Remote factory air pollution emission monitoring method, terminal and system based on space-based Internet of Things

Abstract

The invention provides a far-field factory air pollution emission monitoring method, terminal and system based on space-based Internet of Things. Wherein the monitoring system is used for managing a terminal and training a model, the monitoring system is trained with a PM detection model, and the PM detection model is used for identifying smoke and non-smoke. After generating an instruction for controlling the terminal to work, the system sends instruction data and model data to a ground station and then sends the instruction data and the model data to the terminal through a satellite. The monitoring system is also used for receiving terminal data returned by the satellite; the monitoring terminal comprises a terminal processor, a time service module, a satellite module, a battery module, an infrared sensor, a camera module and a storage module. According to the method, video shooting is carried out on the terminal side, data of more information dimensions are obtained, and air pollution is analyzed. And the accuracy of air pollution analysis is improved through flexible deployment. And data transmission is carried out through satellites without depending on a ground network, so that the defects of the existing system are effectively avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental monitoring, and specifically to a method, a terminal, and a system for monitoring air pollution emissions from remote factories based on a space-based Internet of Things. Background Technique

[0002] Factory emissions are one of the sources of air pollution. According to relevant information, air pollution enterprises generally choose their locations far from densely populated areas, such as remote mountainous areas or rural areas. The complex environment in remote areas poses challenges to air pollution monitoring. In some remote areas, the environment is harsh and not suitable for humans to live and guard for a long time, and humans cannot directly carry out monitoring work. In some remote areas, the infrastructure is imperfect, and it is difficult and costly to build communication facilities. Existing monitoring equipment is particularly dependent on professional equipment and needs to deploy the equipment within the factory park or in areas covered by the ground network, etc. The communication method depends on the ground network. There is a lack of air pollution monitoring in remote areas without ground network coverage or with weak ground network coverage.

[0003] Most of the monitoring equipment in the prior art is composed of various sensors, and the air quality is judged by analyzing the content of various substances in the air. This method has certain limitations and cannot reflect spatial information. The video method can provide more information dimensions. The video can capture the full view image of the factory, and the information can include emission equipment information, time information, weather information, season information, etc., which is more conducive to the analysis of air pollution.

[0004] However, the existing monitoring method depends on the ground network and needs to adopt a camera, a switch, and a server architecture. The data is transmitted to the backend server for analysis through communication methods such as local area network, optical fiber, or 4G. Such an architecture involves more equipment, increases the risk of damage, and improves the maintenance cost. In areas with poor network transmission or when the ground network cannot be used due to natural disasters, the monitoring personnel must go to the scene to manually copy the data, which is time-consuming and laborious, and is accompanied by personal safety risks. At the same time, the delayed transmission of information may lead to potential production safety hazards. The present invention relates to a terminal that does not depend on the ground network and saves maintenance and transmission costs. Summary of the Invention

[0005] Aiming at the deficiencies of the prior art, the purpose of the present invention is to provide a method, a terminal, and a system for monitoring air pollution emissions from remote factories based on a space-based Internet of Things to solve the problems raised in the above background technique. The present invention captures data with more information dimensions through video shooting on the terminal side to analyze air pollution. The accuracy of air pollution analysis is increased through flexible deployment. The data is transmitted through satellites without relying on the ground network, effectively avoiding the defects of the existing system.

[0006] To achieve the above object, the present invention is implemented through the following technical solutions: A remote factory air pollution emission monitoring system based on a space-based Internet of Things. This monitoring system is used to manage and train models for terminals. The monitoring system is trained with a PM detection model, and the PM detection model is used to identify smoke and non-smoke. After the system generates an instruction to control the operation of the terminal, it sends the instruction data and model data to the ground station. When the satellite communicates with the ground station, it is then sent to the terminal through the satellite. The monitoring system is also used to receive the terminal data transmitted back by the satellite.

[0007] Further, after the monitoring system converts the satellite data into data required by the large model, the monitoring system further analyzes the mass concentration of PM, and through a visual front-end interface, displays the smoke image data captured by the terminal, and uses charts to display various indicators of air pollution.

[0008] A remote factory air pollution emission monitoring terminal based on a space-based Internet of Things. This monitoring terminal includes a terminal processor, a timing module, a satellite module, a battery module, an infrared sensor, a camera module, and a storage module. After the monitoring terminal is deployed near the factory, the camera is aimed at the factory emission area for shooting. The monitoring terminal combines with an image recognition module to export the key frames of the collected video images into images. When smoke is recognized, a smoke image is generated, and the image is compressed and stored using a picture compression algorithm.

[0009] Further, the monitoring terminal analyzes the smoke image to obtain the analysis result of the PM mass concentration. When the satellite passes by the terminal, it uploads the processed analysis result through the satellite communication module. At the same time, the monitoring system plans the picture data upload window. The picture data is used for the system to train the smoke recognition model, which is beneficial to model iteration and optimization.

[0010] Further, the terminal processor is built-in with a smoke recognition module, a PM analysis module, and a sample library processing module. The terminal processor communicates with the infrared sensor through an RS-485 interface. After the infrared sensor detects a heat source, it sends a signal to start the camera to shoot video. The camera module communicates with the processor using a USB interface. The storage module communicates with the processor through an SDIO interface and is used to store the captured video.

