Underground water dynamic monitoring and early warning system based on Internet of Things technology

Abstract

The invention relates to the technical field of data processing, in particular to an underground water dynamic monitoring and early warning system based on the Internet of Things technology, which comprises a memory, a processor and a computer program stored in the memory and running on the processor, when executing the computer program, the processor implements the following steps: respectively performing STL time sequence decomposition on time sequence data of any two monitoring indexes in a current period to obtain a corresponding trend item, a corresponding season item and a corresponding residual item; obtaining the similarity degree between any two monitoring indexes according to the similarity of the corresponding trend item, season item and residual item between any two monitoring indexes; all the monitoring indexes are classified according to the similarity degree between every two monitoring indexes of the underground water, then the early warning threshold value of each monitoring index in the next period is set according to the classification result, dynamic abnormal early warning is carried out on the underground water, and the accuracy of monitoring and early warning on the underground water is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a groundwater dynamic monitoring and early warning system based on Internet of Things technology. Background Art

[0002] Groundwater refers to the water stored in the rock voids below the ground surface. Narrowly speaking, it refers to the water in the saturated aquifer below the groundwater table. Groundwater is an important part of water resources. Due to its stable water volume and good water quality, it is one of the important water sources for agricultural irrigation, industrial and mining, and cities. However, under certain conditions, changes in groundwater can also cause adverse natural phenomena such as waterlogging, salinization, landslides, and land subsidence. Therefore, it is necessary to monitor the changes in groundwater in real time.

[0003] Under the traditional method, various sensors are used to monitor groundwater monitoring indicators such as water level, water temperature, conductivity, and pH value. The sensor data is transmitted to a remote data center, and the real-time data is analyzed. Fixed thresholds for each monitoring indicator are set according to the historical data characteristics of each monitoring indicator, and then early warnings are made by judging whether these monitoring indicators exceed the fixed thresholds. However, there are many monitoring indicators for groundwater, and for a sudden and changing environment (such as sudden rainfall or pollution), the data change trends of each monitoring indicator are different, and the fixed thresholds cannot make accurate monitoring and early warnings for the sudden and changing environment.

[0004] Therefore, how to set dynamic thresholds for each monitoring indicator of groundwater to improve the accuracy of groundwater monitoring and early warning has become an urgent problem to be solved. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a groundwater dynamic monitoring and early warning system based on Internet of Things technology to solve the problem of how to set dynamic thresholds for each monitoring indicator of groundwater to improve the accuracy of groundwater monitoring and early warning.

[0006] An embodiment of the present invention provides a groundwater dynamic monitoring and early warning system based on Internet of Things technology, including a memory, a processor, and a computer program stored in the memory and running on the processor. The processor, when executing the computer program, implements the following steps: During the process of monitoring the situation of groundwater, obtain the time series data of all monitoring indicators in the current period. For any two monitoring indicators, perform STL time series decomposition on the time series data of the any two monitoring indicators in the current period to obtain the corresponding trend term, seasonal term, and residual term; Divide the residual terms corresponding to any two of the monitoring indicators into a preset number of residual intervals respectively. According to the outliers and chaos degrees in each residual interval corresponding to any two of the monitoring indicators, as well as the similarity of the trend terms and the similarity of the seasonal terms corresponding to any two of the monitoring indicators, obtain the similarity degree between any two of the monitoring indicators in the current period; Obtain the similarity degree between every two of the monitoring indicators of the groundwater in the current period. Classify all the monitoring indicators of the groundwater according to all the similarity degrees to obtain the classification result of all the monitoring indicators in the current period. According to the classification result and the time series data of all the monitoring indicators in the current period, set the warning threshold of each monitoring indicator in the next period respectively. Conduct dynamic anomaly warning on the groundwater according to the warning threshold of the next period.

[0007] Preferably, the obtaining the similarity degree between any two of the monitoring indicators in the current period according to the outliers and chaos degrees in each residual interval corresponding to any two of the monitoring indicators, as well as the similarity of the trend terms and the similarity of the seasonal terms corresponding to any two of the monitoring indicators includes: According to the chaos degree of each residual interval corresponding to any two of the monitoring indicators, obtain the residual fluctuation similarity index between any two of the monitoring indicators in the current period; By using box plots, obtain the outliers in the residual terms corresponding to any two of the monitoring indicators respectively. According to the distribution of the outliers in the residual terms corresponding to any two of the monitoring indicators, obtain the residual distribution similarity index between any two of the monitoring indicators in the current period; According to the similarity of the trend terms and the similarity of the seasonal terms corresponding to any two of the monitoring indicators, obtain the data overall consistency index between any two of the monitoring indicators in the current period; Perform weighted summation on the residual fluctuation similarity index, the residual distribution similarity index, and the data overall consistency index to obtain the similarity degree between any two of the monitoring indicators in the current period.

