An Automatic Monitoring Data Analysis Method for Suspicious Clues Based on Artificial Intelligence

Abstract

The present invention relates to a method for automatically monitoring and analyzing data of suspicious clues based on artificial intelligence, belonging to the field of electrical digital data processing. The method includes: performing a set number of learning actions on a convolutional neural network to obtain an automatic monitoring model for suspicious clues; using the automatic monitoring model for suspicious clues to intelligently analyze whether the target monitored water surface belongs to a polluted water surface and the suspected distance from the pollution source to the current monitoring point based on time-sharing dynamic water quality-related data and time-sharing dynamic visualization data; through the present invention, aiming at the technical problems that the analysis results of environmental pollution monitoring data are not rich enough and the analysis process is complex, it is possible to use an automatic monitoring model for suspicious clues with a targeted structure design to perform intelligent data analysis on suspicious clues such as whether there is environmental pollution in a fixed water area and the distance from the environmental pollution source to the fixed water area based on time-sharing dynamic water quality data and time-sharing dynamic visualization data, thereby solving the above technical problems.

Description

Technical Field

[0001] The present invention relates to the field of electrical digital data processing, and in particular to an artificial intelligence-based automatic monitoring data analysis method for suspicious clues. Background Art

[0002] With the acceleration of the industrialization process and the enhancement of environmental protection awareness, the automatic monitoring technology of pollution sources has become an indispensable part of environmental management. Traditional pollution monitoring methods mainly rely on manual sampling and laboratory analysis, which are time-consuming, costly, and inefficient. In recent years, with the development of modern information technologies such as the Internet of Things (IoT), big data, and artificial intelligence (AI), automatic monitoring systems have begun to be widely used, enabling real-time monitoring of pollutant emissions, overcoming various defects of traditional pollution monitoring methods, such as high data processing complexity, high false alarm and missed alarm probabilities, insufficient data analysis depth, slow response speed, easy waste of human and material resources, low data utilization rate, poor user experience, unreliable data quality, and poor scalability and flexibility.

[0003] Exemplarily, the invention patent with the publication number CN112101789A proposes an artificial intelligence-based water pollution alarm level identification method, and the method includes the following steps: Step A: Obtain historical data and preprocess the historical data; Step B: Manually mark the risk levels of the historical data to obtain risk label data, and divide it into a training set and a test set according to a ratio; Step C: Build a deep learning model, input the training set data into the deep learning model, and output the trained model; Step D: Use the test set for testing, and if it meets the preset criteria, use it as the water pollution alarm level identification model; Step E: Identify the real-time data of each inlet node to obtain the water pollution alarm level. The advantages of the present invention are as follows: Considering the water quality and water volume data of the water body comprehensively, based on artificial intelligence deep learning technology, discovering the deep relationship between water pollution and multiple parameters, and accurately judging and identifying the pollution alarm level.

[0004] Exemplarily, the invention patent with the publication number CN109631998A proposes an environmental artificial intelligence monitoring system and an environmental artificial intelligence monitoring method. The system includes: a particulate matter concentration parameter acquisition module for acquiring particulate matter concentration parameters; a meteorological five-parameter acquisition module for acquiring meteorological five parameters; a noise parameter acquisition module for acquiring noise parameters; an image parameter acquisition module for acquiring image parameters; a particulate matter pollution judgment module for editing particulate matter pollution judgment results and sending them to an alarm information editing module; a meteorological pollution judgment module for editing meteorological pollution judgment results and sending them to the alarm information editing module; a noise pollution judgment module for editing noise pollution judgment results and sending them to the alarm information editing module; an alarm information editing module for editing corresponding alarm information and sending it to a terminal; and an image display module for displaying the received image parameters. The present invention has the advantages of simple structure and easy implementation.

[0005] It can be seen that in the existing various environmental pollution automatic monitoring data analysis solutions, either although the artificial intelligence mechanism is adopted, the real-time water quality data is analyzed, and it is impossible to judge environmental pollution and preliminarily locate the environmental pollution source according to the time-sharing water quality data and the more abundant visualized data, or although the water quality data and the visualized data are jointly used to judge environmental pollution, the artificial intelligence mechanism is not adopted, and it is also impossible to achieve the preliminary location of the environmental pollution source. Therefore, the analysis results either lack pertinence or still have various defects of the traditional pollution monitoring means. Summary of the Invention

[0006] To solve the technical problems in the prior art, the present invention provides an artificial intelligence-based automatic monitoring data analysis method for suspicious clues. For a fixed water area suspected of water pollution, a suspicious clue automatic monitoring model with a targeted structure design is used, and the time-sharing dynamic water quality data and the time-sharing dynamic visualized data are jointly used as basic data to perform intelligent data analysis on suspicious clues such as whether there is environmental pollution in the fixed water area and the distance from the environmental pollution source to the fixed water area, thereby overcoming various defects of the traditional pollution monitoring means, enriching the analysis results of environmental pollution, and providing more auxiliary information for users to judge and control environmental pollution.

