Optionally, the trigger hierarchical response instruction comprises:

[0083] when the risk level is identified as the attention observation level, an event log record containing the deviation feature is generated;

[0084] Specifically, when the risk level is identified as the attention observation level, the system calls the event log generation interface, binds the parameter deviation feature set with the current timestamp, generates an event log record containing the deviation feature value and the context mode, and stores it to the database.

[0085] when the risk level is identified as the suggestion detection level, a water quality abnormality report is generated and a detection reminder notification is sent;

[0086] Specifically, when the risk level is identified as the suggestion detection level, the potential root cause type identifier and the key deviation feature are extracted from the comprehensive evaluation parameters, the report template engine is called to generate a water quality abnormality report containing an abnormal parameter comparison chart and a root cause speculation text description, and a detection reminder notification containing a report abstract is sent to the administrator terminal through the message push interface.

[0087] when the risk level is identified as the maintenance warning level, a device maintenance proposal is generated and a preset maintenance work order interface is activated;

[0088] Specifically, when the risk level is identified as a maintenance warning level, the potential root cause type is identified to determine the maintenance type, the preset maintenance knowledge base is retrieved to match the maintenance solution, the equipment maintenance proposal is generated with the equipment list and the maintenance work order interface is activated to write the work order data to the external maintenance system queue.

[0089] When the risk level is identified as an emergency warning level, the outlet locking operation is performed and the emergency warning information containing the comprehensive evaluation parameters is pushed.

[0090] Specifically, when the risk level is identified as an emergency warning level, a locking instruction containing the target outlet number is sent to the water dispenser control panel to make the electromagnetic valve power off and close, and the comprehensive evaluation parameter full field is extracted to generate an encrypted data packet, which is pushed to all management terminals through the emergency warning channel.

[0091] Exemplarily, the airport water dispenser detects that the water temperature value continuously drops below the dynamic baseline range of 15℃ and the heating start frequency is abnormal in the peak water consumption situation,

[0092] Comprehensive evaluation parameters Triggering an emergency warning level response:

[0093] The system immediately locks the hot water outlet to prevent low temperature water from being misdrunk, generates an encrypted warning packet containing the heater fault diagnosis graph and data curve, and pushes it to the airport operation and maintenance center in real time; simultaneously activates the equipment maintenance proposal generation process, outputs the heater replacement operation guide and spare parts list. Through the multi-level response mechanism, the risk is accurately controlled, which not only avoids the safety hazards caused by low temperature drinking water, but also provides detailed fault diagnosis basis for maintenance personnel, greatly shortens the equipment downtime and reduces the manual troubleshooting cost, and reflects the synergistic effect of automatic diagnosis and active protection.

[0094] Optionally, the method further comprises:

[0095] Recording the triggered hierarchical response instructions and subsequent processing feedback data;

[0096] Specifically, recording the triggered hierarchical response instructions and subsequent processing feedback data comprises creating a tracking record after each execution of the response operation, storing the response instruction type, trigger timestamp, and corresponding comprehensive evaluation parameter copy;

[0097] According to the subsequent processing feedback data, the dynamic water quality baseline range, the equipment operation baseline range and the dynamic correlation rule base are corrected.

[0098] Specifically, the dynamic water quality baseline range is corrected according to the subsequent processing feedback data by using the weighted moving average algorithm. For the new baseline upper limit , there are:

[0099] ,

[0100] wherein a is a preset smoothing factor value 0.7, is the current baseline upper limit, is the historical maximum upper limit, is the feedback correction amount. The historical maximum upper limit is taken from the maximum value of the water quality parameter in the 30 days before the event in the context, and the feedback correction amount takes a negative deviation compensation value of -5% when the verification tag is a false alarm, and a positive deviation expansion value of +3% when the alarm is true; the modified device operation baseline range uses the same algorithm; updating the dynamic association rule base includes the association rule judgment accuracy rate in the statistical feedback data, and when a rule is misjudged for 3 times in a row, the weight of the rule is reduced by 20%, and when a new root cause combination appears more than 2 times in the feedback data, a new rule entry with a temporary mark is created.

