Industrial park environment monitoring method and system based on internet of things

Abstract

The application discloses an industrial park environment monitoring method and system based on the Internet of Things, and relates to the technical field of scene monitoring.The application establishes a scene monitoring space through the historical production records of a target industrial park, sets a dynamic limited range according to the historical production records of the target industrial park, and sets a patrol cycle route in the scene monitoring space, and then a UAV collects real-time environment data of the target industrial park through the patrol cycle route, and performs data handshake with a ground sensing device, completes data fusion according to the data handshake result, and updates the scene monitoring space, divides the scene monitoring space into a plurality of scene areas, positions a scene area with an anomaly in the scene monitoring space according to the dynamic limited range, establishes an anomaly spreading arc area according to the scene area with the anomaly and adjacent scene areas, and then evaluates the anomaly spreading direction according to the anomaly spreading arc area until the scene area with the anomaly disappears.

Description

Technical Field

[0001] This invention relates to the field of scene monitoring technology, specifically to an IoT-based method and system for environmental monitoring in industrial parks. Background Technology

[0002] With the acceleration of industrialization, the scale and number of industrial parks are constantly increasing, and their production activities are having an increasingly significant impact on the environment. Production processes within industrial parks often involve substantial energy consumption, waste gas emissions, and wastewater treatment, which may pose potential threats to air quality, water quality, soil quality, and other environmental factors within and around the park. Timely and accurate monitoring of the environmental conditions of industrial parks is crucial for ensuring their sustainable development, protecting the surrounding ecological environment, and safeguarding the health of residents.

[0003] Traditional industrial park environmental monitoring methods primarily rely on fixed-location ground-based monitoring equipment. While these devices can provide some environmental data, they suffer from limited monitoring range and cannot provide real-time, comprehensive coverage of every corner of the park. Furthermore, industrial park production activities are dynamic and complex; different time periods and production stages may generate different types and degrees of environmental impact, which traditional static monitoring methods struggle to adapt to. In addition, when environmental anomalies occur within the park, traditional methods often fail to quickly and accurately locate the abnormal area and assess the direction and extent of its spread, thus hindering timely and effective countermeasures. Therefore, this paper proposes an IoT-based industrial park environmental monitoring method and system. Summary of the Invention

[0004] The purpose of this invention is to provide an IoT-based method and system for environmental monitoring in industrial parks, in order to address the shortcomings in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] An IoT-based method for environmental monitoring in industrial parks includes the following steps:

[0007] Step S1: Set up ground sensing devices and drones in the target industrial park to acquire and establish a scene monitoring space based on the historical production records of the target industrial park;

[0008] Step S2: Set a dynamic limit range based on the historical production records of the target industrial park, and set an inspection loop route in the scene monitoring space. Then, the drone collects real-time environmental data of the target industrial park through the inspection loop route, and performs data handshake with the ground sensing device. Based on the data handshake result, complete data fusion and update the scene monitoring space.

[0009] Step S3: Divide the scene monitoring space into several scene areas, locate the scene areas with anomalies in the scene monitoring space according to the dynamically limited range, establish anomaly propagation arc area based on the scene areas with anomalies and their adjacent scene areas, and then evaluate the direction of anomaly propagation based on the anomaly propagation arc area until the scene areas with anomalies disappear.

[0010] Furthermore, the process of setting up ground sensing devices and drones includes:

[0011] Ground sensing devices and drones are set up in the target industrial park, and a fixed data collection area is set for each ground sensing device and drone, so that the data collection areas of ground sensing devices in adjacent spatial locations partially overlap.

[0012] Furthermore, the process of establishing a scenario monitoring space based on the historical production records of the target industrial park includes:

[0013] The target industrial park consists of several production areas. The historical production records of the target industrial park are obtained. The historical production records include historical change records of various historical environmental data in the production process of each production area within the target industrial park.

[0014] A scene monitoring space is established based on the scene distribution of the target industrial park. At the same time, multiple scene monitoring areas are marked in the scene monitoring space according to the production area distribution. In addition, a pipeline visualization image is generated in the scene monitoring space according to the pipeline distribution of the target industrial park.

[0015] In the scene monitoring space, n ground sensing nodes and m aerial sensing nodes are set up, and data channels are established between each ground sensing node and each aerial sensing node and each ground sensing device and UAV in sequence. Each ground sensing node and aerial sensing node marks the scene area associated with the real-time data collection area in the scene monitoring space.

