Machine Learning-Based Early Warning Method and Device for Tap Water Radiation Pollution

Abstract

The present invention relates to a method and device for early warning of tap water radiation pollution based on machine learning. By setting multiple different sampling point positions in the target monitoring area to perform periodic sampling of tap water radiation indicators, a sequence of tap water radiation indicator values at each sampling point position is obtained. Then, the index fluctuation range in the sequence of tap water radiation indicator values at each sampling point position is traversed and found. Based on the index fluctuation range, multiple groups of two sampling point positions with approaching fluctuations are found, and these constitute multiple groups of early warning analysis data. Then, the radiation pollution degree index of each group of early warning analysis data is calculated, and this is compared with the radiation pollution degree threshold to determine whether there is a tap water radiation pollution event, and an alarm is given when there is a tap water radiation pollution event. The present invention suppresses the influence of fluctuations in radiation monitoring indicators caused by the high fluidity of tap water through the above analysis process to ensure the accuracy of the early warning result, and belongs to the technical field related to tap water monitoring.

Description

Technical Field

[0001] The present invention relates to the technical field of tap water monitoring, and particularly to a method and device for early warning of tap water radiation pollution based on machine learning. Background Art

[0002] Currently, the early warning method for tap water radiation pollution often sets a radioactive activity threshold with a concentration unit for α particles and β particles. Once there is a tap water sample exceeding the screening level, an early warning reminder is given.

[0003] In addition, a prior application with the application number 201610153624.6, a method and system for automatic detection and alarm of tap water radiation pollution, proposes to monitor the radioactive count of tap water in a tap water pipeline in real time through a radiation sensor, and store the relevant data in local memory, or transmit it to a mobile phone or computer client, so as to realize timely discovery of potential safety hazards in tap water by continuously obtaining the radioactive data of tap water, and at the same time send the hazard information out in a timely manner through channels such as text messages and the Internet, and can realize audible and visual alarms.

[0004] Although this scheme proposes a relatively mature alarm hardware system, on the one hand, its monitoring effect is not good by using a radiation sensor, and on the other hand, it only gives an alarm by the number of times exceeding the preset radiation sensor reading threshold within the monitoring period, and the accuracy rate is also low.

[0005] Moreover, the above methods do not consider the high fluidity of tap water. When tap water is flowing, especially when the flow rate is particularly large, the radiation particles in it fluctuate with the water flow, sometimes strong and sometimes weak. Therefore, how to give an early warning for each sampling data to improve the accuracy of the early warning result in this situation is an urgent problem to be solved in this field. Summary of the Invention

[0006] The purpose of the present invention is to at least solve one of the deficiencies of the prior art, and provide a method and device for early warning of tap water radiation pollution based on machine learning.

[0007] To achieve the above purpose, the present invention adopts the following technical solutions:

[0008] Specifically, a method for early warning of tap water radiation pollution based on machine learning is proposed, including the following:

[0009] Obtain the tap water radiation index values at different sampling point positions in the target monitoring area through periodic sampling, and obtain the tap water radiation index value sequence at each sampling point position;

[0010] Traverse and find out the index fluctuation range in the tap water radiation index value sequence at each sampling point position;

[0011] Find the position of the second sampling point whose fluctuation is close to that of the first sampling point based on the index fluctuation range, and use the sequence of tap water radiation index values at the positions of the first sampling point and the second sampling point as a set of early warning analysis data, and then obtain multiple sets of early warning analysis data, where the position of the first sampling point is the position of any sampling point;

[0012] Calculate and obtain the radiation pollution degree index of each set of early warning analysis data;

[0013] If the radiation pollution degree index of at least one set of early warning analysis data exceeds the radiation pollution degree threshold, give an early warning reminder.

[0014] Furthermore, specifically, different sampling point positions in the target monitoring area refer to the positions of different tap water outlets in the target monitoring area of the same tap water source.

[0015] Furthermore, specifically, the method for obtaining the tap water radiation index value includes,

[0016] Detect the radioactivity of α or β particles in tap water based on a Geiger counter as the tap water radiation index value.