[0007] According to the present invention, there is provided an artificial intelligence-based automatic monitoring data analysis method for suspicious clues, the method comprising:

[0008] Collecting water quality-related data corresponding to each moment of the target monitoring water surface directly below the current monitoring point, wherein the water quality-related data corresponding to each moment of the target monitoring water surface is the chemical oxygen demand, permanganate index, phosphorus element content, total content of minerals dissolved in water, and water temperature of the target monitoring water surface at that moment;

[0009] Capturing visualization data of a real-time imaging picture corresponding to a target monitoring water surface directly below a current monitoring point, wherein the visualization data of the real-time imaging picture is each cyan component value, each magenta component value, each yellow component value, and each black component value corresponding to each pixel point of the real-time imaging picture;

[0010] Performing a set number of learning actions on the convolutional neural network to obtain a convolutional neural network after completing each learning operation, and outputting the convolutional neural network after completing each learning operation as an automatic monitoring model for suspicious clues, wherein the number of learning operations completed by the convolutional neural network is positively correlated with the water surface area of ​​the target monitored water surface;

[0011] The water surface area of ​​the target monitored water surface, the time length of the preset time interval, the water quality related data corresponding to each time moment evenly spaced within the preset time interval, and the visualization data of each real-time imaging picture corresponding to each time moment evenly spaced within the preset time interval are synchronously input into the suspicious clue automatic monitoring model to run the suspicious clue automatic monitoring model to obtain the pollution mark indicating whether the target monitored water surface is a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model;

[0012] Among them, the water quality related data of the target monitoring water surface directly below the current monitoring point originates from the water body at the target monitoring water surface directly below the current monitoring point;

[0013] Among them, when the pollution mark output by the automatic monitoring model of suspicious clues indicates that the target monitored water surface is not a polluted water surface, the suspected distance from the pollution source to the current monitoring point output at the same time is zero.

[0014] Compared with the prior art, the present invention has at least the following important inventive features:

[0015] Invention point A: For the target monitoring water surface directly below the current monitoring point, an artificial intelligence mechanism is used to intelligently analyze the pollution mark of whether the target monitoring water surface belongs to a polluted water surface and the suspected distance from the pollution source to the current monitoring point based on various time-sharing visual data of the target monitoring water surface and the time-sharing water quality related data of the target monitoring water surface, thereby completing the automatic monitoring, on-site display and wireless reporting of suspicious clue information of water surface environmental pollution including the pollution mark and the suspected distance from the pollution source to the current monitoring point based on the time-sharing dynamic information;

[0016] Inventive Point B: The intelligent analysis of suspicious clue information is based on an automatic monitoring model for suspicious clues. The automatic monitoring model for suspicious clues is a convolutional neural network after each learning operation is completed. Particularly importantly, the number of learning operations completed by the convolutional neural network is positively correlated with the water surface area of the target monitored water surface. Thus, automatic monitoring models with different structures are constructed for monitored water surfaces of different areas, ensuring the stability and effectiveness of the intelligent analysis results;

[0017] Inventive Point C: Specifically, the time-sharing visualization data of the target monitored water surface are the visualization data of each real-time imaging picture corresponding to each moment at evenly spaced intervals within a preset time interval. Each real-time imaging picture is from a fixed downward shooting mechanism at the current monitoring point, and the visualization data of each real-time imaging picture are the cyan component values, magenta component values, yellow component values, and black component values corresponding to each pixel point in the real-time imaging picture. At the same time, the water quality-related data at each moment of the target monitored water surface are the chemical oxygen demand, permanganate index, phosphorus element content, total content of minerals dissolved in water, and water temperature of the water body at the target monitored water surface at that moment. Particularly importantly, the number of moments at evenly spaced intervals within the preset time interval is positively correlated with the water surface area of the target monitored water surface. Thus, a number of comprehensive and sufficient basic information are specifically screened for the intelligent analysis of suspicious clue information, further ensuring the stability and effectiveness of the intelligent analysis results;

[0018] Inventive Point D: In each learning operation performed on the convolutional neural network, the pollution label indicating whether a certain monitored water surface belongs to a polluted water surface and the distance from the nearest pollution source near the certain monitored water surface to the monitoring point corresponding to the certain monitored water surface are used as two output contents of the convolutional neural network. The water surface area of the certain monitored water surface, the time length of the preset time interval, the water quality-related data of the certain monitored water surface corresponding to each moment at evenly spaced intervals within the preset time interval, and the visualization data of each real-time imaging picture of the certain monitored water surface corresponding to each moment at evenly spaced intervals within the preset time interval are used as multiple input contents of the convolutional neural network to complete the current learning operation, thus ensuring the learning effect of each learning operation of the convolutional neural network. Brief Description of the Drawings

[0019] The embodiments of the present invention will be described below with reference to the drawings, where:

[0020] Figure 1 is a technical flow chart of a data analysis method for automatically monitoring suspicious clues based on artificial intelligence according to the present invention;

[0021] Figure 2It is a step flowchart of a method for automatically monitoring and analyzing data of suspicious clues based on artificial intelligence shown in Embodiment 1 of the present invention;

[0022] Figure 3 It is a step flowchart of a method for automatically monitoring and analyzing data of suspicious clues based on artificial intelligence shown in Embodiment 2 of the present invention;

[0023] Figure 4 It is a step flowchart of a method for automatically monitoring and analyzing data of suspicious clues based on artificial intelligence shown in Embodiment 3 of the present invention;

[0024] Figure 5 It is a step flowchart of a method for automatically monitoring and analyzing data of suspicious clues based on artificial intelligence shown in Embodiment 4 of the present invention;

[0025] Figure 6 It is a step flowchart of a method for automatically monitoring and analyzing data of suspicious clues based on artificial intelligence shown in Embodiment 5 of the present invention. Detailed implementation manners

[0026] As Figure 1 shown, a technical flowchart of a method for automatically monitoring and analyzing data of suspicious clues based on artificial intelligence shown in the present invention is given.

