Method, device and equipment for evaluating credibility of monitoring data of water conservancy Internet of Things

Abstract

The invention provides a water conservancy Internet of Things monitoring data credibility evaluation method, device and equipment, and relates to the technical field of water conservancy Internet of Things monitoring, and the method comprises the steps: analyzing the physical characteristics of the time sequence change of a monitored object, determining the change boundary and trend in unit time, and calculating a recommendation value; meanwhile, analyzing a logic relationship between the monitored object and the association factor, and calculating a first credibility for a strong association relationship by adopting a recommendation value method; under the condition that no strong association relationship exists, the association degree method is adopted, the data association degree of the auxiliary monitoring object and the monitoring object is analyzed, association degree coefficients are distributed, and second credibility is calculated; in practical application, a recommendation value method is preferentially adopted, then a correlation degree method is adopted, and a worker judges and corrects a monitored and acquired data value according to a credibility score, so that auxiliary manual judgment and abnormal data correction are realized.

Description

A method, apparatus, and equipment for assessing the reliability of water conservancy Internet of Things (IoT) monitoring data. Technical Field

[0001] This invention relates to the field of water conservancy Internet of Things (IoT) monitoring technology, specifically to a method, apparatus, and equipment for evaluating the reliability of water conservancy IoT monitoring data. Background Technology

[0002] Against the backdrop of the deep integration of water conservancy informatization and Internet of Things (IoT) technology, various business systems such as hydrology, water resources, and water project safety have widely relied on sensor networks deployed in open environments such as rivers, lakes, reservoirs, sluice gates, pumping stations, culverts, dikes, and canal systems to achieve unattended operation and remote sensing. However, the monitoring environment of water conservancy is open, changeable, and susceptible to interference. This is because these monitoring nodes are exposed to the field year-round and are affected by heterogeneous interference from multiple sources, such as seasonal changes, electromagnetic disturbances, biological attachment, siltation and erosion, unstable power supply, communication packet loss, and equipment aging. The collected IoT monitoring data often contains various anomalies such as jumps, drifts, intermittent dead values, slow deviations, and false normalities, leading to misjudgments, interfering with the normal judgment of staff, and invisibly increasing the ineffective workload of staff.

[0003] The accuracy of collected water conservancy data has always been a common problem in the industry. Due to the special nature of the environment, current methods for handling erroneous data mainly rely on extreme value analysis to address extreme anomalies, but there is no truly effective solution. This invention primarily assists manual judgment and the correction of anomalies. However, extreme value analysis can only resolve a small portion of anomalies; it cannot identify most. Furthermore, it cannot assist personnel in correction, meaning they don't know the appropriate correction value. This is because extreme anomalies are generally caused by sensor malfunctions or extreme interference conditions. Most environmental interference does not lead to abnormal values. Personnel cannot correct values ​​based on a single data point; relying on personal experience carries certain risks.

[0004] In view of the above, this application is hereby submitted. Summary of the Invention

[0005] This invention provides a method, apparatus, and equipment for evaluating the reliability of water conservancy Internet of Things (IoT) monitoring data, which can at least partially improve the above-mentioned problems.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a method for evaluating the credibility of water conservancy Internet of Things (IoT) monitoring data, comprising: acquiring the current monitoring object to be monitored, determining auxiliary monitoring objects, collecting monitoring data of the monitoring object, and analyzing the logical relationship between the monitoring object and the auxiliary monitoring object; when the logical relationship is determined to be a strong correlation, acquiring measured values, calculating recommended values, and calculating a first credibility based on the measured values ​​and recommended values; when the logical relationship is determined to be a weak correlation, analyzing the correlation between the monitoring data of the monitoring object and the auxiliary monitoring object, generating analysis results, and calculating a second credibility based on the correlation coefficient corresponding to the current auxiliary monitoring object and the analysis results; comparing and processing the first credibility or the second credibility to generate a corresponding evaluation result.

[0007] This invention also provides a device for evaluating the reliability of water conservancy Internet of Things (IoT) monitoring data, comprising: a data acquisition unit, used to acquire the current monitoring object to be monitored, determine the auxiliary monitoring object, collect the monitoring data of the monitoring object, and analyze the logical relationship between the monitoring object and the auxiliary monitoring object; a strong correlation unit, used to acquire measured values, calculate recommended values, and calculate a first reliability based on the measured values ​​and recommended values ​​when the logical relationship is determined to be a strong correlation; a weak correlation unit, used to analyze the correlation between the monitoring data of the monitoring object and the auxiliary monitoring object when the logical relationship is determined to be a weak correlation, generate analysis results, and calculate a second reliability based on the correlation coefficient corresponding to the current auxiliary monitoring object and the analysis results; and an evaluation unit, used to compare and process the first reliability or the second reliability to generate a corresponding evaluation result.