[0017] Furthermore, specifically, traverse and find the index fluctuation range in the sequence of tap water radiation index values at each sampling point position, including,

[0018] For any sampling point position, if the preset sampling period for periodic sampling is Cycle1, and the tap water radiation index value is denoted as Rad_In;

[0019] From the start time to the end time of periodic sampling, judge whether there is an index fluctuation in the collection time C_exam of Rad_In in the order of sampling successively, find all the collection times C_exam with index fluctuations, and find the 2 collection times with the smallest time interval from the collection times C_exam with index fluctuations, and denote them as the C_begin collection time and the C_close collection time in the order of collection successively;

[0020] Then the range from the C_begin collection time to the C_close collection time is the index fluctuation range;

[0021] Among them, when the following situation exists, it is determined that there is an index fluctuation in the collection time C_exam of Rad_In,

[0022] If the value of Rad_In collected at the collection time C_exam is less than the value of Rad_In collected at the previous collection time, i.e., the collection time C_exam - Cycle1, and the value of Rad_In collected at the collection time C_exam - 2*Cycle1 is less than the value of Rad_In collected at the collection time C_exam - Cycle1, and the value of Rad_In collected at the collection time C_exam is greater than the value of C_exam_Rad_In_Avg;

[0023] Among them, C_exam_Rad_In_Avg represents the average value of the non-zero Rad_In values at all sampling point positions at the collection time C_exam.

[0024] Furthermore, specifically, finding the second sampling point position whose fluctuation is close to that of the first sampling point position based on the index fluctuation range includes,

[0025] Record the maximum tap water radiation index value as Rad_InMaxLoc1 and the minimum tap water radiation index value as Rad_InMinLoc1 within the index fluctuation range range1 corresponding to the first sampling point position; define the variable i as the number of the sampling point position, and make the following judgments,

[0026] If the i-th sampling point position meets the following conditions, then judge that the i-th sampling point position and the first sampling point position have similar fluctuations, and record the i-th sampling point position as the second sampling point position whose fluctuation is close to that of the first sampling point position,

[0027] Rad_InMax(i) ≤ Rad_InMaxLoc1 + recom_Rad_In(i), and,

[0028] Rad_InMax(i) ≤ Rad_InMaxLoc1 + recom_Rad_In(i);

[0029] Among them, Rad_InMax(i) represents the maximum tap water radiation index value within the index fluctuation range range1 of the i-th sampling point position; Rad_InMin(i) represents the minimum tap water radiation index value within the index fluctuation range range1 of the i-th sampling point position; recom_Rad_In(i) represents the index compensation value of the i-th sampling point position, and the calculation formula of recom_Rad_In(i) is as follows:

[0030] ;

[0031] Among them, N is the total number of sampling point positions, Denote the maximum tap water radiation index value within the index fluctuation range corresponding to the i-th sampling point position at the j-th sampling point position; Denote the minimum tap water radiation index value within the index fluctuation range corresponding to the i-th sampling point position at the j-th sampling point position.

[0032] Furthermore, specifically, calculate and obtain the radiation pollution degree index of each group of early warning analysis data, including,

[0033] The calculation formula for the radiation pollution degree index of the q-th group of early warning analysis data denoted as Index_q is as follows:

[0034]

[0035] Wherein, Denote the average value of all non-zero Rad_In values of the first sampling point position within its index fluctuation range range1, Denote the average value of all non-zero Rad_In values of the second sampling point position within the index fluctuation range range1.

[0036] Furthermore, specifically, the way to give early warning reminders is,

[0037] Pre-bind multiple administrator IPs. When early warning reminders are needed, pack all periodic sampling data with the preset text reminder content and send it to all administrator IPs for subsequent verification and analysis.