[0016] Furthermore, the process of setting the dynamic limit range includes:

[0017] Divide a day into several time segments and set a data density frame. The data density frame is a rectangular data frame with a length equal to the duration of the time segment and a width of a preset value.

[0018] Starting from the first data segment in each time interval along the y-axis, the data segments in the time interval are selected and traversed by using a data density box, and the number of data segments selected is counted each time a data segment is selected.

[0019] Select the data segment with the largest number of selected data segments to form the dynamic limit range under the corresponding time segment. Then, stitch the dynamic limit ranges under each time segment together in sequence and mark the stitching result in the corresponding scene monitoring area of ​​the scene monitoring space.

[0020] Furthermore, the process of setting up the inspection cycle route includes:

[0021] Taking the scene monitoring area in the scene monitoring space as the inspection target, select any two spatial boundaries of the scene monitoring space to set m inspection starting points and inspection transfer points, take the diameter of the data collection area of ​​the UAV as the width of the inspection loop route, and then take the overlap of the data collection area radius length of any two inspection loop routes as the first condition, and take the inspection target as the priority target and avoidance of obstacles in the target industrial park as the second condition, and set m inspection loop routes.

[0022] All inspection loop routes are retrieved and inspection flight simulations are performed in the scene monitoring space. It is determined whether the inspection area composed of each inspection loop route covers the entire scene monitoring space. If it is determined that the entire scene monitoring space is covered, the inspection loop routes are assigned to each drone for execution.

[0023] If it is determined that the entire scene monitoring space is not covered, the uncovered part is marked on the scene monitoring space. Then, on the premise of ensuring that there are no gaps between adjacent inspection loop routes, the nearest inspection loop route to the uncovered part is partially shifted so that the uncovered part is covered by the inspection loop route.

[0024] If the condition that the uncovered part is not covered by the inspection loop is not met when the most recent inspection loop route is moved to the next location, the adjacent inspection loop route of the most recent inspection loop route will be retrieved again for translation. The above operation of retrieving inspection loop routes for translation will be repeated until the uncovered part is covered by the inspection loop route, or until there is no inspection loop route to choose from for translation.

[0025] Furthermore, the data handshake implementation process includes:

[0026] The same data update cycle is set for each ground sensing node and air sensing node, and then each UAV collects real-time environmental data within the data collection area through various sensors along the inspection cycle route and the ground sensing device.

[0027] After each data update cycle, the scene monitoring space dynamically updates the spatial position of each ground sensing node and the aerial sensing node in the scene monitoring space based on the real-time location information sent by each ground sensing device and drone.

[0028] Then, based on the associated scene areas marked by the ground sensing nodes and the air sensing nodes, it is determined in real time whether there is an overlap in the data collection area between the ground sensing nodes and the air sensing nodes. If it is determined that there is no overlap, no operation is performed.

[0029] When it is determined that there is overlap, real-time environmental data of the same type collected by the corresponding ground sensing node and air sensing node are retrieved based on the overlapping spatial portion.

[0030] The laser imaging videos of ground sensing nodes and air sensing nodes are matched in time sequence. Based on the content matching results, the common parts of the two are found. Then, the laser imaging videos of the two are stitched together based on the common parts. A three-dimensional image model is generated based on the stitched laser imaging video and overlaid on the scene monitoring space. Other types of environmental data are first spatiotemporally aligned according to the docking relationship of the laser imaging videos, and then overlaid on the three-dimensional image model in the form of texture.

[0031] The scene areas associated with ground sensing nodes and air sensing nodes are divided into several data display areas of the same size. Real-time environmental data of the same type acquired by ground sensing nodes and air sensing nodes are mapped to the corresponding data display areas according to their spatiotemporal location.

[0032] Both ground sensing nodes and air sensing nodes are initially assigned a trust weight of 0.5. A deviation threshold is set for each type of real-time environmental data. The relationship between the difference and the deviation threshold of the same type of real-time environmental data in each data display area is determined. Based on the determination result, the real-time environmental data collected by the ground sensing nodes and air sensing nodes in adjacent spatial locations are retrieved respectively. Then, it is determined again whether the difference of the corresponding type of real-time environmental data between the ground sensing node and the ground sensing nodes in adjacent spatial locations is greater than or equal to the deviation threshold.