[0008] The present invention also provides a device for evaluating the reliability of water conservancy Internet of Things (IoT) monitoring data, comprising: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the water conservancy IoT monitoring data reliability evaluation method as described above.

[0009] In summary, this invention assesses the reliability of alarm data, assisting staff in accurately determining the reliability of alarms. This method analyzes the physical characteristics of the time-series changes of the monitored object to determine the boundary and trend of changes per unit time, calculates recommended values, and simultaneously analyzes the logical relationship between the monitored object and related factors. For strong correlations, the recommended value method is used to calculate the first level of reliability; for cases without strong correlations, the correlation degree method is used to analyze the data correlation between the monitored object and the monitored object, assigning correlation coefficients such as rainfall coefficients to calculate the second level of reliability. In practical applications, the recommended value method is preferred, followed by the correlation degree method. Staff judge and correct the monitored data values ​​based on the reliability score, thus assisting in manual judgment and correcting abnormal data. Attached Figure Description

[0010] Figure 1 is a flowchart illustrating the reliability assessment method for water conservancy IoT monitoring data provided in the first embodiment of the present invention.

[0011] Figure 2 is a schematic flowchart of the reliability assessment method for water conservancy Internet of Things monitoring data provided in the first embodiment of the present invention.

[0012] Figure 3 is a schematic diagram illustrating an example of the hydraulic physical mechanism characteristics provided by the present invention.

[0013] Figure 4 is a schematic diagram of the module of the water conservancy Internet of Things monitoring data credibility assessment device provided in the second embodiment of the present invention. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0015] Referring to Figures 1 and 2, the first embodiment of the present invention discloses a method for evaluating the reliability of water conservancy Internet of Things (IoT) monitoring data. This method can be executed by a water conservancy IoT monitoring data reliability evaluation device (hereinafter referred to as the evaluation device), specifically by one or more processors within the evaluation device, to achieve the following: S1, acquiring the current monitoring object to be monitored, determining auxiliary monitoring objects, collecting monitoring data of the monitoring object, and analyzing the logical relationship between the monitoring object and the auxiliary monitoring object; specifically, step S1 further includes: acquiring the current monitoring object to be monitored, converting the correlation factors of the monitoring object into auxiliary monitoring objects, and collecting monitoring data of the monitoring object, wherein the correlation factors include upstream inflow and downstream flow, and the monitoring data includes reservoir water level and water level fluctuations; based on preset water conservancy physical mechanism characteristics, analyzing the logical relationship between the monitoring object and the auxiliary monitoring object, and determining the logical relationship, wherein the logical relationship includes strong correlation and weak correlation.

[0016] In this embodiment, the current monitoring object is acquired, and monitoring data of the monitoring object is collected. For example, the water level of a reservoir is monitored, and the rise and fall of the water level represents the change in the reservoir capacity. It is related to the inflow and outflow of water. Therefore, the related factors of the monitoring object are transformed into auxiliary monitoring objects, namely the upstream inflow and outflow.

[0017] After data collection, logical relationship analysis is performed on the aforementioned objects based on pre-defined hydraulic physical mechanism characteristics. These hydraulic physical mechanisms are numerous, such as the water balance theory, which can be seen in Figure 3 (where water balance is a physical mechanism, a specific manifestation of the law of conservation of mass in the water cycle). Combining the specific monitoring objects with the hydraulic physical mechanism characteristics, the analysis identifies whether the current logical relationship is strongly or weakly correlated. Different correlation algorithms are used for different logical relationships. This step clarifies whether the recommended value method or the correlation degree method should be used for subsequent credibility assessment, thus avoiding misjudgments caused by the traditional extreme value method, improving the physical rationality of the assessment starting point, and reducing the time spent manually sorting out relationships.

[0018] S2, when the logical relationship is determined to be a strong correlation, the measured value is obtained, the recommended value is calculated, and the first confidence level is calculated based on the measured value and the recommended value; specifically, step S2 further includes: when the logical relationship is determined to be a strong correlation, the measured value obtained by the preset sensor is obtained, and the recommended value is calculated according to the strong correlation algorithm corresponding to the strong correlation; the first confidence level is calculated based on the measured value and the recommended value, and the formula is first confidence level = ABS(recommended value - measured value) / recommended value, where ABS is the absolute value.