[0027] As Figure 1 shown, the specific technical process of the present invention is as follows:

[0028] Technical process 1: For the target monitoring water surface directly below the current monitoring point, obtain various time-sharing visualization data of the target monitoring water surface and time-sharing water quality-related data of the target monitoring water surface, so as to provide sufficient and comprehensive multiple basic data for subsequent intelligent analysis of suspicious clues;

[0029] Exemplarily, the suspicious clues obtained by subsequent intelligent analysis include the pollution mark indicating whether the target monitoring water surface belongs to a polluted water surface and the suspected distance from the pollution source to the current monitoring point;

[0030] Specifically, the time-sharing visualization data of each item of the target monitored water surface are the visualization data of each real-time imaging picture corresponding to each moment evenly spaced within a preset time interval. Each real-time imaging picture is derived from a fixed downward shooting mechanism at the current monitoring point, and the visualization data of each real-time imaging picture are the cyan component values, magenta component values, yellow component values, and black component values corresponding to each pixel point in the real-time imaging picture. At the same time, the water quality-related data at each moment of the target monitored water surface are the chemical oxygen demand, permanganate index, phosphorus element content, total content of minerals dissolved in water, and water temperature of the water body at the target monitored water surface at that moment. Particularly crucial is that the number of moments evenly spaced within the preset time interval is positively correlated with the water surface area of the target monitored water surface, thereby screening a number of comprehensive and sufficient basic information for the intelligent analysis of suspicious clue information, further ensuring the stability and effectiveness of the intelligent analysis results;

[0031] As Figure 1 shown, the downward shooting mechanism at the current monitoring point is fixed by a fixing system including a vertical pole and a horizontal bar;

[0032] And as Figure 1 shown, a control box located on the vertical pole, a solar panel and a radar flow velocity meter located on the horizontal bar are also provided at the current monitoring point;

[0033] In this way, through the intelligent analysis of the change situation of the time-sharing dynamic data of the target monitored water surface, an intelligent and reliable judgment is provided for the environmental pollution degree of the target monitored water surface and the distance of the pollution source;

[0034] Technical process two: Specifically design a suspicious clue automatic monitoring model with a customized structure to perform the subsequent intelligent analysis of suspicious clues;

[0035] Specifically, the structural customization of the suspicious clue automatic monitoring model is mainly manifested in the following aspects:

[0036] First, the suspicious clue automatic monitoring model is a convolutional neural network after completing each learning operation;

[0037] Second, the number of learning operations completed by the convolutional neural network is positively correlated with the water surface area of the target monitored water surface, thereby constructing suspicious clue automatic monitoring models with different structures for monitored water surfaces of different areas, ensuring the stability and effectiveness of the intelligent analysis results;

[0038] Third, in each learning operation performed on the convolutional neural network, the known pollution mark indicating whether a certain monitored water surface is a polluted water surface and the determined distance from the nearest pollution source near the certain monitored water surface to the corresponding monitoring point of the certain monitored water surface are used as two output contents of the convolutional neural network, and the water surface area of ​​the certain monitored water surface, the time length of the preset time interval, the various water quality-related data of the certain monitored water surface corresponding to each time at even intervals within the preset time interval, and the visualization data of each real-time imaging picture of the certain monitored water surface corresponding to each time at even intervals within the preset time interval are used as multiple input contents of the convolutional neural network to complete this learning operation, thereby ensuring the learning effect of each learning operation of the convolutional neural network;

[0039] Technical process three: The automatic monitoring model for suspicious clues designed in technical process two is used to intelligently analyze the pollution mark of the target monitoring water surface and the suspected distance from the pollution source to the current monitoring point based on the various time-sharing visual data of the target monitoring water surface selected in technical process one and the time-sharing water quality-related data of the target monitoring water surface;

[0040] In this way, intelligent analysis and automatic monitoring of suspicious clue information of water surface environmental pollution based on time-sharing dynamic information including pollution identification and suspected distance from pollution source to current monitoring point are completed;

[0041] Technical process 4: Display and wirelessly report suspicious clues obtained by intelligent analysis in technical process 3, including whether the target monitored water surface is a polluted water surface and the suspected distance from the pollution source to the current monitoring point;

[0042] Specifically, a mobile communication interface at the current monitoring point is used to input the pollution mark indicating whether the target monitoring water surface is a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model into a network data packet, and then the network data packet is sent to a remote environmental monitoring server through a mobile communication link;

[0043] Specifically, the on-site display screen at the current monitoring point is used to receive and display in real time the pollution mark output by the suspicious clue automatic monitoring model indicating whether the target monitored water surface is a polluted water surface and the suspected distance from the pollution source to the current monitoring point.

[0044] The key points of the present invention are: targeted screening of dynamic basic data including various time-sharing visualization data of the target monitored water surface and time-sharing water quality-related data of the target monitored water surface, customized structural design of the automatic monitoring model of suspicious clues, and intelligent analysis of suspicious clues including pollution identification of whether the target monitored water surface is a polluted water surface and the suspected distance from the pollution source to the current monitoring point.

[0045] Next, a specific description will be given of the method for automatically monitoring and analyzing data of suspicious clues based on artificial intelligence according to the present invention by way of examples.

[0046] Example 1

[0047] Figure 2 It is a flowchart of steps of a method for automatically monitoring and analyzing data of suspicious clues based on artificial intelligence shown in Example 1 according to the present invention.