[0033] If either party judges the difference to be greater than or equal to the deviation threshold, and the other party judges the difference to be less than the deviation threshold, then the trust weight of the perception node whose judgment difference is greater than or equal to the deviation threshold will be reduced by 10%, and the trust weight of the other party's perception node will be increased by the amount reduced by the other party.

[0034] If both parties agree, no trust weight adjustment will be performed.

[0035] After all data display areas have been judged, the real-time environmental data of the same type in the data display areas are merged according to the trust weight of both parties. The fusion formula is: A=αa+βb, where α and β represent the trust weight of both parties, A is the value of the real-time environmental data after fusion, and a and b are the values ​​of the real-time environmental data before fusion.

[0036] Furthermore, the process for determining abnormal scene areas includes:

[0037] The scene monitoring space is divided into several scene regions. Whenever the scene monitoring space is updated, the corresponding real-time environmental data is retrieved and compared according to the types of dynamic limitation ranges of each scene region. Anomaly labels are set for the scene regions based on the comparison results.

[0038] Furthermore, the process of setting up anomaly propagation arc zones based on abnormal scene areas includes:

[0039] Based on the types of real-time environmental data recorded by the anomaly annotation, the corresponding real-time environmental data is recorded as abnormal real-time environmental data. The fluctuation direction of the corresponding real-time environmental data in the scene area of ​​the adjacent spatial location is determined, and the data fluctuation threshold is set.

[0040] When it is determined that the fluctuation direction of the real-time environmental data of the scene area with adjacent spatial locations is developing towards the value of abnormal real-time environmental data, and the value fluctuation is greater than or equal to the data fluctuation threshold, the scene area with the adjacent spatial location is set with the same abnormal label.

[0041] The scene areas marked with anomalies are connected sequentially to construct the anomaly propagation arc area. Three non-collinear verification points are randomly selected on the edge line of the anomaly propagation arc area, and adjacent verification points are connected to obtain the arc lines X and Y.

[0042] Using the two verification points of arc line X as centers, two reference arc lines are generated with a radius greater than half the length of arc line X. The two reference arc lines are extended to intersect, and the intersection point is connected to obtain the perpendicular bisector of arc line X. The perpendicular bisector of arc line Y is obtained in the same way. The perpendicular bisectors of arc lines X and Y are extended to intersect, and the intersection point is extended and connected to the center position of the edge line of the abnormal spread arc area. The connecting line is recorded as the abnormal spread direction, and then an idle UAV is called to continuously track along the abnormal spread direction.

[0043] An IoT-based industrial park environmental monitoring system, used to implement the IoT-based industrial park environmental monitoring method according to any one of claims 1-8, characterized in that it includes a monitoring scene construction module, a scene data acquisition module, and an anomaly detection module.

[0044] The monitoring scenario construction module is used to establish a scenario monitoring space based on historical production records, update the scenario monitoring space based on real-time environmental data from the scenario data acquisition module, and generate an inspection loop route for the drone in the scenario monitoring space.

[0045] The scene data acquisition module is equipped with a data handshake mechanism, which is used to perform spatiotemporal alignment and data fusion of real-time environmental data collected by ground sensing devices and drones.

[0046] The anomaly detection module is used to construct a dynamically limited range based on historical production records and to divide the scene monitoring space into several scene areas. Then, it uses the dynamically limited range to determine whether there are anomalies in the real-time environmental data of each scene area. Based on the determination result, it constructs an anomaly propagation arc area and evaluates the direction of anomaly propagation based on the anomaly propagation arc area until the scene area with anomalies disappears.

[0047] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0048] 1. This invention combines the advantages of ground-based sensing devices and drones in the target industrial park to achieve comprehensive monitoring of the park environment, making up for the limited monitoring range of ground-based monitoring equipment. At the same time, it sets a dynamic range limit based on the historical production records of the target industrial park and updates the scene monitoring space. Since the production activities in the industrial park are dynamic, the production situation may be different at different times. The dynamic range limit can better adapt to this change, making the monitoring more accurately reflect the actual environmental conditions of the park and improving the effectiveness and adaptability of the monitoring.

[0049] 2. This invention divides the scene monitoring space into several scene areas, locates the scene areas with anomalies according to the dynamically defined range, establishes an anomaly propagation arc zone, and assesses the direction of anomaly propagation. To a certain extent, it achieves rapid and accurate location of environmental anomaly areas and predicts the spread trend of anomalies, effectively preventing the further expansion of anomalies and reducing the impact on the park and surrounding environment. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0051] Figure 1 This is a flowchart of the IoT-based industrial park environmental monitoring method described in this invention.