[0019] In this embodiment, when the logical relationship is determined to be a strong correlation, a recommended value method is adopted, which uses prior values ​​that can be calculated by algorithms based on physical characteristics, such as parabolic curves, lookup tables, etc. First, the measured values ​​output by the preset sensor at the current moment are read; then, the correlation algorithm corresponding to the strong correlation is called to calculate the recommended value. These correlation algorithms include the generalization method, average method, linear method, extreme value method, cubic spline method, nearest neighbor method, ARIMA model, etc. After obtaining two sets of values, arithmetic operations are performed according to the formula "First confidence level = ABS(recommended value - measured value) / recommended value". The result can be displayed intuitively as a percentage on the duty interface. For example, the flow rate can be calculated from the water level using the water level as an example. Similarly, the flow rate can be calculated using the water balance algorithm using the inflow / outflow and water level / reservoir capacity curves.

[0020] S3, when it is determined that the logical relationship is not a strong correlation, the correlation between the monitoring data of the monitoring object and the auxiliary monitoring object is analyzed, an analysis result is generated, and a second confidence level is calculated based on the correlation coefficient corresponding to the current auxiliary monitoring object and the analysis result; specifically, step S3 further includes: when it is determined that the logical relationship is not a strong correlation, the correlation between the monitoring data of the monitoring object and the auxiliary monitoring object is analyzed, and it is determined whether the changing trend of the auxiliary monitoring object and the monitoring data is consistent, and an analysis result is generated. Wherein, when the changing trend of the auxiliary monitoring object and the monitoring data is consistent, it is true, and the value of the analysis result is 1; when the changing trend of the auxiliary monitoring object and the monitoring data is inconsistent, it is false, and the value of the analysis result is 0; the correlation coefficient corresponding to the current auxiliary monitoring object is obtained, and combined with the analysis result, the second confidence level is calculated, and the formula is: second confidence level = Where n is the number of auxiliary monitoring objects, Let be the correlation coefficient of the i-th auxiliary monitoring object, which is assigned according to the strength of the correlation.

[0021] In this embodiment, when the logical relationship is determined to be a weak correlation, meaning no related object can be found for the algorithm to calculate the prior value, the correlation degree method is used. Since there are multiple auxiliary monitoring objects in the actual processing, a corresponding correlation degree coefficient is assigned to each auxiliary monitoring object based on the strength of the correlation. For example, the coefficient for auxiliary monitoring object 1 is X1, the coefficient for object 2 is X2, the coefficient for object N is XN, and X1 + X2 + ... + XN = 1. That is, the coefficient represents the weight of the correlation relationship; the larger the coefficient, the higher the weight, i.e., the higher the correlation degree. For example, in a reservoir without direct upstream water, if the reservoir water level rises, the related object is rainfall. Analyzing the rainfall over the past 24 hours, there are coefficients for heavy rain (0.5), moderate rain (0.4), and light rain (0.1).

[0022] Next, the consistency of the monitoring data and the auxiliary monitoring object's trend is judged. If the monitoring data and the auxiliary monitoring object's trend are consistent, it is recorded as "true" and assigned 1; otherwise, it is recorded as "false" and assigned 0. For example, regarding the relationship between reservoir water level and rainfall, a rising water level should correspond to rainfall, and rainfall is consistent with the rising water level trend. Heavy rain is also considered a true indicator. Subsequently, according to the formula, the second confidence level is determined... Complete the calculation.

[0023] S4, compare the first confidence level or the second confidence level to generate the corresponding evaluation result.

[0024] Specifically, step S4 further includes: when the logical relationship is determined to be a strong correlation, the first confidence level is compared with a preset first threshold, and it is determined whether the first confidence level is less than the first threshold, and a corresponding evaluation result is generated; if yes, it indicates that the measured value is reliable and the evaluation result is normal; if no, it indicates that the measured value is abnormal and the evaluation result is a warning.

[0025] In this embodiment, the smaller the absolute value of the first confidence level, the more reliable the measured value. If it exceeds the first threshold, the measured value can be judged as abnormal. The first threshold can be 0.3. When the first confidence level exceeds this threshold, the interface immediately pops up a red "Measured Value Abnormal" alarm and displays the predicted values ​​side by side for the duty officer to adopt with one click or perform secondary verification. By following the order of "first calculating the physical reasonable value, then comparing the sensor readings," this step moves the traditional ex-post experience judgment to online quantitative comparison, which not only avoids the inaction of the extreme value method on non-extreme deviations, but also significantly reduces the workload of manual review, ensuring that scheduling instructions are based on more reliable data.

[0026] When the logical relationship is determined to be a non-strong correlation, the second confidence level is compared with a preset second threshold to determine whether the second confidence level is greater than the second threshold, and a corresponding evaluation result is generated; if yes, it indicates that the measured value is reliable and the evaluation result is normal; if no, it indicates that the measured value is abnormally reliable and the evaluation result is a warning.