[0038] The present invention also proposes a device for early warning of tap water radiation pollution based on machine learning, including the following:

[0039] A data acquisition module, used to obtain the tap water radiation index values at different sampling point positions in the target monitoring area through periodic sampling, and obtain the tap water radiation index value sequence at each sampling point position;

[0040] An index fluctuation range calculation module, used to traverse and find out the index fluctuation range in the tap water radiation index value sequence at each sampling point position;

[0041] A data generation module, used to find out the second sampling point position whose fluctuation is close to that of the first sampling point position based on the index fluctuation range, and use the tap water radiation index value sequences of the first sampling point position and the second sampling point position together as a group of early warning analysis data, and then obtain multiple groups of early warning analysis data, where the first sampling point position is any sampling point position;

[0042] An index calculation module, used to calculate and obtain the radiation pollution degree index of each group of early warning analysis data;

[0043] An early warning reminder module is used to give an early warning reminder when the radiation pollution degree index of at least one set of early warning analysis data exceeds the radiation pollution degree threshold.

[0044] The beneficial effects of the present invention are as follows:

[0045] The present invention provides a method and device for early warning of tap water radiation pollution based on machine learning. By setting multiple different sampling point positions in the target monitoring area to perform periodic sampling of tap water radiation indicators, a sequence of tap water radiation indicator values at each sampling point position is obtained. Then, the index fluctuation range in the sequence of tap water radiation indicator values at each sampling point position is traversed and found. Based on the index fluctuation range, multiple pairs of two sampling point positions with approaching fluctuations are found, and these form multiple sets of early warning analysis data. Then, the radiation pollution degree index of each set of early warning analysis data is calculated, and this is compared with the radiation pollution degree threshold to determine whether there is a tap water radiation pollution event, and an alarm is given when there is a tap water radiation pollution event. The present invention suppresses the influence of fluctuations in radiation monitoring indicators caused by the high fluidity of tap water through the above analysis process, and ensures the accuracy of the early warning result as much as possible. Description of the Drawings

[0046] By elaborating on the embodiments shown in combination with the drawings, the above and other features of the present disclosure will become more obvious. The same reference numerals in the drawings of the present disclosure represent the same or similar elements. Obviously, the drawings in the following description are only some embodiments of the present disclosure. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings. In the drawings:

[0047] Figure 1 Shows a flowchart of the method for early warning of tap water radiation pollution based on machine learning of the present invention;

[0048] Figure 2 Shows a structural schematic diagram of the device for early warning of tap water radiation pollution based on machine learning of the present invention. Detailed Embodiments

[0049] The following will clearly and completely describe the concept, specific structure and technical effects generated by the present invention in combination with the embodiments and the drawings, so as to fully understand the purpose, solution and effects of the present invention. It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. The same reference numerals used throughout the drawings indicate the same or similar parts.

[0050] Embodiment 1, referring to Figure 1 , the present invention provides a method for early warning of tap water radiation pollution based on machine learning, including the following:

[0051] Step 110: Obtain the tap water radiation index values at different sampling point positions in the target monitoring area through periodic sampling, and obtain the tap water radiation index value sequence at each sampling point position;

[0052] Step 120: Traverse and find out the index fluctuation range in the tap water radiation index value sequence at each sampling point position;

[0053] Step 130: Based on the index fluctuation range, find out the second sampling point position that fluctuates close to the first sampling point position, and use the tap water radiation index value sequences of the first sampling point position and the second sampling point position together as a set of early warning analysis data, and then obtain multiple sets of early warning analysis data, where the first sampling point position is any sampling point position;

[0054] Step 140: Calculate and obtain the radiation pollution degree index of each set of early warning analysis data;

[0055] Step 150: If the radiation pollution degree index of at least one set of early warning analysis data exceeds the radiation pollution degree threshold, give an early warning reminder.

[0056] In this Embodiment 1, by setting multiple different sampling point positions in the target monitoring area for periodic sampling of the tap water radiation index, the tap water radiation index value sequence at each sampling point position is obtained. Then, traverse and find out the index fluctuation range in the tap water radiation index value sequence at each sampling point position. Based on the index fluctuation range, find out multiple sets of two sampling point positions with similar fluctuations, and form multiple sets of early warning analysis data with them. Then calculate the radiation pollution degree index of each set of early warning analysis data, and compare it with the radiation pollution degree threshold to judge whether there is a tap water radiation pollution event, and give an alarm when there is a tap water radiation pollution event. The present invention suppresses the influence of the fluctuation of the radiation monitoring index caused by the high fluidity of tap water through the above analysis process, and ensures the accuracy of the early warning result as much as possible.