[0008] Preferably, the obtaining the residual fluctuation similarity index between any two of the monitoring indicators in the current period according to the chaos degree of each residual interval corresponding to any two of the monitoring indicators includes: Denote any one of the two monitoring indicators as the first monitoring indicator and the other monitoring indicator as the second monitoring indicator. Round the information entropy of each residual interval corresponding to the first monitoring indicator to obtain the target information entropy. Form a first residual information entropy sequence with all the target information entropies, and obtain the second residual information entropy sequence corresponding to the second monitoring indicator; Construct a line chart based on the first residual information entropy sequence. Among them, the ordinate of the line chart is the target information entropy, and the abscissa is the residual interval. Obtain the slope between every two adjacent data in the line chart to obtain the slope sequence corresponding to the first residual information entropy sequence, denoted as the first slope sequence. Obtain the second slope sequence corresponding to the second residual information entropy sequence, calculate the absolute value of the first difference between the two data at the same position in the first slope sequence and the second slope sequence, and perform linear normalization on the reciprocal of the average value of all the absolute values of the first differences to obtain the first variable; In the first residual information entropy sequence and the second residual information entropy sequence, calculate the absolute value of the second difference between the two data at the same position, and perform linear normalization on the reciprocal of the average value of all the absolute values of the second differences to obtain the second variable; Calculate the average value between the first variable and the second variable to obtain the residual fluctuation similarity index between any two monitoring indicators in the current period.

[0009] Preferably, obtaining the residual distribution similarity index between any two monitoring indicators in the current period according to the distribution of outliers in the residual terms corresponding to any two monitoring indicators includes: Form a first outlier sequence with the outliers in the residual term corresponding to the first monitoring indicator, and form a second outlier sequence with the outliers in the residual term corresponding to the second monitoring indicator; Calculate the absolute value of the third difference between the number of all elements in the first outlier sequence and the number of all elements in the second outlier sequence, and perform linear normalization on the reciprocal of the sum of the constant 1 and the absolute value of the third difference to obtain the third variable; According to the sampling time corresponding to each data in the first outlier sequence, calculate the time interval between every two adjacent data and accumulate them to obtain the first outlier distribution characteristic value of the first residual term. According to the sampling time corresponding to each data in the second outlier sequence, obtain the second outlier distribution characteristic value of the second residual term. Calculate the absolute value of the fourth difference between the first outlier distribution characteristic value and the second outlier distribution characteristic value, and perform linear normalization on the reciprocal of the absolute value of the fourth difference to obtain the fourth variable; Calculate the average value of all data except outliers in the residual term corresponding to the first monitoring indicator, denoted as the first average value. Calculate the average value of all data except outliers in the residual term corresponding to the second monitoring indicator, denoted as the second average value. Calculate the absolute value of the fifth difference between the first average value and the second average value, and perform linear normalization on the reciprocal of the sum of the constant 1 and the absolute value of the fifth difference to obtain the fifth variable; Calculate the average value among the third variable, the fourth variable, and the fifth variable to obtain the residual distribution similarity index between any two monitoring indicators during the current period.

[0010] Preferably, obtaining the overall data consistency index between any two monitoring indicators during the current period according to the similarity of the trend terms and the similarity of the seasonal terms corresponding to the any two monitoring indicators includes: Calculate the DTW distance between the trend terms corresponding to any two monitoring indicators, denoted as the first DTW distance, and perform linear normalization on the reciprocal of the sum of the constant 1 and the first DTW distance to obtain the sixth variable; Calculate the DTW distance between the seasonal terms corresponding to any two monitoring indicators, denoted as the second DTW distance, and perform linear normalization on the reciprocal of the sum of the constant 1 and the second DTW distance to obtain the seventh variable; Calculate the average value between the sixth variable and the seventh variable to obtain the overall data consistency index between any two monitoring indicators during the current period.

[0011] Preferably, classifying all the monitoring indicators of the groundwater according to all the similarity degrees to obtain the classification result of all the monitoring indicators during the current period includes: For the similarity degree between any two monitoring indicators, if the similarity degree between the two monitoring indicators is greater than the preset similarity degree threshold, then the two monitoring indicators are regarded as the same category.

[0012] Preferably, setting the warning threshold for each monitoring indicator in the next period according to the classification result and the time series data of all the monitoring indicators during the current period includes: For any category in the classification result, calculate the average value of the time series data of all the monitoring indicators in the any category during the current period, denoted as the target average value, and use the target average value as the warning threshold for all the monitoring indicators in the any category in the next period.