[0101] Exemplarily, the drinking water machine in the office building triggers a maintenance early warning level response due to the slow increase of TDS in the off-peak period, and the subsequent maintenance work order feedback confirms that the filter core is normally replaced due to aging; the system extracts the TDS value when the event occurs as 280ppm, the original dynamic water quality baseline upper limit is 260ppm, and the new upper limit is calculated according to the algorithm ; at the same time, the rule of "low peak long standby TDS slow rise→filter core failure" in the association rule base is verified as true, and the weight is increased by 10%. Through the feedback loop, the dynamic evolution of the baseline model is realized, so that the system can adapt to the parameter drift caused by the performance degradation of the device, continuously improve the monitoring accuracy, and at the same time, optimize the weight of the association rule based on the empirical data, significantly reduce the false alarm probability in the initial stage of the new filter core installation, and form the self-adaptive ability of getting more accurate with use.

[0102] Optionally, the generating a water quality anomaly report comprises:

[0103] associating dynamic water quality baseline historical data in the current water use context mode;

[0104] Specifically, the association of dynamic water quality baseline historical data in the current water use context mode comprises retrieving all water quality parameter baseline records of the context mode in the last 30 days from the context baseline library, and extracting the median value in the baseline as the historical reference line.

[0105] Comparing the deviation amplitude of the real-time water quality parameter value with the dynamic water quality baseline historical data;

[0106] Specifically, the comparison of the deviation amplitude of the real-time water quality parameter value with the dynamic water quality baseline historical data adopts a relative deviation algorithm:

[0107] ,

[0108] Wherein, D represents the deviation percentage of real-time water quality parameter value relative to historical baseline value, R represents the specific value of water quality parameter obtained by real-time monitoring; H represents the median value extracted from the baseline record of water quality parameter in the last 30 days as the historical reference value.

[0109] Generate a readability report containing situational correlation analysis based on the deviation amplitude and the potential root cause type identification.

[0110] Specifically, generating a readability report in combination with potential root cause type identification includes mapping root cause identification to a preset natural language description template, dynamically inserting specific water quality parameter name, deviation percentage value and occurrence period, and displaying the comparison chart of historical baseline value curve and real-time monitoring value curve in association, and finally outputting a water quality anomaly report document containing situational correlation analysis paragraph, data visualization chart and maintenance suggestion text.

[0111] For example, the factory break area water dispenser detects TDS value 380 ppm under the regular water use situation, the system retrieves the historical baseline value of this situation as 210 ppm, and calculates the deviation percentage as The potential root cause type identification is water source pollution, and the mapping description template generates "detecting TDS value anomaly increase of 81% under regular water use period, significantly deviating from historical baseline level of 210 ppm, suspected external water source pollution event"; the report simultaneously outputs the comparison chart of TDS change curve in this week and the same situation in the past four weeks. Through historical data visualization comparison, the administrator can intuitively master the severity of the anomaly, accurately locate the pollution event combined with natural language description, avoid confusion with slow-rising anomalies caused by filter failure, greatly improve the efficiency of anomaly disposal and reduce the time delay of water quality inspection.

[0112] Optionally, the device operating parameter value includes:

[0113] At least two combined parameters in heating start frequency, single heating duration, cumulative working period and standby time.

[0114] Specifically, obtaining the heating start frequency in the device operating parameter value includes counting the number of times of triggering heating when the water temperature monitored by the temperature sensor is lower than the preset threshold, and obtaining the heating start frequency value by counting the number of times of triggering in unit time; collecting single heating duration includes recording the time interval from the start of each heating action to reaching the set temperature, and taking the heating duration value; cumulative working period is obtained by cumulative device power-on total duration through real-time clock; standby time records the time difference value from the end of the last water use event to the current time; combined parameter processing uses independent channel analysis: setting the first monitoring channel for heating start frequency to calculate its deviation degree from the device operating baseline range , setting the second monitoring channel for single heating duration to calculate The rest of the parameters are independently calculated, and finally a multi-dimensional deviation feature vector is formed as the basis for analyzing the device operating parameter values.