[0057] As a preferred embodiment of the present invention, specifically, the different sampling point positions in the target monitoring area refer to the positions of different tap water outlets in the target monitoring area of the same tap water source.

[0058] In this preferred embodiment, considering the fluidity problem of tap water, if data sampling and monitoring of tap water are carried out simply from a certain position, the monitoring result will be inaccurate. Therefore, considering the actual application scenario of tap water use, when it is necessary to monitor the tap water in the target monitoring area (i.e., the same tap water source), by analyzing the tap water sampling at multiple pre-selected tap water outlets, the above problems can be solved to a certain extent.

[0059] As a preferred embodiment of the present invention, specifically, the method for obtaining the tap water radiation index value includes,

[0060] Detect the radioactivity of α or β particles in tap water based on a Geiger counter as the radiation index value of tap water.

[0061] In this preferred embodiment, obtaining the radiation index value of tap water based on a Geiger counter specifically includes the following processes and precautions.

[0062] 1. Understand the working principle of the Geiger counter

[0063] The Geiger counter detects the interaction between radioactive particles (such as α and β particles) and the internal gas of the detector, generating an ionization effect and thus producing signals. These signals are used by the counter to record the number of particles passing through within a specific time (count).

[0064] 2. The penetration abilities of α and β particles are different

[0065] α particles: Have a weak penetration ability and can only penetrate thin paper or the surface of the skin. Since α particles can only penetrate a very short distance, it is necessary to ensure that the window of the Geiger counter is thin enough for effective detection.

[0066] β particles: Have a stronger penetration ability and can penetrate thinner materials. For β particles, a Geiger counter with a thicker window is usually required to avoid excessive interference from α particles.

[0067] 3. Configure a suitable Geiger counter

[0068] Detection of α particles: Use a Geiger counter with a thin metal window, which can effectively detect α particles. For the detection of α particles, it is necessary to ensure that the window of the Geiger counter can transmit α particles while blocking heavier β particles.

[0069] Detection of β particles: A Geiger counter with a thicker window is required to prevent excessive α particles from affecting the detection result. Since β particles can penetrate a thicker window, an ordinary Geiger counter can be used for detection.

[0070] 4. Measure the radioactivity concentration

[0071] Calibrate the counter: First, ensure that the Geiger counter has been calibrated to ensure the accuracy of its readings. A known radioactive substance or standard source can be used for calibration.

[0072] Perform the measurement: Place the Geiger counter close to the sample or radioactive source and record the count value per unit time. Usually, the unit time is counts per minute (cpm).

[0073] Separating α and β signals: If it is necessary to distinguish between α and β particles, it can be achieved by using different detectors (e.g., α - specific window and β - specific window) or a Geiger counter with a filter. Generally, if the Geiger counter does not have a separate window, the count of α particles will be significantly lower than that of β particles.

[0074] 5. Calculating the radioactive concentration

[0075] Based on the known relationship between the activity of a substance and the counting rate, the concentrations of α and β radioactive substances in the sample can be calculated. Using a standard calibration curve (generated by standard sources with known activities), according to the results read by the counter, the concentration of radioactive substances in the sample can be deduced.

[0076] 6. Precautions

[0077] Sample handling: Ensure that the sample is properly handled and placed within the measurement area of the counter to avoid errors.

[0078] Effect of background radiation: Background radiation may interfere with the measurement results. Therefore, it is necessary to record the background radiation first without the sample and subtract this background value from the measurement results.

[0079] Through the above steps, a Geiger counter can effectively detect the concentrations of α and β radioactive substances.