[0013] The beneficial effects of the embodiments of the present invention compared with the prior art are: In the process of monitoring the groundwater situation, the present invention obtains the time-series data of all monitoring indicators in the current period. For any two monitoring indicators, the time-series data of the two monitoring indicators in the current period are respectively subjected to STL time-series decomposition to obtain the corresponding trend term, seasonal term, and residual term. The residual terms corresponding to the two monitoring indicators are respectively divided into a preset number of residual intervals. According to the outliers and chaos degree in each residual interval corresponding to the two monitoring indicators, as well as the similarity of the trend terms and the similarity of the seasonal terms corresponding to the two monitoring indicators, the similarity degree between the two monitoring indicators in the current period is obtained. The similarity degrees between every two monitoring indicators of the groundwater in the current period are obtained. According to all the similarity degrees, all the monitoring indicators of the groundwater are classified to obtain the classification result of all the monitoring indicators in the current period. According to the classification result and the time-series data of all the monitoring indicators in the current period, the warning threshold of each monitoring indicator in the next period is respectively set, and the groundwater is dynamically and abnormally warned according to the warning threshold of the next period. Among them, obtaining the time-series data of each monitoring indicator in the current period makes the subsequently set warning threshold pay more attention to the short-term data fluctuation characteristics and improves the accuracy of monitoring and warning the groundwater. Further, considering the potential similarity between the detection indicators, the time-series data of each monitoring indicator in the current period are decomposed by STL to analyze the similarity characteristics of each decomposition item (trend item, residual item, and seasonal item) corresponding to each monitoring indicator, so as to obtain the similarity degree between each monitoring indicator in the current period. Then, each monitoring indicator is classified, and the warning threshold is set for the monitoring indicators in each category according to the classification result to monitor and warn the groundwater situation in the next period, which improves the accuracy of monitoring the groundwater and greatly reduces the singularity and workload of setting thresholds for each monitoring indicator in the traditional method, and improves the efficiency of monitoring the groundwater. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0015] Figure 1 It is a flowchart of a method for dynamically monitoring and warning groundwater based on Internet of Things technology provided in Embodiment 1 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] Embodiments of the present disclosure will be described in detail below, and examples of the embodiments are shown in the accompanying drawings. The embodiments described below by referring to the accompanying drawings are exemplary and are intended to explain the present disclosure, and should not be construed as limiting the present disclosure.

[0017] It should be noted that the terms "first", "second", etc. in the specification of the present disclosure and the above-mentioned accompanying drawings are used to distinguish similar objects and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances so that the embodiments of the present disclosure described here can be implemented in an order different from those illustrated or described here. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present disclosure.

[0018] In order to illustrate the technical solution of the present invention, specific embodiments will be used for illustration below.

[0019] The specific scenario targeted by the present invention is as follows: During the process of real-time monitoring of the changes in groundwater, fixed thresholds are set respectively according to the historical data characteristics of various monitoring indicators of groundwater, and then early warnings are made by judging whether these monitoring indicators exceed the fixed thresholds. However, there are many monitoring indicators for groundwater, and in a sudden and changeable environment (such as sudden rainfall or pollution), the data change trends of various monitoring indicators are different, and the fixed thresholds cannot make accurate monitoring and early warnings for the sudden and changeable environment. Therefore, the present invention sets the early warning thresholds of various monitoring indicators in the next cycle according to the data fluctuation characteristics between various monitoring indicators in the current cycle, so as to improve the accuracy of monitoring and early warning of groundwater.

[0020] An embodiment of the present invention provides a groundwater dynamic monitoring and early warning system based on Internet of Things technology, including a processor and a memory. The processor executes the computer program stored in the memory to implement a groundwater dynamic monitoring and early warning method based on Internet of Things technology, as Figure 1 shown, this groundwater dynamic monitoring and early warning method based on Internet of Things technology includes the following steps: Step S101, during the process of monitoring the groundwater situation, obtain the time series data of all monitoring indicators in the current cycle. For any two monitoring indicators, perform STL time series decomposition on the time series data of the any two monitoring indicators in the current cycle respectively to obtain the corresponding trend term, seasonal term, and residual term.

[0021] In the process of monitoring groundwater under the traditional method, the whole process of groundwater monitoring and early warning is carried out by setting fixed thresholds for each monitoring index. However, for a sudden and changeable environment (such as sudden rainfall or pollution), the data change trends of each monitoring index are different, and the fixed threshold cannot be adjusted for the sudden and changeable environment, thus making it impossible to achieve accurate monitoring and early warning. Therefore, in the embodiments of the present invention, the duration of each period is set to 10 minutes, and the sampling frequency is 1 time per second. There is no limitation here, and the implementer can set it according to the specific scenario. The period containing the current moment is used as the current period, and through various sensors, the data of each monitoring index (such as water level, water temperature, conductivity, pH value, etc.) of groundwater in the current period are obtained in real time. Then, the early warning thresholds of each monitoring index are set according to the data in the current period, so that the early warning thresholds pay more attention to the short-term data fluctuation characteristics, in order to warn about the situation of groundwater in the next period, realize the dynamic monitoring of groundwater, and improve the accuracy of groundwater monitoring and early warning.

[0022] Since the dimensions of the data corresponding to each monitoring index are different, for the convenience of subsequent analysis, the data of each monitoring index obtained within 10 minutes are linearly normalized to obtain the time series data of each monitoring index in the current period. Then, the early warning thresholds of each monitoring index are set according to the time series data in the current period, in order to warn about the situation of groundwater in the next period.

[0023] For the establishment of the early warning thresholds of each monitoring index of groundwater, under the traditional method, fixed thresholds are set respectively according to the historical data characteristics of each monitoring index. However, there are many monitoring indexes of groundwater, and analyzing the historical data characteristics of each monitoring index and setting the early warning thresholds greatly increases the workload. Considering the potential similarity in the data changes among the monitoring indexes of groundwater, therefore, the monitoring indexes can be classified according to the similarity among them, and then the early warning thresholds are set for each monitoring index according to the classification results, so as to reduce the singularity and workload of setting the early warning thresholds for each monitoring index in the traditional method.

[0024] Considering that STL time series decomposition can be used to reflect the different components of time series data, as well as the basic structure and law of time series data, therefore, the time series data of each monitoring index in the current period can be decomposed by STL time series decomposition to obtain the corresponding trend term, seasonal term and residual term. By judging the similarity among the trend term, seasonal term and residual term corresponding to each monitoring index, the monitoring indexes of groundwater can be classified.