[0115] For example, the hotel room water dispenser records 18 times / hour of heating start frequency during the 8-9 am peak period (baseline upper limit 15 times), the average single heating duration is 4 minutes (baseline range 3-5 minutes), the cumulative working period is 1200 hours, and the standby duration is 8 minutes; independently calculate , (in the baseline upper limit), forming a feature vector . By independent analysis of multiple parameters, dimension confusion is avoided, both abnormal load conditions with high frequency heating are identified, and single heating efficiency is verified to be normal, excluding the possibility of misjudgment of heater aging, providing fine data support for device health diagnosis, and significantly improving the recognition accuracy of compound faults.

[0116] Based on the same inventive concept, as shown in Figure 5 , the present application also provides a water dispenser state monitoring system based on data analysis, which comprises:

[0117] A data acquisition module is used to acquire real-time water quality parameter values, device operating parameter values and user water consumption behavior data of the water dispenser, and generate a multi-source monitoring data set;

[0118] A context recognition module is used to perform pattern recognition on the user water consumption behavior data in the multi-source monitoring data set to obtain a current water consumption context mode, and extract a dynamic water quality baseline range and a device operating baseline range corresponding to the current water consumption context mode in a preset context baseline library to generate a contextualized baseline parameter set;

[0119] A dynamic analysis module is used to calculate the deviation degree of the real-time water quality parameter values and the device operating parameter values from the dynamic water quality baseline range and the device operating baseline range in the contextualized baseline parameter set to generate a parameter deviation feature set;

[0120] A comprehensive evaluation module is used to verify the correlation of the parameter deviation feature set based on a preset dynamic correlation rule library to generate a comprehensive evaluation parameter;

[0121] A hierarchical response module is used to map the risk level based on the comprehensive evaluation parameter to obtain a risk level identifier, and trigger a hierarchical response instruction matching the risk level identifier.

[0122] It should be noted that the electrical connection between the above-mentioned units does not necessarily represent the direct connection of the line, and the indirect connection mode can be applied to the embodiments of the present application as long as the purpose of the present application is achieved. The above-mentioned is only an exemplary embodiment of the present application, and cannot limit the scope of the present application.

[0123] That is, any equivalents of the above described subject matter, as well as other modifications and variations that can occur to those skilled in the art are intended to be covered. Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the application being indicated by the following claims.

Claims

1. A method for monitoring the status of a water dispenser based on data analysis, characterized in that, The method comprises: obtaining real-time water quality parameter values, equipment operation parameter values and user water use behavior data of the water dispenser, and generating a multi-source monitoring data set; performing pattern recognition on the user water use behavior data in the multi-source monitoring data set to obtain a current water use context mode, and extracting a dynamic water quality baseline range and an equipment operation baseline range corresponding to the current water use context mode in a preset context baseline library to generate a contextualized baseline parameter set; calculating the deviation of the real-time water quality parameter values and the equipment operation parameter values from the dynamic water quality baseline range and the equipment operation baseline range in the contextualized baseline parameter set to generate a parameter deviation feature set; verifying the parameter deviation feature set based on a preset dynamic correlation rule library to generate a comprehensive evaluation parameter; mapping the risk level based on the comprehensive evaluation parameter to obtain a risk level identifier, and triggering a hierarchical response instruction matched with the risk level identifier.

2. The data analysis based water cooler condition monitoring method as claimed in claim 1 wherein, The acquisition of the user water use behavior data comprises: collecting water use time period distribution characteristics, single water use amount distribution characteristics and water use interval characteristics within a preset time window; performing pattern recognition on the water use time period distribution characteristics, the single water use amount distribution characteristics and the water use interval characteristics to generate the user water use behavior data.