[0080] As a preferred embodiment of the present invention, specifically, traverse and find the index fluctuation range in the sequence of tap - water radiation index values at each sampling point position, including,

[0081] For any sampling point position, if the preset sampling period for periodic sampling is Cycle1, and the tap - water radiation index value is denoted as Rad_In;

[0082] From the start time to the end time of periodic sampling, judge whether there is an index fluctuation in the collection time C_exam of Rad_In in the order of sampling successively. Find all the collection times C_exam with index fluctuations, and find the two collection times with the smallest time interval from the collection times C_exam with index fluctuations. Denote them as the C_begin collection time and the C_close collection time in the order of collection;

[0083] Then the range from the C_begin collection time to the C_close collection time is the index fluctuation range;

[0084] Among them, when the following situation exists, it is determined that there is an index fluctuation in the collection time C_exam of Rad_In,

[0085] If the value of Rad_In collected at the collection time C_exam is less than the value of Rad_In collected at the previous collection time, i.e., the collection time C_exam - Cycle1, and the value of Rad_In collected at the collection time C_exam - 2*Cycle1 is less than the value of Rad_In collected at the collection time C_exam - Cycle1, and the value of Rad_In collected at the collection time C_exam is greater than the value of C_exam_Rad_In_Avg;

[0086] Wherein, C_exam_Rad_In_Avg represents the average value of non-zero Rad_In values at all sampling point positions at the collection time C_exam.

[0087] In this preferred embodiment, through the above analysis method, based on the adjacent sampling interval data analysis, the index fluctuation range in the tap water radiation index value sequence at any sampling point position can be accurately found.

[0088] As a preferred embodiment of the present invention, specifically, finding the second sampling point position whose fluctuation approaches that of the first sampling point position based on the index fluctuation range includes,

[0089] Record the maximum tap water radiation index value in the corresponding index fluctuation range range1 of the first sampling point position as Rad_InMaxLoc1, and the minimum tap water radiation index value as Rad_InMinLoc1; define the variable i as the number of the sampling point position, and make the following judgments,

[0090] If the i-th sampling point position satisfies the following conditions, it is determined that the i-th sampling point position and the first sampling point position fluctuate closely, and the i-th sampling point position is recorded as the second sampling point position whose fluctuation approaches that of the first sampling point position,

[0091] Rad_InMax(i) ≤ Rad_InMaxLoc1 + recom_Rad_In(i), and,

[0092] Rad_InMax(i) ≤ Rad_InMaxLoc1 + recom_Rad_In(i);

[0093] Wherein, Rad_InMax(i) represents the maximum tap water radiation index value of the i-th sampling point position in the index fluctuation range range1; Rad_InMin(i) represents the minimum tap water radiation index value of the i-th sampling point position in the index fluctuation range range1; recom_Rad_In(i) represents the index compensation value of the i-th sampling point position, and the calculation formula of recom_Rad_In(i) is as follows:

[0094]

[0095]

[0096] Among them, N is the total number of sampling point positions, represents the maximum tap water radiation index value within the index fluctuation range corresponding to the i-th sampling point position at the j-th sampling point position; represents the minimum tap water radiation index value within the index fluctuation range corresponding to the i-th sampling point position at the j-th sampling point position.

[0097] In this preferred embodiment, on the one hand, the proximity relationship of the tap water radiation index values between two sampling point positions is compared through a preset rule, and on the other hand, an index compensation value is added to the data being compared, and the proximity relationship of the tap water radiation index values between two sampling point positions is corrected by associating all the data of the sampling point positions; try to make the determination result of the fluctuation relationship between the first sampling point position and the second sampling point position more accurate.

[0098] As a preferred embodiment of the present invention, specifically, the radiation pollution degree index of each group of early warning analysis data is calculated and obtained. Specifically, the radiation pollution degree index of each group of early warning analysis data is calculated and obtained, including,

[0099] The calculation formula for the radiation pollution degree index of the q-th group of early warning analysis data, denoted as Index_q, is as follows:

[0100]

[0101] Among them, represents the average value of all non-zero Rad_In values of the first sampling point position within its index fluctuation range range1, represents the average value of all non-zero Rad_In values of the second sampling point position within the index fluctuation range range1.