[0025] In an embodiment of the present invention, taking the ith monitoring index and the jth monitoring index of groundwater as examples, STL time series decomposition is respectively performed on the time series data of the ith monitoring index and the jth monitoring index in the current period to obtain the corresponding trend term, seasonal term, and residual term. Since STL time series decomposition is a prior art, it will not be elaborated here.

[0026] Step S102: Divide the residual terms corresponding to any two monitoring indexes into a preset number of residual intervals respectively. According to the outliers and chaos degrees in each residual interval corresponding to the two monitoring indexes, as well as the similarity of the trend terms and the similarity of the seasonal terms corresponding to the two monitoring indexes, obtain the similarity degree between the two monitoring indexes in the current period.

[0027] Since the residual terms obtained after STL decomposition of time series data represent the remaining part of the time series data after removing the seasonal term and trend term components, usually manifested as random fluctuations or noises. If the similarity between the residual terms of the ith monitoring index and the jth monitoring index of groundwater is relatively high, it means that after removing the influence of the seasonal term and trend term, these two monitoring indexes still have similar fluctuation characteristics, further indicating that these two monitoring indexes may be interfered by the same external factors or have some internal connection. At the same time, the similarity of the trend terms and seasonal terms between the ith monitoring index and the jth monitoring index is easier to judge than the similarity of the residual terms, and the similarity of the trend terms and seasonal terms can be used to judge the similarity of the ith monitoring index and the jth monitoring index in the overall direction, but it is more difficult to reflect the internal connection between the ith monitoring index and the jth monitoring index. Therefore, taking the similarity between the residual terms corresponding to the ith monitoring index and the jth monitoring index of groundwater as the main factor and the similarity between the trend term and the seasonal term as the auxiliary factor, obtain the similarity degree between the ith monitoring index and the jth monitoring index in the current period, which is used to judge whether there is potential similarity in the data changes between the ith monitoring index and the jth monitoring index of groundwater in the current period, and further judge whether the ith monitoring index and the jth monitoring index of groundwater are of the same category in the current period. The specific steps to obtain the similarity degree are as follows: (1) Obtain the residual fluctuation similarity index between the ith monitoring index and the jth monitoring index in the current period according to the chaos degree of the residual terms corresponding to the ith monitoring index and the jth monitoring index of groundwater.

[0028] In an embodiment of the present invention, first, the residual terms corresponding to the i-th monitoring index and the j-th monitoring index are respectively divided into 10 intervals, and each interval is denoted as a residual interval. There is no limitation here, and the implementer can set the number of intervals according to the specific scenario. Then, the information entropy of each residual interval is calculated. Since the change in information entropy is small, all the information entropies are rounded to obtain the target information entropy to amplify the similarity between each residual interval. Finally, according to the target information entropy of each residual interval corresponding to the i-th monitoring index and the j-th monitoring index, the residual fluctuation similarity index between the i-th monitoring index and the j-th monitoring index in the current period is obtained, which is used to characterize the similarity of the data fluctuations of the time series data of the i-th monitoring index and the j-th monitoring index in the current period. Specifically: The target information entropies of each residual interval corresponding to the i-th monitoring index form a first residual information entropy sequence, and the target information entropies of each residual interval corresponding to the j-th monitoring index form a second residual information entropy sequence; Construct a line chart according to the first residual information entropy sequence, where the ordinate of the line chart is the target information entropy and the abscissa is the residual interval. Obtain the slope between every two adjacent data in the line chart to get the slope sequence corresponding to the first residual information entropy sequence, denoted as the first slope sequence. Obtain the second slope sequence corresponding to the second residual information entropy sequence, calculate the absolute value of the first difference between the two data at the same position in the first slope sequence and the second slope sequence, and perform linear normalization on the reciprocal of the average value of all the absolute values of the first differences to obtain a first variable; In the first residual information entropy sequence and the second residual information entropy sequence, calculate the absolute value of the second difference between the two data at the same position, and perform linear normalization on the reciprocal of the average value of all the absolute values of the second differences to obtain a second variable; Calculate the average value between the first variable and the second variable to obtain the residual fluctuation similarity index between the i-th monitoring index and the j-th monitoring index in the current period.

[0029] In an embodiment, the calculation formula for the residual fluctuation similarity index between the i-th monitoring index and the j-th monitoring index in the current period is:

[0030] Wherein, represents the residual fluctuation similarity index between the i-th monitoring index and the j-th monitoring index in the current period, represents the m-th data in the first slope sequence, represents the m-th data in the second slope sequence, Denote the nth data in the target information entropy (the first residual information entropy sequence) corresponding to each residual interval of the ith monitoring index. Denote the nth data in the target information entropy (the first residual information entropy sequence) corresponding to each residual interval of the jth monitoring index. M represents the number of all data in the first slope sequence or the second slope sequence, and N represents the number of all data in the first residual information entropy sequence or the second residual information entropy sequence. Denote the absolute value symbol. Denote the linear normalization function.