[0052] Figure 2 This is a system framework diagram of the IoT-based industrial park environmental monitoring system described in this invention. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0054] Please see Figure 1 As shown, the IoT-based environmental monitoring method for industrial parks includes the following steps:

[0055] Step S1: Set up ground sensing devices and drones in the target industrial park to acquire and establish a scene monitoring space based on the historical production records of the target industrial park;

[0056] Step S2: Set a dynamic limit range based on the historical production records of the target industrial park, and set an inspection loop route in the scene monitoring space. Then, the drone collects real-time environmental data of the target industrial park through the inspection loop route, and performs data handshake with the ground sensing device. Based on the data handshake result, complete data fusion and update the scene monitoring space.

[0057] Step S3: Divide the scene monitoring space into several scene areas, locate the scene areas with anomalies in the scene monitoring space according to the dynamically limited range, establish anomaly propagation arc area based on the scene areas with anomalies and their adjacent scene areas, and then evaluate the direction of anomaly propagation based on the anomaly propagation arc area until the scene areas with anomalies disappear.

[0058] Furthermore, step S1 is implemented through the following process:

[0059] Step S101: Set up the ground sensing device and the drone. The specific process includes:

[0060] Set up n ground sensing devices and m drones in the target industrial park. The ground sensing devices are evenly distributed in the target industrial park at fixed spatial intervals, and n and m are natural numbers greater than 20.

[0061] The ground sensing device integrates a temperature and humidity sensor, a multi-parameter air quality sensor (used to collect the concentration of fine air particles such as nitrogen oxides and volatile organic compounds), a wind speed and direction sensor, a water quality sensor (used to collect the pH value, ammonia nitrogen, total phosphorus and other substances content in the drainage from the sewage outlet, circulating water pool, rainwater pipe network and other locations in the target industrial park), a soil sensor (used to collect parameters such as soil moisture content, heavy metal content and other parameters from the raw material storage area, solid waste landfill area and soil around the production workshop in the target industrial park), and a laser imager.

[0062] It should be noted that the types of sensors that the ground sensing device operates in real time vary depending on the installation location of the ground sensing device. For example, a ground sensing device located at a sewage discharge outlet operates a water quality sensor, a temperature and humidity sensor, and a laser imager (used to acquire laser imaging video) in real time.

[0063] The drone is equipped with a temperature and humidity sensor, a multi-parameter air quality sensor, a wind speed and direction sensor, and a laser imager (for acquiring laser imaging video).

[0064] Each ground sensing device and UAV is assigned the numbers a1, a2, ..., a n b1, b2, ..., b m ;

[0065] At the same time, a fixed data acquisition area is set for each ground sensing device and drone, and the data acquisition areas of ground sensing devices in adjacent spatial locations partially overlap.

[0066] Step S102: Establish a scenario monitoring space based on the historical production records of the target industrial park. The specific process includes:

[0067] The target industrial park consists of several production areas, and the products produced in each production area are not exactly the same, such as producing waste materials, laundry detergent, and auto parts.

[0068] Obtain historical production records of the target industrial park, which include historical change records of various historical environmental data such as waste gas and wastewater emitted during the production process in each production area of ​​the target industrial park;

[0069] A scene monitoring space is established based on the scene distribution of the target industrial park. At the same time, multiple scene monitoring areas are marked in the scene monitoring space according to the production area distribution. In addition, a pipeline visualization image is generated in the scene monitoring space according to the pipeline distribution of the target industrial park.

[0070] In the scene monitoring space, n ground sensing nodes and m aerial sensing nodes are set up, and data channels are established between each ground sensing node and each aerial sensing node and each ground sensing device and UAV in sequence. Each ground sensing node and aerial sensing node marks the scene area associated with the real-time data collection area in the scene monitoring space, wherein the data collection area is circular.

[0071] Furthermore, step S2 is implemented through the following process:

[0072] Step S201: Set the dynamic limit range. The specific process includes:

[0073] Obtain k historical change records for each production area and establish a two-dimensional coordinate system. Map the same historical environmental data in each historical change record to the same two-dimensional coordinate system in days, where k is a natural number of 100.