[0102] In this preferred embodiment, each set of early warning analysis data found through the above solution is equivalent to the ROI data to be analyzed. Then, Index_q obtained by averaging the early warning analysis data within the range of range1 in the ROI data can basically accurately represent the value to be compared in this sampling. By comparing it with the screening level standard value of the radioactive screening index set by the World Health Organization for drinking water quality, an accurate early warning result can be output. Additionally, according to the differences in the monitored α and β particles, the corresponding screening indicators are adjusted accordingly, that is, the radiation pollution degree threshold changes synchronously with the radioactive screening indicators set by the World Health Organization for drinking water quality, namely total α and total β radioactivity, and the corresponding screening levels are 0.5 Bq / L and 1 Bq / L respectively. Additionally, as a preference, an adjustment coefficient P can also be set according to the differences in different target monitoring areas, and P * radiation pollution degree threshold is used to eliminate the differences in different target monitoring areas, where P is a preset coefficient determined through preliminary experiments.

[0103] As a preferred embodiment of the present invention, specifically, the method of giving early warning reminders is as follows:

[0104] Pre-bind multiple administrator IPs. When early warning reminders are required, all periodic sampling data and preset text reminder content are packaged and sent to all administrator IPs for subsequent verification and analysis.

[0105] In this preferred embodiment, by presetting the administrator IP, when early warning reminders are required, relevant administrators are informed through the preset text reminder content to conduct further verification and analysis, and at the same time, all periodic sampling data is attached for administrators to screen and use, and then the final result determination is made.

[0106] Referring to Figure 2 , Example 2, the present invention also proposes a device for early warning of tap water radiation pollution based on machine learning, including the following:

[0107] A data acquisition module 100, which is used to obtain the tap water radiation index values at different sampling point positions in the target monitoring area through periodic sampling, and obtain a sequence of tap water radiation index values at each sampling point position;

[0108] An index fluctuation range calculation module 200, which is used to traverse and find out the index fluctuation range in the sequence of tap water radiation index values at each sampling point position;

[0109] A data generation module 300, which is used to find a second sampling point position whose fluctuation is close to that of the first sampling point position based on the index fluctuation range, and use the sequences of tap water radiation index values at the first sampling point position and the second sampling point position together as a set of early warning analysis data, and then obtain multiple sets of early warning analysis data, where the first sampling point position is any sampling point position;

[0110] An index calculation module 400 is configured to calculate and obtain the radiation pollution degree index of each group of early warning analysis data;

[0111] An early warning reminder module 500 is configured to give an early warning reminder if the radiation pollution degree index of at least one group of early warning analysis data exceeds the radiation pollution degree threshold.

[0112] In this Embodiment 2, which is consistent with the method embodiment proposed by the present invention, by setting multiple different sampling point positions in the target monitoring area to perform periodic sampling of the tap water radiation index, a sequence of tap water radiation index values at each sampling point position is obtained. Then, the index fluctuation range in the sequence of tap water radiation index values at each sampling point position is traversed and found. Based on the index fluctuation range, multiple groups of two sampling point positions with approaching fluctuations are found, and these form multiple groups of early warning analysis data. Then, the radiation pollution degree index of each group of early warning analysis data is calculated, and compared with the radiation pollution degree threshold to determine whether there is a tap water radiation pollution event, and an alarm is given when there is a tap water radiation pollution event. The present invention suppresses the influence of the radiation monitoring index fluctuation caused by the high fluidity of tap water through the above analysis process, and ensures the accuracy of the early warning result as much as possible.

[0113] In addition, in each embodiment of the present invention, the various functional modules can be integrated in a processing module, or each module can exist physically alone, or two or more modules can be integrated in one module. The above integrated modules can be implemented in the form of hardware or in the form of software functional modules.

[0114] If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, to implement all or part of the processes in the above method embodiments of the present invention, it can also be completed by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or system, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc., that can carry the computer program code.