[0031] It should be noted that The smaller the value of, the more similar the change trends of the target information entropy of the residual intervals corresponding to the ith monitoring index and the jth monitoring index are. Furthermore The larger the value of, the higher the similarity of the fluctuations of the trend terms corresponding to the ith monitoring index and the jth monitoring index of the groundwater is. The smaller the value of, the more similar the degrees of chaos of the residual intervals corresponding to the ith monitoring index and the jth monitoring index are. Furthermore The larger the value of, the higher the similarity of the fluctuations of the trend terms corresponding to the ith monitoring index and the jth monitoring index of the groundwater is.

[0032] (2) According to the distribution of outliers in the residual terms corresponding to the ith monitoring index and the jth monitoring index of the groundwater, obtain the residual distribution similarity index between the ith monitoring index and the jth monitoring index in the current period.

[0033] Considering that the time-series data of each monitoring index of the groundwater may be affected by the same external factors, in the embodiments of the present invention, first, through box plots, outliers are respectively screened out in the residual terms corresponding to the ith monitoring index and the jth monitoring index, and the situation of being affected by external factors in the time-series data is shown by the outliers. Among them, the box plot is a prior art and will not be elaborated here. Then, according to the distribution of outliers in the residual terms, obtain the residual distribution similarity index between the ith monitoring index and the jth monitoring index in the current period, which is used to characterize the probability that the time-series data of the ith monitoring index and the jth monitoring index are affected by the same external factors in the current period. The larger the residual distribution similarity index is, the more similar the distributions of the residual terms corresponding to the ith monitoring index and the jth monitoring index are, the greater the probability that the time-series data of the ith monitoring index and the jth monitoring index are affected by the same external factors in the current period, and the more potentially similar the data changes between the ith monitoring index and the jth monitoring index in the current period are.

[0034] Then the specific way to obtain the residual distribution similarity index is as follows: The outliers in the residual terms corresponding to the \(i\)-th monitoring index are grouped into a first outlier sequence, and the outliers in the residual terms corresponding to the \(j\)-th monitoring index are grouped into a second outlier sequence; Calculate the absolute value of the third difference between the number of all elements in the first outlier sequence and the number of all elements in the second outlier sequence, and perform linear normalization on the reciprocal of the sum of the constant 1 and the absolute value of the third difference to obtain a third variable; According to the sampling time corresponding to each data in the first outlier sequence, calculate the time interval between every two adjacent data and accumulate them to obtain the first outlier distribution characteristic value of the first residual term. According to the sampling time corresponding to each data in the second outlier sequence, obtain the second outlier distribution characteristic value of the second residual term. Calculate the absolute value of the fourth difference between the first outlier distribution characteristic value and the second outlier distribution characteristic value, and perform linear normalization on the reciprocal of the absolute value of the fourth difference to obtain a fourth variable; Calculate the average value of all data except outliers in the residual term corresponding to the \(i\)-th monitoring index, denoted as the first average value. Calculate the average value of all data except outliers in the residual term corresponding to the \(j\)-th monitoring index, denoted as the second average value. Calculate the absolute value of the fifth difference between the first average value and the second average value, and perform linear normalization on the reciprocal of the sum of the constant 1 and the absolute value of the fifth difference to obtain a fifth variable; Calculate the average value among the third variable, the fourth variable, and the fifth variable to obtain the residual distribution similarity index between the \(i\)-th monitoring index and the \(j\)-th monitoring index in the current period.

[0035] In an embodiment, the calculation formula for the residual distribution similarity index between the \(i\)-th monitoring index and the \(j\)-th monitoring index in the current period is:

[0036] where, represents the residual distribution similarity index between the \(i\)-th monitoring index and the \(j\)-th monitoring index in the current period, represents the number of all elements in the first outlier sequence, represents the number of all elements in the second outlier sequence, represents the time interval between the \(e\)-th data and its right adjacent data in the first outlier sequence composed of outliers in the residual term corresponding to the \(i\)-th monitoring index, represents the time interval between the \(f\)-th data and its right adjacent data in the second outlier sequence composed of outliers in the residual term corresponding to the \(j\)-th monitoring index, where \(f\) is the number of the data, represents the average value (the first average value) of all data except the outliers in the residual term corresponding to the i-th monitoring index, represents the average value (the second average value) of all data except the outliers in the residual term corresponding to the j-th monitoring index, and 1 represents a constant, represents the absolute value symbol, represents the linear normalization function.

[0037] It should be noted that, the smaller it is, the closer the number of outliers in the residual terms corresponding to the i-th monitoring index and the j-th monitoring index is, and the more similar the situations of the i-th monitoring index and the j-th monitoring index being abnormally interfered are. Furthermore, the larger it is, the more similar the distributions of the residual terms corresponding to the i-th monitoring index and the j-th monitoring index of groundwater are; the smaller it is, the more similar the distributions of the outliers in the residual terms corresponding to the i-th monitoring index and the j-th monitoring index are. Furthermore, the larger it is, the more similar the distributions of the residual terms corresponding to the i-th monitoring index and the j-th monitoring index of groundwater are; the smaller it is, the more similar the distributions of the normal values after removing the outliers in the residual terms corresponding to the i-th monitoring index and the j-th monitoring index are. Furthermore, the larger it is, the more similar the distributions of the residual terms corresponding to the i-th monitoring index and the j-th monitoring index of groundwater are.