3. The data analysis based water cooler condition monitoring method as claimed in claim 1 wherein, The method further comprises: obtaining a multi-dimensional monitoring data set in a historical period, which comprises historical water quality parameter values, historical equipment operation parameter values and historical user water use behavior data; dividing typical context mode categories according to the historical user water use behavior data; calculating water quality parameter fluctuation threshold intervals and equipment operation parameter fluctuation threshold intervals under each typical context mode category to form dynamic water quality baseline ranges and equipment operation baseline ranges, and constructing a context baseline library; based on the dynamic water quality baseline ranges and the equipment operation baseline ranges, analyzing the correlation between water quality parameter change trends and equipment operation states to generate a dynamic correlation rule library.

4. The data analysis based water cooler condition monitoring method as claimed in claim 1 wherein, The correlation verification based on the preset dynamic correlation rule library comprises: extracting a correlation verification rule matched with the current water use context mode from the dynamic correlation rule library; inputting the parameter deviation feature set into the correlation verification rule for logical verification to output an abnormality credibility score and a potential root cause type identifier; combining the abnormality credibility score, the potential root cause type identifier and the parameter deviation feature set to generate the comprehensive evaluation parameter.

5. The data analysis based water cooler condition monitoring method as claimed in claim 4, wherein, The correlation verification rule comprises: a low water use context rule, a peak water use context rule and a cross verification rule; wherein, the low water use context rule is that when a long standby period is detected and the water quality parameter value shows a slow rising trend, it is determined as a normal concentration feature; the peak water use context rule is that when a concentrated water use period is detected and the water quality parameter value abnormally rises, an equipment fault suspicion identifier is triggered; the cross verification rule is that when water quality parameter deviation features and equipment operation abnormality features exist at the same time, the abnormality credibility score level is improved.

6. The data analytics based water cooler condition monitoring method as claimed in claim 1 wherein, The triggering of the hierarchical response instruction comprises: when the risk level identifier is the observation level, an event log record containing the deviation feature is generated; When the risk level is identified as the suggestion detection level, a water quality anomaly report is generated and a detection reminder notification is sent; When the risk level is identified as the maintenance warning level, a device maintenance proposal is generated and a preset maintenance work order interface is activated; When the risk level is identified as the emergency warning level, an outlet locking operation is performed and emergency warning information containing comprehensive evaluation parameters is pushed.

7. The data analytics based water cooler condition monitoring method as claimed in claim 3, wherein, The method further comprises: Recording triggered hierarchical response instructions and subsequent processing feedback data; According to the subsequent processing feedback data, the dynamic water quality baseline range, the device operation baseline range and the dynamic correlation rule base are corrected.

8. The data analytics based water cooler condition monitoring method as claimed in claim 6, wherein, The generation of the water quality anomaly report comprises: Correlating dynamic water quality baseline historical data under the current water use context mode; Comparing the deviation amplitude of the real-time water quality parameter value and the dynamic water quality baseline historical data; Based on the deviation amplitude and the potential root cause type identification, a readable report containing context correlation analysis is generated.

9. The data analytics based water cooler condition monitoring method as claimed in claim 1 wherein, The device operation parameter value comprises: At least two combined parameters in heating start frequency, single heating duration, cumulative working period and standby duration.

10. A data analysis based water dispenser condition monitoring system applied to the data analysis based water dispenser condition monitoring method according to any one of claims 1-9, characterized in that, The system comprises: A data acquisition module for acquiring real-time water quality parameter values, device operation parameter values and user water use behavior data of the drinking water machine, and generating a multi-source monitoring data set; A context recognition module for performing pattern recognition on the user water use behavior data in the multi-source monitoring data set to obtain a current water use context mode, and extracting a dynamic water quality baseline range and a device operation baseline range corresponding to the current water use context mode from a preset context baseline library to generate a contextual baseline parameter set; A dynamic analysis module for calculating the deviation degree of the real-time water quality parameter values and the device operation parameter values from the dynamic water quality baseline range and the device operation baseline range in the contextual baseline parameter set to generate a parameter deviation feature set; A comprehensive evaluation module for correlating the parameter deviation feature set based on a preset dynamic correlation rule base to generate comprehensive evaluation parameters; A hierarchical response module for mapping the risk level according to the comprehensive evaluation parameters to obtain a risk level identification, and triggering hierarchical response instructions matched with the risk level identification.

Images