[0115] Although the description of the present invention has been quite detailed and several of the described embodiments have been described in particular, it is not intended to be limited to any of these details or embodiments or any particular embodiment, but rather should be regarded as providing a broad interpretation of these claims in light of the prior art by reference to the appended claims, so as to effectively cover the intended scope of the present invention. In addition, the present invention has been described above in terms of embodiments foreseeable by the inventors for the purpose of providing a useful description, and non-substantive modifications to the present invention that are not currently foreseeable may still represent equivalent modifications of the present invention.

[0116] As described above, these are only the preferred embodiments of the present invention. The present invention is not limited to the above-described embodiments. As long as the same means are used to achieve the technical effects of the present invention, they should fall within the protection scope of the present invention. Within the protection scope of the present invention, various different modifications and changes may be made to its technical solutions and / or embodiments.

Claims

1. A tap water radiation pollution early warning method based on machine learning, characterized in that: Includes the following: The tap water radiation index values ​​at different sampling points in the target monitoring area are obtained by periodic sampling, and a tap water radiation index value sequence at each sampling point is obtained; wherein the different sampling point positions in the target monitoring area refer to the positions of different tap water outlets in the target monitoring area of ​​the same tap water source; Traverse and find the fluctuation range of the tap water radiation index value sequence at each sampling point; specifically include: For any sampling point location, if the preset sampling period of periodic sampling is Cycle1, and the tap water radiation index value is recorded as Rad_In; From the start time to the end time of periodic sampling, the collection time of Rad_In is recorded as C_exam in the order of sampling to determine whether there is index fluctuation, and all collection times C_exam with index fluctuation are found. From the collection times C_exam with index fluctuation, the two collection times with the smallest time interval are found, and they are recorded as C_begin collection time and C_close collection time in the order of collection; Then the range from the C_begin collection time to the C_close collection time is the indicator fluctuation range; Among them, when the following conditions exist, it is determined that there is an index fluctuation at the collection time C_exam of Rad_In, If the value of Rad_In collected at the collection time C_exam is less than the value of Rad_In collected at the previous collection time, i.e., the collection time C_exam-Cycle1, and the value of Rad_In collected at the collection time C_exam-2*Cycle1 is less than the value of Rad_In collected at the collection time C_exam-Cycle1, and the value of Rad_In collected at the collection time C_exam is greater than the value of C_exam_Rad_In_Avg; Wherein, C_exam_Rad_In_Avg represents the average value of non-zero Rad_In values ​​at all sampling points at the acquisition time C_exam; Based on the index fluctuation range, find the second sampling point position that is close to the first sampling point position, and use the tap water radiation index value sequences of the first sampling point position and the second sampling point position as a set of early warning analysis data, thereby obtaining multiple sets of early warning analysis data, wherein the first sampling point position is an arbitrary sampling point position; Calculate and obtain the radiation contamination level index for each set of early warning analysis data; If at least one set of radiation contamination level indicators of early warning analysis data exceeds the radiation contamination level threshold, an early warning reminder will be issued.

2. The method for early warning of tap water radiation pollution based on machine learning according to claim 1 is characterized in that: Specifically, the method of obtaining the tap water radiation index value includes: The radioactivity of α or β particles in tap water is detected by a Geiger counter as the tap water radiation index value.

3. The method for early warning of tap water radiation pollution based on machine learning according to claim 1 is characterized in that: Specifically, based on the indicator fluctuation range, finding the second sampling point position that is close to the first sampling point position fluctuation includes: The maximum tap water radiation index value of the first sampling point in the corresponding index fluctuation range is recorded as range1 as Rad_InMaxLoc1, and the minimum tap water radiation index value is Rad_InMinLoc1; define the variable i as the number of the sampling point position, and make the following judgment: If the position of the i-th sampling point satisfies the following conditions, it is determined that the position of the i-th sampling point and the first sampling point are close to each other in terms of fluctuation, and the position of the i-th sampling point is recorded as the second sampling point position close to the first sampling point position in terms of fluctuation. Rad_InMax(i)≤Rad_InMaxLoc1+recom_Rad_In(i), and Rad_InMax(i)≤Rad_InMaxLoc1+recom_Rad_In(i); Among them, Rad_InMax(i) represents the maximum tap water radiation index value of the i-th sampling point within the index fluctuation range range1; Rad_InMin(i) represents the minimum tap water radiation index value of the i-th sampling point within the index fluctuation range range1; recom_Rad_In(i) represents the index compensation value of the i-th sampling point, and the calculation formula of recom_Rad_In(i) is as follows: Wherein, N is the total number of sampling point locations, Maxrange(i,j) represents the maximum tap water radiation index value at the j-th sampling point location within the index fluctuation range corresponding to the i-th sampling point location; Minrange(i,j) represents the minimum tap water radiation index value at the j-th sampling point location within the index fluctuation range corresponding to the i-th sampling point location.