[0038] (3) According to the trend terms and seasonal terms corresponding to the i-th monitoring index and the j-th monitoring index of groundwater, obtain the overall data consistency index between the i-th monitoring index and the j-th monitoring index in the current period.

[0039] Considering that the DTW (Dynamic Time Warping) algorithm is an algorithm for calculating the similarity of two time series, and it measures the similarity between them by non-linearly aligning the time series. Therefore, use the DTW algorithm to obtain the DTW distance between the trend terms corresponding to the i-th monitoring index and the j-th monitoring index of groundwater, and the DTW distance between the seasonal terms, and then obtain the overall data consistency index between the i-th monitoring index and the j-th monitoring index in the current period, which is used to characterize the similarity between the i-th monitoring index and the j-th monitoring index in the overall direction. Specifically: Calculate the DTW distance between the trend terms corresponding to the i-th monitoring index and the j-th monitoring index, denoted as the first DTW distance, and perform linear normalization on the reciprocal of the sum of the constant 1 and the first DTW distance to obtain the sixth variable; Calculate the DTW distance between the seasonal terms corresponding to the \(i\)-th monitoring indicator and the \(j\)-th monitoring indicator, denoted as the second DTW distance, and perform linear normalization on the reciprocal of the sum between the constant 1 and the second DTW distance to obtain the seventh variable; Calculate the average value between the sixth variable and the seventh variable to obtain the overall data consistency index between any two monitoring indicators in the current period. Among them, the DTW distance is a prior art and will not be elaborated here.

[0040] In an implementation manner, the calculation formula for the overall data consistency index between the \(i\)-th monitoring indicator and the \(j\)-th monitoring indicator in the current period is:

[0041] Among them, represents the overall data consistency index between the \(i\)-th monitoring indicator and the \(j\)-th monitoring indicator in the current period, represents the trend term corresponding to the \(i\)-th monitoring indicator, represents the trend term corresponding to the \(j\)-th monitoring indicator, represents the DTW distance (the first DTW distance) between the trend terms corresponding to the \(i\)-th monitoring indicator and the \(j\)-th monitoring indicator, represents the seasonal term corresponding to the \(i\)-th monitoring indicator, represents the seasonal term corresponding to the \(j\)-th monitoring indicator, represents the DTW distance (the second DTW distance) between the seasonal terms corresponding to the \(i\)-th monitoring indicator and the \(j\)-th monitoring indicator, 1 represents a constant, represents the linear normalization function.

[0042] It should be noted that the smaller, the smaller the difference between the trend terms corresponding to the \(i\)-th monitoring indicator and the \(j\)-th monitoring indicator, and thus the larger, the greater the similarity of the data of the \(i\)-th monitoring indicator and the \(j\)-th monitoring indicator of the groundwater in the overall direction in the current period; the smaller, the smaller the difference between the seasonal terms corresponding to the \(i\)-th monitoring indicator and the \(j\)-th monitoring indicator, and thus the larger, the greater the similarity of the data of the \(i\)-th monitoring indicator and the \(j\)-th monitoring indicator of the groundwater in the overall direction in the current period.

[0043] (4) Combine the residual fluctuation similarity index, the residual distribution similarity index, and the overall data consistency index between the \(i\)-th monitoring indicator and the \(j\)-th monitoring indicator of the groundwater in the current period to obtain the similarity degree between the \(i\)-th monitoring indicator and the \(j\)-th monitoring indicator in the current period.

[0044] Specifically, a weighted sum of the residual fluctuation similarity index, the residual distribution similarity index, and the overall data consistency index is calculated to obtain the similarity degree between any two monitoring indicators in the current period.

[0045] In one embodiment, the calculation formula for the similarity degree between the i-th monitoring indicator and the j-th monitoring indicator in the current period is:

[0046] Wherein, represents the similarity degree between the i-th monitoring indicator and the j-th monitoring indicator in the current period, represents the first weight, represents the residual fluctuation similarity index between the i-th monitoring indicator and the j-th monitoring indicator in the current period, represents the second weight, represents the residual distribution similarity index between the i-th monitoring indicator and the j-th monitoring indicator in the current period, represents the third weight, represents the overall data consistency index between the i-th monitoring indicator and the j-th monitoring indicator in the current period.

[0047] It should be noted that setting , , there is no restriction here, and the implementer can set it according to the specific scenario. The larger is, the higher the fluctuation similarity of the trend terms corresponding to the i-th monitoring indicator and the j-th monitoring indicator, and thus The larger is, the more similar the time series data of the i-th monitoring indicator and the j-th monitoring indicator of groundwater in the current period; the larger is, the more similar the distributions of the residual terms corresponding to the i-th monitoring indicator and the j-th monitoring indicator, and thus

[0048] The larger

[0049] is, the more similar the time series data of the i-th monitoring indicator and the j-th monitoring indicator of groundwater in the current period; the larger is, the greater the similarity of the data of the i-th monitoring indicator and the j-th monitoring indicator in the overall direction in the current period, and thus is, the more similar the time series data of the i-th monitoring indicator and the j-th monitoring indicator of groundwater in the current period. Thus, the similarity degree between the i-th monitoring indicator and the j-th monitoring indicator of groundwater in the current period is obtained.