[0048] As Figure 2 shown, the method for automatically monitoring and analyzing data of suspicious clues based on artificial intelligence includes the following specific steps:

[0049] Step S201: Collect water quality-related data corresponding to each moment of the target monitored water surface directly below the current monitoring point. The water quality-related data corresponding to each moment of the target monitored water surface is the chemical oxygen demand, permanganate index, phosphorus element content, total content of minerals dissolved in water, and water temperature of the target monitored water surface at that moment;

[0050] Specifically, the chemical oxygen demand is an index to measure the content of reducing substances in water. The amount of oxidant consumed by these substances reacting with strong oxidants under specific conditions. A high COD value indicates a high content of organic matter in water, which may cause environmental pollution;

[0051] Specifically, the permanganate index is calculated by the amount consumed by potassium permanganate to oxidize certain organic and inorganic reducing substances in the water sample, and it reflects the degree of pollution of the water by reducing substances;

[0052] Specifically, the phosphorus element content, that is, the total phosphorus, is the total content of phosphorus elements in water, including dissolved and particulate phosphorus. A high phosphorus content may lead to eutrophication of the water body, causing a large number of algae to multiply and affecting water quality;

[0053] And specifically, the total dissolved solids, that is, the total content of minerals dissolved in water. A too high value of the total dissolved solids indicates a large amount of minerals in the water, which may affect the taste and health of the water quality;

[0054] Step S202: Capture the visualized data of the real-time imaging picture corresponding to the target monitored water surface directly below the current monitoring point. The visualized data of the real-time imaging picture are the respective cyan component values, respective magenta component values, respective yellow component values, and respective black component values corresponding to each pixel point of the real-time imaging picture;

[0055] Step S203: Perform a set number of learning actions on the convolutional neural network to obtain the convolutional neural network after each learning operation, and use the convolutional neural network after each learning operation as the output of the suspicious clue automatic monitoring model. The number of learning operations completed by the convolutional neural network is positively correlated with the water surface area of the target monitored water surface;

[0056] Exemplarily, the fact that the number of learning operations completed by the convolutional neural network is positively correlated with the water surface area of the target monitored water surface includes: when the water surface area of the target monitored water surface is 20 square meters, the number of learning operations completed by the convolutional neural network is 30 times; when the water surface area of the target monitored water surface is 50 square meters, the number of learning operations completed by the convolutional neural network is 50 times; when the water surface area of the target monitored water surface is 100 square meters, the number of learning operations completed by the convolutional neural network is 70 times, and so on;

[0057] Step S204: Synchronously input the water surface area of the target monitored water surface, the time length of the preset time interval, each piece of water quality-related data corresponding to each equally spaced moment within the preset time interval, and the visualization data of each real-time imaging picture corresponding to each equally spaced moment within the preset time interval into the suspicious clue automatic monitoring model to run the suspicious clue automatic monitoring model, and obtain the pollution flag indicating whether the target monitored water surface belongs to a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model;

[0058] Exemplarily, synchronously inputting the water surface area of the target monitored water surface, the time length of the preset time interval, each piece of water quality-related data corresponding to each equally spaced moment within the preset time interval, and the visualization data of each real-time imaging picture corresponding to each equally spaced moment within the preset time interval into the suspicious clue automatic monitoring model to run the suspicious clue automatic monitoring model, and obtaining the pollution flag indicating whether the target monitored water surface belongs to a polluted water surface and the suspected distance from the pollution source to the current monitoring point includes: the preset time interval is the time interval between 9:00 am and 10:00 am on the same day, and each equally spaced moment within the preset time interval is a 10-minute time point within the time interval between 9:00 am and 10:00 am on the same day, that is, 9:10 am, 9:20 am, 9:30 am, 9:40 am, 9:50 am, and 10:00 am on the same day;

[0059] Among them, the water quality-related data of the target monitored water surface directly below the current monitoring point is sourced from the water body at the target monitored water surface directly below the current monitoring point;

[0060] Specifically, the water quality-related data of the target monitored water surface directly below the current monitoring point is sourced from the water body at the target monitored water surface directly below the current monitoring point, including: the water quality-related data of the target monitored water surface directly below the current monitoring point is all taken from the water body at the target monitored water surface directly below the current monitoring point;

[0061] Among them, when the pollution flag output by the suspicious clue automatic monitoring model indicates that the target monitored water surface is not a polluted water surface, the suspected distance from the pollution source to the current monitoring point output simultaneously is zero;

[0062] In this way, when the pollution flag output by the suspicious clue automatic monitoring model indicates that the target monitored water surface is a polluted water surface, the suspected distance from the pollution source to the current monitoring point output simultaneously is the actual distance from the pollution source to the current monitoring point that needs to be confirmed later;

[0063] Among them, performing a set number of learning actions on the convolutional neural network to obtain the convolutional neural network after each learning operation, and using the convolutional neural network after each learning operation as the output of the suspicious clue automatic monitoring model. The number of learning operations completed by the convolutional neural network is positively correlated with the water surface area of the target monitored water surface, including: in each learning operation performed on the convolutional neural network, taking the pollution flag indicating whether a certain monitored water surface is a polluted water surface and the determined distance from the nearest pollution source near the certain monitored water surface to the corresponding monitoring point of the certain monitored water surface as the two output contents of the convolutional neural network, and taking the water surface area of the certain monitored water surface, the time length of the preset time interval, the water quality-related data of each part of the certain monitored water surface corresponding to each evenly spaced moment within the preset time interval, and the visualized data of each real-time imaging picture of the certain monitored water surface corresponding to each evenly spaced moment within the preset time interval as the multiple input contents of the convolutional neural network, and completing the current learning operation;

[0064] Among them, synchronously inputting the water surface area of the target monitored water surface, the time length of the preset time interval, the water quality-related data of each part corresponding to each evenly spaced moment within the preset time interval, and the visualized data of each real-time imaging picture corresponding to each evenly spaced moment within the preset time interval into the suspicious clue automatic monitoring model, including: the number of evenly spaced moments within the preset time interval is positively correlated with the water surface area of the target monitored water surface;