4. The method for early warning of tap water radiation pollution based on machine learning according to claim 3 is characterized in that: Specifically, the radiation contamination index of each set of early warning analysis data is calculated and obtained, including: The calculation formula of the radiation contamination index of the qth group of early warning analysis data, marked as Index_q, is as follows: Wherein, Rad_In_range1_Avg1 represents the average value of all non-zero Rad_In values ​​at the first sampling point in its index fluctuation range range1, and Rad_In_range1_Avg2 represents the average value of all non-zero Rad_In values ​​at the second sampling point in its index fluctuation range range1.

5. The method for early warning of tap water radiation pollution based on machine learning according to claim 1 is characterized in that: Specifically, the method of providing early warning reminder is: Pre-bind multiple administrator IPs. When an early warning reminder is needed, all periodic sampling data and preset text reminder content are packaged and sent to all administrator IPs for subsequent verification and analysis.

6. A device for early warning of tap water radiation contamination based on machine learning, characterized in that: The method for early warning of tap water radiation contamination based on machine learning as claimed in claim 1 comprises the following: The data acquisition module is used to obtain the tap water radiation index values ​​at different sampling points in the target monitoring area by periodic sampling, and obtain the tap water radiation index value sequence at each sampling point; wherein the different sampling point positions in the target monitoring area refer to the positions of different tap water outlets in the target monitoring area of ​​the same tap water source; The index fluctuation range calculation module is used to traverse and find the index fluctuation range in the tap water radiation index value sequence at each sampling point; specifically includes: For any sampling point location, if the preset sampling period of periodic sampling is Cycle1, and the tap water radiation index value is recorded as Rad_In; From the start time to the end time of periodic sampling, the collection time of Rad_In is recorded as C_exam in the order of sampling to determine whether there is index fluctuation, and all collection times C_exam with index fluctuation are found. From the collection times C_exam with index fluctuation, the two collection times with the smallest time interval are found, and they are recorded as C_begin collection time and C_close collection time in the order of collection; Then the range from the C_begin collection time to the C_close collection time is the indicator fluctuation range; Among them, when the following conditions exist, it is determined that there is an index fluctuation at the collection time C_exam of Rad_In, If the value of Rad_In collected at the collection time C_exam is less than the value of Rad_In collected at the previous collection time, i.e., the collection time C_exam-Cycle1, and the value of Rad_In collected at the collection time C_exam-2*Cycle1 is less than the value of Rad_In collected at the collection time C_exam-Cycle1, and the value of Rad_In collected at the collection time C_exam is greater than the value of C_exam_Rad_In_Avg; Wherein, C_exam_Rad_In_Avg represents the average value of non-zero Rad_In values ​​at all sampling points at the acquisition time C_exam; A data generation module is used to find a second sampling point position that is close to the first sampling point position based on the index fluctuation range, and use the tap water radiation index value sequences of the first sampling point position and the second sampling point position as a set of early warning analysis data, thereby obtaining multiple sets of early warning analysis data, wherein the first sampling point position is an arbitrary sampling point position; An index calculation module is used to calculate and obtain the radiation contamination degree index of each set of early warning analysis data; The early warning reminder module is used to issue an early warning reminder when the radiation pollution level index of at least one set of early warning analysis data exceeds the radiation pollution level threshold.

Images