[0049] Step S103: Obtain the similarity degree between every two monitoring indicators of the groundwater during the current period. Classify all the monitoring indicators of the groundwater according to all the similarity degrees to obtain the classification result of all the monitoring indicators during the current period. According to the classification result and the time series data of all the monitoring indicators during the current period, set the warning threshold for each monitoring indicator in the next period respectively. Conduct dynamic anomaly warning on the groundwater according to the warning threshold of the next period.

[0050] Set the preset similarity degree threshold to 0.7. There is no limit here, and the implementer can set it according to the specific scenario. If the similarity degree between the i-th monitoring indicator and the j-th monitoring indicator of the groundwater during the current period , it indicates that the similarity between the i-th monitoring indicator and the j-th monitoring indicator is relatively high during the current period, and they can be classified into the same category. Similarly, according to Step S102, obtain the similarity degree between every two monitoring indicators of the groundwater during the current period, and then classify all the monitoring indicators of the groundwater according to all the similarity degrees. Taking the T-th category as an example, use the average value or the maximum value of the time series data of all the monitoring indicators in the T-th category during the current period as the warning threshold of all the monitoring indicators in the T-th category in the next period. There is no limit here, and the implementer can set it according to the specific scenario to realize dynamic warning of the groundwater situation and improve the accuracy of monitoring and warning of the groundwater.

[0051] In summary, during the process of monitoring the groundwater condition, the invention obtains the time series data of all monitoring indicators in the current period. For any two monitoring indicators, the time series data of the two monitoring indicators in the current period are respectively subjected to STL time series decomposition to obtain the corresponding trend term, seasonal term, and residual term; the residual terms corresponding to the two monitoring indicators are respectively divided into a preset number of residual intervals, and according to the outliers and chaos degree in each residual interval corresponding to the two monitoring indicators, as well as the similarity of the trend terms and the similarity of the seasonal terms corresponding to the two monitoring indicators, the similarity degree between the two monitoring indicators in the current period is obtained; the similarity degree between every two monitoring indicators of the groundwater in the current period is obtained, and all monitoring indicators of the groundwater are classified according to all similarity degrees to obtain the classification result of all monitoring indicators in the current period. According to the classification result and the time series data of all monitoring indicators in the current period, the warning threshold of each monitoring indicator in the next period is set respectively, and dynamic anomaly warning of the groundwater is carried out according to the warning threshold of the next period. Among them, obtaining the time series data of each monitoring indicator in the current period makes the subsequent set warning threshold pay more attention to the short-term data fluctuation characteristics and improves the accuracy of monitoring and warning of groundwater; further, considering the potential similarity between the detection indicators, the time series data of each monitoring indicator in the current period are decomposed by STL to analyze the similarity characteristics of each decomposition item (trend item, residual item, and seasonal item) corresponding to each monitoring indicator, so as to obtain the similarity degree between each monitoring indicator in the current period, and then classify each monitoring indicator, and set the warning threshold for the monitoring indicators in each category according to the classification result to monitor and warn the groundwater condition in the next period, which improves the accuracy of monitoring groundwater, and at the same time greatly reduces the singularity and workload of setting thresholds for each monitoring indicator in the traditional method, and improves the efficiency of monitoring groundwater.

[0052] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. The groundwater dynamic monitoring and early warning system based on Internet of Things technology is characterized in that Comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that when the processor executes the computer program, the following steps are implemented: During the process of monitoring the groundwater situation, obtain the time-series data of all monitoring indicators in the current period. For any two monitoring indicators, perform STL time-series decomposition on the time-series data of the two monitoring indicators in the current period respectively to obtain the corresponding trend term, seasonal term, and residual term; Divide the residual terms corresponding to the two monitoring indicators into a preset number of residual intervals respectively. According to the outliers and chaos degrees in each residual interval corresponding to the two monitoring indicators, as well as the similarity of the trend terms and the similarity of the seasonal terms corresponding to the two monitoring indicators, obtain the similarity degree between the two monitoring indicators in the current period; Obtain the similarity degree between every two monitoring indicators of the groundwater in the current period. Classify all the monitoring indicators of the groundwater according to all the similarity degrees to obtain the classification result of all the monitoring indicators in the current period. According to the classification result and the time-series data of all the monitoring indicators in the current period, set the warning threshold for each monitoring indicator in the next period respectively. Perform dynamic anomaly warning on the groundwater according to the warning threshold in the next period.

2. The groundwater dynamic monitoring and early warning system based on the Internet of Things technology according to claim 1, characterized in that The obtaining of the similarity degree between the two monitoring indicators in the current period according to the outliers and chaos degrees in each residual interval corresponding to the two monitoring indicators, as well as the similarity of the trend terms and the similarity of the seasonal terms corresponding to the two monitoring indicators, includes: According to the chaos degree of each residual interval corresponding to the two monitoring indicators, obtain the residual fluctuation similarity index between the two monitoring indicators in the current period; Through a box plot, obtain the outliers in the residual terms corresponding to the two monitoring indicators respectively. According to the distribution of the outliers in the residual terms corresponding to the two monitoring indicators, obtain the residual distribution similarity index between the two monitoring indicators in the current period; According to the similarity of the trend terms and the similarity of the seasonal terms corresponding to the two monitoring indicators, obtain the data overall consistency index between the two monitoring indicators in the current period; Perform weighted summation on the residual fluctuation similarity index, the residual distribution similarity index, and the data overall consistency index to obtain the similarity degree between the two monitoring indicators in the current period.