[0027] In this embodiment, the obtained values ​​can be refreshed on the interface in real time. The larger the value, the more reliable the current combination. When the value of the second reliability is lower than the second threshold (e.g., 0.5), a yellow prompt can be triggered. By transforming the unformulaic correlation into a quantitative score of "trend + weight", this step fills the gaps that strong correlation algorithms cannot cover, enabling duty officers to quickly identify suspicious data even in the face of complex meteorological interference, and significantly reducing the risk of misjudgment caused by the omission of progressive anomalies due to a single threshold.

[0028] In practical applications, for data where a recommendation value can be calculated, the recommendation value method should be prioritized for determining credibility, as it is more reliable. For data where a recommendation value cannot be calculated, the correlation degree method can be used. Staff can then use the credibility score to make judgments and correct the monitored data values.

[0029] In summary, this method relies on data of related objects and physical change characteristics, using digital metrics to measure reliability in a clear and intuitive manner. It aims to address the reliability assessment of data collected through water conservancy IoT monitoring. This includes judgment criteria and, based on the characteristics of the time series of physical changes in the monitored objects, uses algorithms to provide recommended reliability values ​​to assist staff in making judgments and correcting data.

[0030] In summary, for alarm data, the logical relationship between the monitored object and auxiliary quantities such as upstream inflow, downstream flow, and rainfall is first divided into two categories based on the hydraulic physics mechanism: strongly correlated and uncorrelated. If strongly correlated, the recommended value method is used to calculate the first confidence level, and the ABS difference ratio directly shows the degree of deviation. If the deviation exceeds the threshold, an alarm is immediately triggered and a recommended value is given, which the duty officer can adopt with one click, avoiding the extreme value method from ignoring gradual anomalies. If uncorrelated, the correlation degree method is used, and the second confidence level is obtained by multiplying the trend consistency with the pre-set correlation degree coefficient. The lower the value, the more abnormal it is, so that hidden errors under meteorological interference can also be quantitatively captured. The two methods are linked, with the recommended value method used first, followed by the correlation degree method. Finally, the confidence score and recommended value are pushed to the interface side by side. Staff members judge and correct the collected data accordingly, realizing the transformation from "experience-based review" to "online quantification", significantly reducing the number of on-site verifications, reducing the ineffective workload caused by misjudgment, and improving the data quality and decision reliability of the water conservancy Internet of Things.

[0031] Please refer to Figure 4. A second embodiment of the present invention provides a device for evaluating the reliability of water conservancy Internet of Things (IoT) monitoring data. The device includes: a data acquisition unit 101, used to acquire the current monitoring object to be monitored, determine auxiliary monitoring objects, collect monitoring data of the monitoring object, and analyze the logical relationship between the monitoring object and the auxiliary monitoring object; a strong correlation unit 102, used to acquire measured values, calculate recommended values, and calculate a first reliability based on the measured values ​​and recommended values ​​when the logical relationship is determined to be a strong correlation; a weak correlation unit 103, used to analyze the correlation between the monitoring data of the monitoring object and the auxiliary monitoring object when the logical relationship is determined to be a weak correlation, generate analysis results, and calculate a second reliability based on the correlation coefficient corresponding to the current auxiliary monitoring object and the analysis results; and an evaluation unit 104, used to compare and process the first reliability or the second reliability to generate a corresponding evaluation result.

[0032] A third embodiment of the present invention provides a device for evaluating the reliability of water conservancy Internet of Things (IoT) monitoring data, comprising: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the water conservancy IoT monitoring data reliability evaluation method as described above.

[0033] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for evaluating the reliability of water conservancy Internet of Things (IoT) monitoring data, characterized in that, include: The system acquires the current monitoring target, identifies auxiliary monitoring targets, collects monitoring data of the monitoring target, and analyzes the logical relationship between the monitoring target and the auxiliary monitoring targets. When the logical relationship is determined to be a strong correlation, the measured value is obtained, the recommended value is calculated, and the first credibility is calculated based on the measured value and the recommended value. When it is determined that the logical relationship is not a strong correlation, the correlation between the monitoring data of the monitoring object and the auxiliary monitoring object is analyzed, the analysis results are generated, and the second confidence level is calculated based on the correlation coefficient corresponding to the current auxiliary monitoring object and the analysis results. The first or second confidence level is compared and processed to generate the corresponding evaluation result.