[0074] Divide a day into several time segments and set a data density frame. The data density frame is a rectangular data frame with a length equal to the duration of the time segment and a width of a preset value.

[0075] Starting from the first data segment in each time interval along the y-axis, the data segments in the time interval are selected and traversed by using a data density box, and the number of data segments selected is counted each time a data segment is selected.

[0076] Select the data segment with the largest number of selected data segments to form the dynamic limit range under the corresponding time segment. Then, stitch the dynamic limit ranges under each time segment together in sequence and mark the stitching result in the corresponding scene monitoring area of ​​the scene monitoring space.

[0077] Step S202: Set the inspection loop route. The specific process includes:

[0078] Taking the scene monitoring area in the scene monitoring space as the inspection target, select any two spatial boundaries of the scene monitoring space to set m inspection starting points and inspection transfer points, take the diameter of the data collection area of ​​the UAV as the width of the inspection loop route, and then take the overlap of the data collection area radius length of any two inspection loop routes as the first condition, and take the inspection target as the priority target and avoidance of obstacles in the target industrial park as the second condition, and set m inspection loop routes.

[0079] All inspection loop routes are retrieved and inspection flight simulations are performed in the scene monitoring space. It is determined whether the inspection area composed of each inspection loop route covers the entire scene monitoring space. If it is determined that the entire scene monitoring space is covered, the inspection loop routes are assigned to each drone for execution.

[0080] If it is determined that the entire scene monitoring space is not covered, the uncovered part is marked on the scene monitoring space. Then, on the premise of ensuring that there are no gaps between adjacent inspection loop routes, the nearest inspection loop route to the uncovered part is partially shifted so that the uncovered part is covered by the inspection loop route.

[0081] If the condition that the uncovered part is not covered by the inspection loop is not met when the most recent inspection loop route is moved to the next location, the adjacent inspection loop route of the most recent inspection loop route will be retrieved again for translation. The above operation of retrieving inspection loop routes for translation will be repeated until the uncovered part is covered by the inspection loop route, or until there is no inspection loop route to choose from for translation.

[0082] Step S203: Collect real-time environmental data and perform a data handshake. The specific process includes:

[0083] After synchronizing the system clocks of all ground sensing devices and UAVs, the same data update cycle is set for all ground sensing nodes and airborne sensing nodes, with the data update cycle typically lasting 5 to 10 seconds.

[0084] Then, each drone collects real-time environmental data within the collection area by various sensors along the inspection loop route and ground sensing devices.

[0085] As the drone flies along the inspection loop, at the beginning of each data update cycle, if it is determined that the data sensing area of ​​a drone overlaps with the data sensing area of ​​any ground sensing device, the data handshake mechanism is automatically triggered.

[0086] The data handshake mechanism includes: after each data update cycle, the scene monitoring space dynamically updates the spatial position of each ground sensing node and the air sensing node in the scene monitoring space based on the real-time location information sent by each ground sensing device and the drone.

[0087] Then, based on the associated scene areas marked by the ground sensing nodes and the air sensing nodes, it is determined in real time whether there is an overlap in the data collection area between the ground sensing nodes and the air sensing nodes. If it is determined that there is no overlap, no operation is performed.

[0088] When it is determined that there is overlap, real-time environmental data of the same type collected by the corresponding ground sensing node and air sensing node are retrieved based on the overlapping spatial portion.

[0089] Since the overlapping areas of the ground sensing devices and UAV data sensing areas corresponding to the ground sensing nodes and the air sensing nodes correspond to the same scene target, the laser imaging videos acquired by the ground sensing nodes and the air sensing nodes must have the same parts.

[0090] The laser imaging videos of ground sensing nodes and air sensing nodes are matched in time sequence. Based on the content matching results, the common parts of the two are found. Then, the laser imaging videos of the two are stitched together based on the common parts. A three-dimensional image model is generated based on the stitched laser imaging video and overlaid on the scene monitoring space. Other types of environmental data are first spatiotemporally aligned according to the docking relationship of the laser imaging videos, and then overlaid on the three-dimensional image model in the form of texture.

[0091] The scene areas associated with ground sensing nodes and air sensing nodes are divided into several data display areas of the same size. Real-time environmental data of the same type acquired by ground sensing nodes and air sensing nodes are mapped to the corresponding data display areas according to their spatiotemporal location.