[0065] Exemplarily, the number of evenly spaced moments within a preset time interval being positively correlated with the water surface area of the target monitored water surface includes: when the water surface area of the target monitored water surface is 20 square meters, the number of evenly spaced moments within the preset time interval is 6; when the water surface area of the target monitored water surface is 50 square meters, the number of evenly spaced moments within the preset time interval is 12; when the water surface area of the target monitored water surface is 100 square meters, the number of evenly spaced moments within the preset time interval is 18, and so on;

[0066] Wherein, water quality-related data corresponding to each moment of the target monitored water surface directly below the current monitoring point is collected. The water quality-related data corresponding to each moment of the target monitored water surface is the total sum of the chemical oxygen demand, permanganate index, phosphorus element content, dissolved mineral content in water, and water temperature of the target monitored water surface at that moment, including: the permanganate index of the target monitored water surface represents the degree of pollution of the water body at the target monitored water surface by reducing substances;

[0067] Wherein, visualization data of a real-time imaging picture corresponding to the target monitored water surface directly below the current monitoring point is captured. The visualization data of the real-time imaging picture is the respective cyan component values, magenta component values, yellow component values, and black component values corresponding to each pixel point of the real-time imaging picture, including: the real-time imaging picture corresponding to the target monitored water surface directly below the current monitoring point is a real-time picture captured by a fixed downward shooting mechanism at the current monitoring point;

[0068] And wherein, visualization data of a real-time imaging picture corresponding to the target monitored water surface directly below the current monitoring point is captured. The visualization data of the real-time imaging picture is the respective cyan component values, magenta component values, yellow component values, and black component values corresponding to each pixel point of the real-time imaging picture, including: the cyan component value, magenta component value, yellow component value, and black component value corresponding to each pixel point of the real-time imaging picture are the C component value, M component value, Y component value, and K component value of the pixel point in the CMYK color space.

[0069] Embodiment 2

[0070] Figure 3 It is a step flowchart of an automatic monitoring data analysis method for suspicious clues based on artificial intelligence shown in Embodiment 2 of the present invention.

[0071] As Figure 3 shown, and Figure 2Different from the embodiment in, before capturing the visualization data of the real-time imaging picture corresponding to the target monitoring water surface directly below the current monitoring point, the visualization data of the real-time imaging picture being each cyan component value, each magenta component value, each yellow component value and each black component value corresponding to each pixel point of the real-time imaging picture, that is, before step S202, the method further includes:

[0072] Step S301: using a fixed overhead imaging mechanism at the current monitoring point to perform an overhead imaging operation on the target monitoring water surface to obtain and output a corresponding real-time imaging picture;

[0073] Specifically, a fixed overhead camera mechanism at the current monitoring point is used to perform an overhead camera operation on the target monitored water surface to obtain and output a corresponding real-time imaging picture, including: the overhead camera mechanism includes a CMOS sensor, an overhead camera lens, a flexible circuit board and a filter.

[0074] Example 3

[0075] Figure 4 The present invention is a flowchart showing a method for automatically monitoring suspicious clue data and analyzing it based on artificial intelligence according to Embodiment 3 of the present invention.

[0076] like Figure 4 As shown, Figure 2 Different from the embodiment in, after the surface area of ​​the target monitored water surface, the time length of the preset time interval, the respective water quality related data corresponding to the respective moments evenly spaced within the preset time interval, and the visual data of the respective real-time imaging pictures corresponding to the respective moments evenly spaced within the preset time interval are synchronously input into the suspicious clue automatic monitoring model to run the suspicious clue automatic monitoring model, and obtain the pollution mark indicating whether the target monitored water surface is a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model, that is, after step S204, the method further includes:

[0077] Step S401: using the mobile communication interface at the current monitoring point to input the pollution mark indicating whether the target monitoring water surface is a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model into a network data packet, and then sending the network data packet to a remote environmental monitoring server through a mobile communication link;

[0078] Specifically, the mobile communication interface at the current monitoring point is used to package the pollution identification indicating whether the target monitored water surface belongs to a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model into a network data packet, and then send the network data packet to the remote environmental monitoring server through the mobile communication link. The environmental monitoring server includes one of a big data monitoring network element, a cloud computing monitoring network element, or a blockchain monitoring network element.

[0079] Embodiment 4

[0080] Figure 5 It is a step flowchart of a method for automatically monitoring and analyzing suspicious clues based on artificial intelligence shown in Embodiment 4 of the present invention.

[0081] As Figure 5 shown, different from the embodiments in Figure 2 after synchronously inputting the water surface area of the target monitored water surface, the time length of the preset time interval, each piece of water quality-related data corresponding to each moment evenly spaced within the preset time interval, and the visualization data of each real-time imaging screen corresponding to each moment evenly spaced within the preset time interval into the suspicious clue automatic monitoring model to run the suspicious clue automatic monitoring model, and obtaining the pollution identification indicating whether the target monitored water surface belongs to a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model, that is, after step S204, the method further includes:

[0082] Step S501: Use the on-site display screen at the current monitoring point to receive and real-time display the pollution identification indicating whether the target monitored water surface belongs to a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model

[0083] Specifically, using the on-site display screen at the current monitoring point to receive and real-time display the pollution identification indicating whether the target monitored water surface belongs to a polluted water surface and the suspected distance from the pollution source to the current monitoring point includes: the on-site display screen is an LCD display array or an LED display array.