2. The method for evaluating the reliability of water conservancy Internet of Things monitoring data according to claim 1, characterized in that, The process involves: acquiring the current monitoring target, identifying auxiliary monitoring targets, collecting monitoring data from the monitoring targets, and analyzing the logical relationship between the monitoring targets and the auxiliary monitoring targets. Specifically, this includes: acquiring the current monitoring target, converting the related factors of the monitoring target into auxiliary monitoring targets, and collecting monitoring data from the monitoring targets. The related factors include upstream inflow and downstream outflow, and the monitoring data includes reservoir water level and water level fluctuations. Based on preset hydraulic physical mechanism characteristics, the logical relationship between the monitoring targets and the auxiliary monitoring targets is analyzed to determine the logical relationship, which includes strong correlation and weak correlation.

3. The method for evaluating the reliability of water conservancy Internet of Things monitoring data according to claim 1, characterized in that, When the logical relationship is determined to be a strong correlation, the measured value is obtained, the recommended value is calculated, and the first confidence level is calculated based on the measured value and the recommended value. Specifically, when the logical relationship is determined to be a strong correlation, the measured value obtained by the preset sensor is obtained, and the recommended value is calculated based on the strong correlation algorithm corresponding to the strong correlation. The first confidence level is calculated based on the measured value and the recommended value. The formula is: First confidence level = ABS(recommended value - measured value) / recommended value, where ABS is the absolute value.

4. The method for evaluating the reliability of water conservancy Internet of Things monitoring data according to claim 1, characterized in that, When the logical relationship is determined to be a weak correlation, the correlation between the monitoring data of the monitored object and the auxiliary monitoring object is analyzed to generate an analysis result. Based on the correlation coefficient corresponding to the current auxiliary monitoring object and the analysis result, a second confidence level is calculated. Specifically: when the logical relationship is determined to be a weak correlation, the correlation between the monitoring data of the monitored object and the auxiliary monitoring object is analyzed to determine whether the changing trends of the auxiliary monitoring object and the monitoring data are consistent, generating an analysis result. Wherein, if the changing trends of the auxiliary monitoring object and the monitoring data are consistent, it is considered true, and the value of the analysis result is 1; if the changing trends of the auxiliary monitoring object and the monitoring data are inconsistent, it is considered false, and the value of the analysis result is 0. The correlation coefficient corresponding to the current auxiliary monitoring object is obtained, and combined with the analysis result, the second confidence level is calculated using the formula: Second Confidence Level = Where n is the number of auxiliary monitoring objects, Let be the correlation coefficient of the i-th auxiliary monitoring object, which is assigned according to the strength of the correlation.

5. The method for evaluating the reliability of water conservancy Internet of Things monitoring data according to claim 1, characterized in that, The first confidence level or the second confidence level is compared and processed to generate a corresponding evaluation result. Specifically, when the logical relationship is determined to be a strong correlation, the first confidence level is compared with a preset first threshold to determine whether the first confidence level is less than the first threshold and generate a corresponding evaluation result. If yes, it indicates that the measured value is reliable and the evaluation result is normal. If no, it indicates that the measured value is abnormal and the evaluation result is a warning.

6. The method for evaluating the reliability of water conservancy Internet of Things monitoring data according to claim 5, characterized in that, Also includes: When the logical relationship is determined to be a non-strong correlation, the second confidence level is compared with a preset second threshold to determine whether the second confidence level is greater than the second threshold, and a corresponding evaluation result is generated; if yes, it indicates that the measured value is reliable and the evaluation result is normal; if no, it indicates that the measured value is abnormally reliable and the evaluation result is a warning.

7. A device for evaluating the reliability of water conservancy Internet of Things (IoT) monitoring data, characterized in that, include: The data acquisition unit is used to acquire the current monitoring object to be monitored, determine the auxiliary monitoring object, collect the monitoring data of the monitoring object, and analyze the logical relationship between the monitoring object and the auxiliary monitoring object; A strong correlation unit is used to obtain a measured value, calculate a recommended value, and calculate a first credibility based on the measured value and the recommended value when the logical relationship is determined to be a strong correlation. The non-strong correlation unit is used to analyze the correlation between the monitoring data of the monitoring object and the auxiliary monitoring object when the logical relationship is determined to be non-strong correlation, generate analysis results, and calculate the second confidence level based on the correlation coefficient corresponding to the current auxiliary monitoring object and the analysis results. The evaluation unit is used to compare and process the first confidence level or the second confidence level, and generate the corresponding evaluation result.

8. A device for evaluating the reliability of water conservancy Internet of Things (IoT) monitoring data, characterized in that, The system includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the water conservancy Internet of Things monitoring data reliability assessment method as described in any one of claims 1 to 6.

Images