[0092] Both ground sensing nodes and air sensing nodes are initially assigned a trust weight of 0.5. A deviation threshold is set for each type of real-time environmental data. At the same time, the relationship between the difference and the deviation threshold between the real-time environmental data of the same type in each data display area is determined.

[0093] If the difference is less than the deviation threshold, it is determined that there is no difference in the real-time environmental data of the corresponding type.

[0094] If the difference is greater than or equal to the deviation threshold, then the real-time environmental data collected by the ground sensing node and the air sensing node in the adjacent spatial location are retrieved respectively, and then it is determined again whether the difference between the real-time environmental data of the corresponding type of the ground sensing node and the ground sensing node in the adjacent spatial location is greater than or equal to the deviation threshold.

[0095] If either party judges the difference to be greater than or equal to the deviation threshold, and the other party judges the difference to be less than the deviation threshold, then the trust weight of the perception node whose judgment difference is greater than or equal to the deviation threshold will be reduced by 10%, and the trust weight of the other party's perception node will be increased by the amount reduced by the other party.

[0096] If both parties agree, no trust weight adjustment will be performed.

[0097] After all data display areas have been judged, the real-time environmental data of the same type in the data display areas are merged according to the trust weight of both parties. The fusion formula is: A=αa+βb, where α and β represent the trust weight of both parties, A is the value of the real-time environmental data after fusion, and a and b are the values ​​of the real-time environmental data before fusion.

[0098] Then, when the data update cycle ends, the scene monitoring space is updated based on the real-time environmental data collected by each ground sensing node and air sensing node, as well as the data handshake results.

[0099] Furthermore, step S3 is implemented through the following process:

[0100] Step S301: Determine the scene area where anomalies exist. The specific process includes:

[0101] The scene monitoring space is divided into several scene regions. Whenever the scene monitoring space is updated, the corresponding real-time environmental data is retrieved and compared according to the type of dynamic limit range of each scene region. If it is determined that the real-time environmental data of the scene region is within its corresponding dynamic limit range, the scene region is judged to be normal.

[0102] If the real-time environmental data of the scene area is determined to be outside its corresponding dynamic limit range, then the scene area is determined to be abnormal, and an abnormal label is set for the scene area according to the type of real-time environmental data that is not outside the dynamic limit range.

[0103] Step S302: Set the abnormal propagation arc zone according to the abnormal scene area. The specific process includes:

[0104] Based on the types of real-time environmental data recorded by the anomaly annotation, the corresponding real-time environmental data is recorded as abnormal real-time environmental data. The fluctuation direction of the corresponding real-time environmental data in the scene area of ​​the adjacent spatial location is determined, and the data fluctuation threshold is set.

[0105] When it is determined that the fluctuation direction of the real-time environmental data of the scene area with adjacent spatial locations is developing towards the value of abnormal real-time environmental data, and the value fluctuation is greater than or equal to the data fluctuation threshold, the scene area with the adjacent spatial location is set with the same abnormal label.

[0106] The scene areas marked with anomalies are connected sequentially to construct the anomaly propagation arc area. Three non-collinear verification points are randomly selected on the edge line of the anomaly propagation arc area, and adjacent verification points are connected to obtain the arc lines X and Y.

[0107] Using the two verification points of arc line X as centers, generate two reference arc lines with a radius greater than half the length of arc line X. Extend the two reference arc lines to intersect, and connect the intersection point to obtain the perpendicular bisector of arc line X. Obtain the perpendicular bisector of arc line Y in the same process. Extend the perpendicular bisectors of arc lines X and Y to intersect, and extend the intersection point to connect it with the center position of the edge line of the abnormal spread arc area. Record the connecting line as the abnormal spread direction, and then call up an idle drone to continuously track along the abnormal spread direction until the abnormal labeling of the scene area with abnormal labeling disappears.

[0108] It should be noted that the abnormal propagation arc area is updated after each data update cycle, and the real-time propagation direction is reselected.

[0109] Please see Figure 2 As shown, the IoT-based industrial park environmental monitoring system includes a monitoring scenario construction module, a scenario data acquisition module, and an anomaly detection module.

[0110] The monitoring scenario construction module is used to establish a scenario monitoring space based on historical production records, update the scenario monitoring space based on real-time environmental data from the scenario data acquisition module, and generate an inspection loop route for the drone in the scenario monitoring space.

[0111] The scene data acquisition module is equipped with a data handshake mechanism, which is used to perform spatiotemporal alignment and data fusion of real-time environmental data collected by ground sensing devices and drones.