[0084] Embodiment 5

[0085] Figure 6 It is a step flowchart of a method for automatically monitoring and analyzing suspicious clues based on artificial intelligence shown in Embodiment 5 of the present invention.

[0086] As Figure 6 shown, different from Figure 2Different from the embodiments, after performing a set number of learning actions on the convolutional neural network to obtain the convolutional neural network after each learning operation, and using the convolutional neural network after each learning operation as the suspicious clue automatic monitoring model output, after the number of learning operations completed by the convolutional neural network is positively correlated with the water surface area of the target monitored water surface, that is, after step S203, the method further includes:

[0087] Step S601: Store various model parameters of the suspicious clue automatic monitoring model to implement model storage of the suspicious clue automatic monitoring model;

[0088] Exemplarily, storing various model parameters of the suspicious clue automatic monitoring model to implement model storage of the suspicious clue automatic monitoring model includes: It can be selected to use an SD storage device, a TF storage device, or an MMC storage device to store various model parameters of the suspicious clue automatic monitoring model to implement model storage of the suspicious clue automatic monitoring model.

[0089] Next, continue to describe each method embodiment of the present invention in detail.

[0090] In the method for analyzing suspicious clue automatic monitoring data based on artificial intelligence according to each method embodiment of the present invention:

[0091] Synchronously input the water surface area of the target monitored water surface, the time length of the preset time interval, each piece of water quality-related data corresponding to each moment evenly spaced within the preset time interval, and the visualization data of each real-time imaging screen corresponding to each moment evenly spaced within the preset time interval into the suspicious clue automatic monitoring model to run the suspicious clue automatic monitoring model, and obtain the pollution identification indicating whether the target monitored water surface belongs to a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model, including: After performing binary numerical conversion on the water surface area of the target monitored water surface, the time length of the preset time interval, each piece of water quality-related data corresponding to each moment evenly spaced within the preset time interval, and the visualization data of each real-time imaging screen corresponding to each moment evenly spaced within the preset time interval, synchronously input them into the suspicious clue automatic monitoring model;

[0092] Specifically, after performing binary numerical conversion on the water surface area of the target monitored water surface, the time length of the preset time interval, each piece of water quality-related data corresponding to each moment evenly spaced within the preset time interval, and the visualization data of each real-time imaging picture corresponding to each moment evenly spaced within the preset time interval, and then synchronously inputting them into the suspicious clue automatic monitoring model includes: selecting a synchronous driving mechanism implemented by a programmable logic device to achieve the synchronous input;

[0093] Synchronously inputting the water surface area of the target monitored water surface, the time length of the preset time interval, each piece of water quality-related data corresponding to each moment evenly spaced within the preset time interval, and the visualization data of each real-time imaging picture corresponding to each moment evenly spaced within the preset time interval into the suspicious clue automatic monitoring model to run the suspicious clue automatic monitoring model, and obtaining the pollution identification indicating whether the target monitored water surface belongs to a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model further includes: the pollution identification indicating whether the target monitored water surface belongs to a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model are both in the form of binary numerical representation.

[0094] In the method for analyzing suspicious clue automatic monitoring data based on artificial intelligence according to various method embodiments of the present invention:

[0095] Performing a set number of learning actions on the convolutional neural network to obtain the convolutional neural network after each learning operation is completed, and using the convolutional neural network after each learning operation is completed as the output of the suspicious clue automatic monitoring model. The number of learning operations completed by the convolutional neural network is positively correlated with the water surface area of the target monitored water surface, including: using a numerical mapping formula to represent the numerical mapping relationship in which the number of learning operations completed by the convolutional neural network is positively correlated with the water surface area of the target monitored water surface;

[0096] Specifically, it is possible to select to use the MATLAB toolbox to implement the test and simulation of the data processing process of using a numerical mapping formula to represent the numerical mapping relationship in which the number of learning operations completed by the convolutional neural network is positively correlated with the water surface area of the target monitored water surface;

[0097] Among them, using a numerical mapping formula to represent the numerical mapping relationship in which the number of learning operations completed by the convolutional neural network is positively correlated with the water surface area of the target monitored water surface includes: in the numerical mapping formula, the water surface area of the target monitored water surface is the input value of the numerical mapping formula, and the number of learning operations completed by the convolutional neural network corresponding to the water surface area of the target monitored water surface is the output value of the numerical mapping formula.

[0098] and in the method for automatically monitoring and analyzing data of suspicious clues based on artificial intelligence according to various method embodiments of the present invention:

[0099] The cyan component value, magenta component value, yellow component value, and black component value corresponding to each pixel point of the real-time imaging screen are the C component value, M component value, Y component value, and K component value of the pixel point in the CMYK color space, including: the value range of any one of the cyan component value, magenta component value, yellow component value, and black component value corresponding to each pixel point of the real-time imaging screen is between 0 and 255.