3. The groundwater dynamic monitoring and early warning system based on the Internet of Things technology according to claim 2, characterized in that, The obtaining of the residual fluctuation similarity index between the two monitoring indicators in the current period according to the chaos degree of each residual interval corresponding to the two monitoring indicators, includes: Denote any one of the two monitored indicators as the first monitored indicator and the other as the second monitored indicator. Round the information entropy of each residual interval corresponding to the first monitored indicator to obtain the target information entropy. Combine all the target information entropies to form the first residual information entropy sequence, and obtain the second residual information entropy sequence corresponding to the second monitored indicator. Construct a line chart based on the first residual information entropy sequence. Here, the vertical axis of the line chart is the target information entropy, and the horizontal axis is the residual interval. Obtain the slope between every two adjacent data in the line chart to get the slope sequence corresponding to the first residual information entropy sequence, denoted as the first slope sequence. Obtain the second slope sequence corresponding to the second residual information entropy sequence. Calculate the absolute value of the first difference between the two data at the same position in the first slope sequence and the second slope sequence. Perform linear normalization on the reciprocal of the average value of all the absolute values of the first differences to obtain the first variable. In the first residual information entropy sequence and the second residual information entropy sequence, calculate the absolute value of the second difference between the two data at the same position. Perform linear normalization on the reciprocal of the average value of all the absolute values of the second differences to obtain the second variable. Calculate the average value between the first variable and the second variable to obtain the residual fluctuation similarity index between the two monitored indicators in the current period.

4. The groundwater dynamic monitoring and early warning system based on the Internet of Things technology according to claim 3, wherein The obtaining of the residual distribution similarity index between the two monitored indicators in the current period according to the distribution of the outliers in the residual terms corresponding to the two monitored indicators includes: Form the first outlier sequence with the outliers in the residual term corresponding to the first monitored indicator, and form the second outlier sequence with the outliers in the residual term corresponding to the second monitored indicator. Calculate the absolute value of the third difference between the number of all elements in the first outlier sequence and the number of all elements in the second outlier sequence. Perform linear normalization on the reciprocal of the sum between the constant 1 and the absolute value of the third difference to obtain the third variable. According to the sampling time corresponding to each data in the first outlier sequence, calculate the time interval between every two adjacent data and accumulate them to obtain the first outlier distribution characteristic value of the first residual term. According to the sampling time corresponding to each data in the second outlier sequence, obtain the second outlier distribution characteristic value of the second residual term. Calculate the absolute value of the fourth difference between the first outlier distribution characteristic value and the second outlier distribution characteristic value. Perform linear normalization on the reciprocal of the absolute value of the fourth difference to obtain the fourth variable. Calculate the average value of all the data except the outliers in the residual term corresponding to the first monitored indicator, denoted as the first average value. Calculate the average value of all the data except the outliers in the residual term corresponding to the second monitored indicator, denoted as the second average value. Calculate the absolute value of the fifth difference between the first average value and the second average value. Perform linear normalization on the reciprocal of the sum between the constant 1 and the absolute value of the fifth difference to obtain the fifth variable. Calculate the average value among the third variable, the fourth variable, and the fifth variable to obtain the residual distribution similarity index between any two monitoring indicators during the current period.

5. The groundwater dynamic monitoring and early warning system based on the Internet of Things technology according to claim 2, characterized in that The obtaining of the overall data consistency index between any two monitoring indicators during the current period according to the similarity of the trend terms and the seasonal term similarity corresponding to any two monitoring indicators includes: Calculate the DTW distance between the trend terms corresponding to any two monitoring indicators, denoted as the first DTW distance, and perform linear normalization on the reciprocal of the sum of the constant 1 and the first DTW distance to obtain the sixth variable; Calculate the DTW distance between the seasonal terms corresponding to any two monitoring indicators, denoted as the second DTW distance, and perform linear normalization on the reciprocal of the sum of the constant 1 and the second DTW distance to obtain the seventh variable; Calculate the average value between the sixth variable and the seventh variable to obtain the overall data consistency index between any two monitoring indicators during the current period.

6. The groundwater dynamic monitoring and early warning system based on the Internet of Things technology according to claim 1, characterized in that, The classifying of all monitoring indicators of the groundwater according to all similarity degrees to obtain the classification result of all monitoring indicators during the current period includes: For the similarity degree between any two monitoring indicators, if the similarity degree between the two monitoring indicators is greater than the preset similarity degree threshold, then the two monitoring indicators are regarded as the same category.

7. The groundwater dynamic monitoring and early warning system based on the Internet of Things technology according to claim 1, characterized in that, The setting of the warning threshold for each monitoring indicator in the next period according to the classification result and the time series data of all monitoring indicators during the current period includes: For any category in the classification result, calculate the average value of the time series data of all monitoring indicators in the current period in the any category, denoted as the target average value, and use the target average value as the warning threshold for all monitoring indicators in the any category in the next period.

Images