[0112] The anomaly detection module is used to construct a dynamically limited range based on historical production records and to divide the scene monitoring space into several scene areas. Then, it uses the dynamically limited range to determine whether there are anomalies in the real-time environmental data of each scene area. Based on the determination result, it constructs an anomaly propagation arc area and evaluates the direction of anomaly propagation based on the anomaly propagation arc area until the scene area with anomalies disappears.

[0113] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An industrial park environmental monitoring method based on the Internet of Things, characterized in that, Includes the following steps: Step S1: Set up ground sensing devices and drones in the target industrial park to acquire and establish a scene monitoring space based on the historical production records of the target industrial park; Step S2: Set a dynamic limit range based on the historical production records of the target industrial park, and set an inspection loop route in the scene monitoring space. Then, the drone collects real-time environmental data of the target industrial park through the inspection loop route, and performs data handshake with the ground sensing device. Based on the data handshake result, complete data fusion and update the scene monitoring space. Step S3: Divide the scene monitoring space into several scene areas, locate the scene areas with anomalies in the scene monitoring space according to the dynamically limited range, establish an anomaly propagation arc area based on the scene areas with anomalies and their adjacent scene areas, and then evaluate the direction of anomaly propagation based on the anomaly propagation arc area until the scene areas with anomalies disappear. The process of setting up an anomaly propagation arc zone based on the anomaly scene area includes: Based on the types of real-time environmental data recorded by the anomaly annotation, the corresponding real-time environmental data is recorded as abnormal real-time environmental data. The fluctuation direction of the corresponding real-time environmental data in the scene area of ​​the adjacent spatial location is determined, and the data fluctuation threshold is set. When it is determined that the fluctuation direction of the real-time environmental data of the scene area with adjacent spatial locations is developing towards the value of abnormal real-time environmental data, and the value fluctuation is greater than or equal to the data fluctuation threshold, the scene area with the adjacent spatial location is set with the same abnormal label. The scene areas marked with anomalies are connected sequentially to construct the anomaly propagation arc area. Three non-collinear verification points are randomly selected on the edge line of the anomaly propagation arc area, and adjacent verification points are connected to obtain the arc lines X and Y. Using the two verification points of arc line X as centers, two reference arc lines are generated with a radius greater than half the length of arc line X. The two reference arc lines are extended to intersect, and the intersection point is connected to obtain the perpendicular bisector of arc line X. The perpendicular bisector of arc line Y is obtained in the same way. The perpendicular bisectors of arc lines X and Y are extended to intersect, and the intersection point is extended and connected to the center position of the edge line of the abnormal spread arc area. The connecting line is recorded as the abnormal spread direction, and then an idle UAV is called to continuously track along the abnormal spread direction.

2. The industrial park environmental monitoring method based on the Internet of Things according to claim 1, characterized in that, A fixed data acquisition area is set for each ground sensing device and the UAV, and the data acquisition areas of ground sensing devices in adjacent spatial locations partially overlap.

3. The method for environmental monitoring of industrial parks based on the Internet of Things according to claim 2, characterized in that, The process of establishing a scene monitoring space based on historical production records includes: The target industrial park consists of several production areas. The historical production records of the target industrial park are obtained. The historical production records include historical change records of various historical environmental data in the production process of each production area within the target industrial park. Establish a scene monitoring space based on the scene distribution of the target industrial park, and mark multiple scene monitoring areas in the scene monitoring space according to the distribution of production areas. In the scene monitoring space, n ground sensing nodes and m aerial sensing nodes are set up, and data channels are established between each ground sensing node and each aerial sensing node and each ground sensing device and UAV in sequence. Each ground sensing node and each aerial sensing node marks the scene area associated with the real-time data collection area in the scene monitoring space, where n and m are natural numbers greater than 20.

4. The industrial park environmental monitoring method based on the Internet of Things according to claim 3, characterized in that, The process of setting dynamically limited ranges includes: Divide a day into several time segments and set a data density frame. The data density frame is a rectangular data frame with a length equal to the duration of the time segment and a width of a preset value. Starting from the first data segment in each time interval along the y-axis, the data segments in the time interval are selected and traversed by using a data density box, and the number of data segments selected is counted each time a data segment is selected. Select the data segment with the largest number of selected data segments to form the dynamic limit range under the corresponding time segment. Then, stitch the dynamic limit ranges under each time segment together in sequence and mark the stitching result in the corresponding scene monitoring area of ​​the scene monitoring space.