[0100] In addition, the present invention can also cite the following technical content to highlight the significant technical progress of the present invention:

[0101] Collect the water quality-related data corresponding to each moment of the target monitoring water surface directly below the current monitoring point. The water quality-related data corresponding to each moment of the target monitoring water surface is the chemical oxygen demand, permanganate index, phosphorus element content, total content of minerals dissolved in water, and water temperature of the target monitoring water surface at that moment, including: using a plurality of real-time measurement components arranged at the current monitoring point, when each moment arrives, for respectively measuring the chemical oxygen demand, permanganate index, phosphorus element content, total content of minerals dissolved in water, and water temperature of the target monitoring water surface corresponding to that moment;

[0102] Specifically, using a plurality of real-time measurement components arranged at the current monitoring point, when each moment arrives, for respectively measuring the chemical oxygen demand, permanganate index, phosphorus element content, total content of minerals dissolved in water, and water temperature of the target monitoring water surface corresponding to that moment, including: a timing service component and a synchronous driving component are also arranged at the current monitoring point. The timing service component is used to determine whether each moment arrives. The synchronous driving component is connected to the timing service component and is also connected to the plurality of real-time measurement components, and is used to drive the plurality of real-time measurement components to respectively measure the chemical oxygen demand, permanganate index, phosphorus element content, total content of minerals dissolved in water, and water temperature of the target monitoring water surface corresponding to that moment when each moment arrives;

[0103] And among them, capture the visualization data of the real-time imaging screen corresponding to the target monitoring water surface directly below the current monitoring point. The visualization data of the real-time imaging screen is the respective cyan component values, respective magenta component values, respective yellow component values, and respective black component values corresponding to each pixel point of the real-time imaging screen, including: keeping the downward shooting lens of the fixed downward shooting mechanism at the current monitoring point in the horizontal plane.

[0104] The embodiments of the present invention have been described above in conjunction with the accompanying drawings. However, the present invention is not limited to the above specific embodiments. The above specific embodiments are merely illustrative rather than restrictive. Under the inspiration of the present invention, those of ordinary skill in the art can also make many forms without departing from the spirit of the present invention and the scope protected by the claims, and all of them fall within the protection scope of the present invention.

Claims

1. A suspicious clue automatic monitoring data analysis method based on artificial intelligence, characterized in that: The method comprises: Collecting water quality related data corresponding to the target monitoring water surface directly below the current monitoring point at each moment, wherein the water quality related data corresponding to the target monitoring water surface at each moment is the chemical oxygen demand, permanganate index, phosphorus content, total content of minerals dissolved in water, and water body temperature of the target monitoring water surface at the moment; Capturing visualization data of a real-time imaging picture corresponding to a target monitoring water surface directly below a current monitoring point, wherein the visualization data of the real-time imaging picture is each cyan component value, each magenta component value, each yellow component value, and each black component value corresponding to each pixel point of the real-time imaging picture; Performing a set number of learning actions on the convolutional neural network to obtain a convolutional neural network after completing each learning operation, and outputting the convolutional neural network after completing each learning operation as an automatic monitoring model for suspicious clues, wherein the number of learning operations completed by the convolutional neural network is positively correlated with the water surface area of ​​the target monitored water surface; The water surface area of ​​the target monitored water surface, the time length of the preset time interval, the water quality related data corresponding to each time moment evenly spaced within the preset time interval, and the visualization data of each real-time imaging picture corresponding to each time moment evenly spaced within the preset time interval are synchronously input into the suspicious clue automatic monitoring model to run the suspicious clue automatic monitoring model to obtain the pollution mark indicating whether the target monitored water surface is a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model; Among them, the water quality related data of the target monitoring water surface directly below the current monitoring point originates from the water body at the target monitoring water surface directly below the current monitoring point; Wherein, when the pollution mark output by the suspicious clue automatic monitoring model indicates that the target monitored water surface is not a polluted water surface, the suspected distance outputted from the pollution source to the current monitoring point is zero; Wherein, the overhead camera lens of the overhead camera mechanism fixed at the current monitoring point is kept in the horizontal plane; Wherein, the number of evenly spaced moments within the preset time interval is positively correlated with the water surface area of ​​the target monitored water surface; Among them, in each learning operation performed on the convolutional neural network, a known pollution mark indicating whether a certain monitored water surface is a polluted water surface and a determined distance from the nearest pollution source near the certain monitored water surface to the corresponding monitoring point of the certain monitored water surface are used as two output contents of the convolutional neural network, and the water surface area of ​​the certain monitored water surface, the time length of a preset time interval, various water quality-related data of the certain monitored water surface corresponding to each moment evenly spaced within the preset time interval, and various visualization data of real-time imaging pictures of the certain monitored water surface corresponding to each moment evenly spaced within the preset time interval are used as multiple input contents of the convolutional neural network.

2. The method for automatically monitoring suspicious clues based on artificial intelligence as claimed in claim 1, characterized in that: Collecting water quality related data corresponding to the target monitoring water surface directly below the current monitoring point at each moment, wherein the water quality related data corresponding to the target monitoring water surface at each moment are the chemical oxygen demand, permanganate index, phosphorus content, the total content of minerals dissolved in water, and water body temperature of the target monitoring water surface at the moment, including: the permanganate index of the target monitoring water surface indicates the degree of water pollution by reducing substances at the target monitoring water surface; Wherein, the visualization data of the real-time imaging picture corresponding to the target monitoring water surface directly below the current monitoring point is captured, and the visualization data of the real-time imaging picture is each cyan component value, each magenta component value, each yellow component value and each black component value respectively corresponding to each pixel point of the real-time imaging picture, including: the real-time imaging picture corresponding to the target monitoring water surface directly below the current monitoring point is the real-time picture taken by the overhead imaging mechanism fixed at the current monitoring point; Among them, the visualization data of the real-time imaging picture corresponding to the target monitoring water surface directly below the current monitoring point is captured, and the visualization data of the real-time imaging picture is each cyan component value, each magenta component value, each yellow component value and each black component value corresponding to each pixel point of the real-time imaging picture, including: the cyan component value, magenta component value, yellow component value and black component value corresponding to each pixel point of the real-time imaging picture are the C component value, M component value, Y component value and K component value of the pixel point in the CMYK color space.

3. The method for automatically monitoring suspicious clues based on artificial intelligence as claimed in claim 2 is characterized in that: Before capturing visualization data of a real-time imaging picture corresponding to a target monitoring water surface directly below a current monitoring point, wherein the visualization data of the real-time imaging picture is each cyan component value, each magenta component value, each yellow component value, and each black component value corresponding to each pixel point of the real-time imaging picture, the method further includes: A fixed overhead camera mechanism at the current monitoring point is used to perform an overhead camera operation on the target monitoring water surface to obtain and output the corresponding real-time imaging picture.

4. The method for automatically monitoring suspicious clues based on artificial intelligence as claimed in claim 2 is characterized in that: After synchronously inputting the water surface area of ​​the target monitored water surface, the time length of the preset time interval, the respective water quality related data corresponding to the respective moments evenly spaced within the preset time interval, and the respective visualization data of the respective real-time imaging pictures corresponding to the respective moments evenly spaced within the preset time interval into the suspicious clue automatic monitoring model to run the suspicious clue automatic monitoring model, and obtaining the pollution mark indicating whether the target monitored water surface is a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model, the method further includes: The mobile communication interface at the current monitoring point is used to input the pollution mark indicating whether the target monitored water surface is a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model into a network data packet, and then the network data packet is sent to a remote environmental monitoring server through a mobile communication link.

5. The method for automatically monitoring suspicious clues based on artificial intelligence as claimed in claim 2 is characterized in that: After synchronously inputting the water surface area of ​​the target monitored water surface, the time length of the preset time interval, the respective water quality related data corresponding to the respective moments evenly spaced within the preset time interval, and the respective visualization data of the respective real-time imaging pictures corresponding to the respective moments evenly spaced within the preset time interval into the suspicious clue automatic monitoring model to run the suspicious clue automatic monitoring model, and obtaining the pollution mark indicating whether the target monitored water surface is a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model, the method further includes: The on-site display screen at the current monitoring point is used to receive and display in real time the pollution mark output by the suspicious clue automatic monitoring model indicating whether the target monitored water surface is a polluted water surface and the suspected distance from the pollution source to the current monitoring point.

6. The method for automatically monitoring suspicious clues based on artificial intelligence as claimed in claim 2, characterized in that: After performing a set number of learning actions on the convolutional neural network to obtain a convolutional neural network after completing each learning operation, and outputting the convolutional neural network after completing each learning operation as a suspicious clue automatic monitoring model, and the number of learning operations completed by the convolutional neural network is positively correlated with the water surface area of ​​the target monitored water surface, the method further includes: The model storage of the suspicious clue automatic monitoring model is realized by storing various model parameters of the suspicious clue automatic monitoring model.

7. The method for automatically monitoring suspicious clues based on artificial intelligence according to any one of claims 2 to 6, characterized in that: The water surface area of ​​the target monitored water surface, the time length of the preset time interval, the water quality related data corresponding to each moment evenly spaced within the preset time interval, and the visual data of each real-time imaging screen corresponding to each moment evenly spaced within the preset time interval are synchronously input into the suspicious clue automatic monitoring model to run the suspicious clue automatic monitoring model, and obtain the pollution mark indicating whether the target monitored water surface is a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model, including: performing binary value conversion on the water surface area of ​​the target monitored water surface, the time length of the preset time interval, the water quality related data corresponding to each moment evenly spaced within the preset time interval, and the visual data of each real-time imaging screen corresponding to each moment evenly spaced within the preset time interval, and then synchronously inputting them into the suspicious clue automatic monitoring model; The water surface area of ​​the target monitored water surface, the time length of the preset time interval, the water quality-related data corresponding to each moment evenly spaced within the preset time interval, and the visualization data of each real-time imaging picture corresponding to each moment evenly spaced within the preset time interval are synchronously input into the suspicious clue automatic monitoring model to run the suspicious clue automatic monitoring model, and obtain the pollution mark indicating whether the target monitored water surface is a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model. It also includes: the pollution mark indicating whether the target monitored water surface is a polluted water surface and the suspected distance from the pollution source to the current monitoring point output by the suspicious clue automatic monitoring model are both in the form of binary numerical values.

8. The method for automatically monitoring suspicious clues based on artificial intelligence as claimed in any one of claims 2 to 6, characterized in that: Performing a set number of learning actions on the convolutional neural network to obtain a convolutional neural network after completing each learning operation, and outputting the convolutional neural network after completing each learning operation as a suspicious clue automatic monitoring model, wherein the number of learning operations completed by the convolutional neural network is positively correlated with the water surface area of ​​the target monitored water surface, including: using a numerical mapping formula to represent a numerical mapping relationship of the number of learning operations completed by the convolutional neural network and the water surface area of ​​the target monitored water surface; Among them, the numerical mapping relationship in which the number of learning operations completed by the convolutional neural network and the water surface area of ​​the target monitored water surface are expressed by a numerical mapping formula includes: in the numerical mapping formula, the water surface area of ​​the target monitored water surface is the input value of the numerical mapping formula, and the number of learning operations completed by the convolutional neural network corresponding to the water surface area of ​​the target monitored water surface is the output value of the numerical mapping formula.

9. The method for automatically monitoring suspicious clues based on artificial intelligence as claimed in any one of claims 2 to 6, characterized in that: The cyan component value, magenta component value, yellow component value and black component value corresponding to each pixel point of the real-time imaging picture are the C component value, M component value, Y component value and K component value of the pixel point in the CMYK color space, including: the value range of any color component value among the cyan component value, magenta component value, yellow component value and black component value corresponding to each pixel point of the real-time imaging picture is between 0-255.

Images