5. The industrial park environmental monitoring method based on the Internet of Things according to claim 4, characterized in that, The process of setting up the inspection loop route includes: Taking the scene monitoring area in the scene monitoring space as the inspection target, select any two spatial boundaries of the scene monitoring space to set m inspection starting points, take the diameter of the data collection area of ​​the UAV as the width of the inspection loop route, and then take the overlap of the data collection area radius length of any two inspection loop routes as the first condition, and take the inspection target as the priority target and avoid the obstacles in the target industrial park as the second condition, and set m inspection loop routes. All inspection loop routes are retrieved and inspection flight simulations are performed in the scene monitoring space. It is determined whether the inspection area composed of each inspection loop route covers the entire scene monitoring space. If it is determined that the entire scene monitoring space is covered, the inspection loop routes are assigned to each drone for execution. If it is determined that the entire scene monitoring space is not covered, the uncovered part is marked on the scene monitoring space. Then, on the premise of ensuring that there are no gaps between adjacent inspection loop routes, the nearest inspection loop route to the uncovered part is partially shifted so that the uncovered part is covered by the inspection loop route. If the condition that the uncovered part is not covered by the inspection cycle route is not met when the most recent inspection cycle route is moved to the next location, the adjacent inspection cycle route of the most recent inspection cycle route will be retrieved and moved again until the uncovered part is covered by the inspection cycle route, or until there is no inspection cycle route to choose from for moving.

6. The industrial park environmental monitoring method based on the Internet of Things according to claim 5, characterized in that, The data handshake process includes: Set the same data update cycle for each ground sensing node and air sensing node, and determine in real time whether there is overlap in the data collection area of ​​the ground sensing node and air sensing node based on the associated scene area marked by the ground sensing node and air sensing node. Based on the judgment results, the laser imaging videos of ground sensing nodes and air sensing nodes are matched in time sequence. Based on the content matching results, other types of environmental data are first spatiotemporally aligned according to the docking relationship of the laser imaging videos. The scene areas associated with ground sensing nodes and air sensing nodes are divided into several data display areas of the same size. Real-time environmental data of the same type acquired by ground sensing nodes and air sensing nodes are mapped to the corresponding data display areas according to their spatiotemporal location. Trust weights and deviation thresholds are set for ground sensing nodes and air sensing nodes. The relationship between the difference and the deviation threshold between real-time environmental data of the same type in each data display area is judged. Based on the judgment result, the real-time environmental data collected by ground sensing nodes and air sensing nodes in adjacent spatial locations are retrieved respectively. Then, it is judged again whether the difference between the real-time environmental data of the corresponding type of ground sensing node and the ground sensing node in the adjacent spatial location is greater than or equal to the deviation threshold. Based on the judgment result, the trust weight adjustment operation is performed, and the real-time environmental data of the same type in the data display area are merged according to the trust weight of both parties.

7. The industrial park environmental monitoring method based on the Internet of Things according to claim 6, characterized in that, The anomaly detection process for a scene area includes: The scene monitoring space is divided into several scene regions. Whenever the scene monitoring space is updated, the corresponding real-time environmental data is retrieved and compared according to the types of dynamic limitation ranges of each scene region. Anomaly labels are set for the scene regions based on the comparison results.

8. An IoT-based industrial park environmental monitoring system, used to implement the IoT-based industrial park environmental monitoring method according to any one of claims 1-7, characterized in that, It includes a monitoring scenario construction module, a scenario data acquisition module, and an anomaly detection module; The monitoring scenario construction module is used to establish a scenario monitoring space based on historical production records, update the scenario monitoring space based on real-time environmental data from the scenario data acquisition module, and generate an inspection loop route for the drone in the scenario monitoring space. The scene data acquisition module is equipped with a data handshake mechanism, which is used to perform spatiotemporal alignment and data fusion of real-time environmental data collected by ground sensing devices and drones. The anomaly detection module is used to construct a dynamically limited range based on historical production records and to divide the scene monitoring space into several scene areas. Then, it uses the dynamically limited range to determine whether there are anomalies in the real-time environmental data of each scene area. Based on the determination result, it constructs an anomaly propagation arc area and evaluates the direction of anomaly propagation based on the anomaly propagation arc area until the scene area with anomalies disappears.

Images