[0011] A remote factory air pollution emission monitoring method based on a space-based Internet of Things, including a method for reducing the energy consumption of the terminal based on an infrared sensor: using the infrared sensor to detect temperature changes. When the factory emits smoke, there is a phenomenon of generating heat through combustion or machinery. When the infrared sensor recognizes the heat source, it starts the camera to shoot, and in other cases, the camera is in a sleep state.

[0012] Furthermore, it also includes a method for reducing transmission costs: compressing the image through the monitoring terminal to reduce the communication bandwidth occupancy, and then processing the redundant data of the image to reduce the number of samples and reduce the total time of transmitting the image.

[0013] Furthermore, when the system receives the image data, it gives feedback and notifies the terminal to destroy the transmitted image to save storage space.

[0014] Furthermore, it also includes a method for achieving continuous energy supply based on solar panels: charging and storing energy in terminal batteries through the constructed solar panels to support long-term use of the terminal in remote areas.

[0015] Furthermore, it also includes a communication method: using Xingyun satellite to provide a communication link, and the communication between the terminal and the satellite depends on the satellite communication module. The control center plans the communication time window between the satellite and the terminal and a series of instructions to control the terminal through the system. When the satellite passes over the terminal, if it is within the communication time window, the control software controls the communication module to establish a connection with the satellite and transmit the data to the satellite. After the connection is established, if there is a control instruction on the satellite, it will also be downlinked to the terminal at the same time to update the software in the terminal and update the schedule for the module in the terminal to establish a connection with the satellite. After the communication time window ends, the terminal turns off the communication module.

[0016] Beneficial effects of the present invention:

[0017] 1. The present invention can be used in remote areas without relying on the ground network, and can be deployed at any location without being restricted by the ground network; it supports flexible deployment, integrates intelligent recognition models, can obtain information in more dimensions, and provide richer air quality analysis results; it does not rely on the ground network, reduces dependence on professional equipment, and reduces ground network maintenance costs and equipment use costs.

[0018] 2. Compared with existing air pollution monitoring methods, the present invention can expand the field of view in the time dimension through the combination of multiple frames of images, which contains more information than a single image or traditional sensor. Video analysis technology can effectively eliminate environmental interference factors (such as lighting changes, shadows, etc.) through image processing and machine learning algorithms to ensure the accuracy and reliability of monitoring results.

[0019] 3. The present invention has excellent flexibility and wide versatility, and can be migrated and widely used in multiple image recognition fields such as geological disaster early warning, ecological environment monitoring and public safety prevention and control, specifically covering typical application scenarios such as intelligent early warning monitoring of landslide deformation, continuous tracking and analysis of river and lake ecological parameters, and intelligent diagnosis of urban infrastructure safety. In these applications, it can ensure the acquisition and transmission of remote data, greatly expanding its application scope and enhancing its practical value. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is the flowchart of a remote factory air pollution emission monitoring system based on the space-based Internet of Things according to the present invention;

[0021] Figure 2 This is the composition structure diagram of a remote factory air pollution emission monitoring terminal based on the space-based Internet of Things according to the present invention;

[0022] Figure 3 This is the identification flowchart of the monitoring terminal according to the present invention;

[0023] Figure 4 This is the communication flowchart of the monitoring method according to the present invention;

[0024] Figure 5 This is the iterative process diagram from the general model to the specific model according to the present invention;

[0025] Figure 6 This is the data parsing flowchart according to the present invention. Specific embodiments

[0026] To make the technical means, creative features, achieved purposes and functions of the present invention easy to understand, the present invention will be further described below in conjunction with specific embodiments.

[0027] Please refer to Figures 1 to 6 , the present invention provides the following technical solutions: a remote factory air pollution emission monitoring method, terminal and system based on the space-based Internet of Things, wherein the composition of the monitoring terminal is as Figure 2 shown, mainly including a terminal processor, a camera module, a power module, a solar module, a storage module, a timing module and corresponding antennas, a Xingyun satellite communication module and corresponding antennas. A smoke recognition module, a PM analysis module and a sample library processing module are built in the terminal processor.

[0028] The terminal processor communicates with the infrared sensor through the RS-485 interface. After the infrared sensor detects a heat source, it sends a signal to start the camera to shoot a video. The camera module communicates with the processor using the USB interface. The storage module communicates with the processor through the SDIO interface and is used to store the shot video, and the storage capacity can be expanded. The image sample obtained after shooting the video is processed by the sample library processing module and communicates with the processor through the USB interface. The smoke recognition module is a software that runs in the terminal processor and is responsible for recognizing the smoke in the video. Both the sample library processing module and the PM analysis module are software that runs in the processor. The timing module and the satellite module both communicate with the processor through the UART interface. The terminal processor module obtains the time provided by the timing module, which is used to support the use of time information in intelligent recognition. The power module supplies power to all the above modules.

[0029] The smoke recognition module uses the Support Vector Machine (SVM) method to analyze smoke frames and non-smoke frames, and then uses the K-means algorithm to cluster and segment the frame pixels in the smoke frame data to obtain the characteristic data of the smoke. The PM analysis module analyzes the smoke characteristic data and estimates the PM mass concentration of the captured smoke data through a pre-set linear regression model. The sample library processing module is responsible for operations such as data cleaning, vectorization, and normalization of the newly generated smoke data to facilitate intelligent analysis by other modules. The program codes of the smoke recognition module, the PM analysis module, and the sample library processing module can be pre-set before deployment or updated by downloading the program code via satellite later. The recognition workflow is as Figure 3 shown.

[0030] This embodiment also provides an intelligent terminal and a communication method between the terminal and the satellite. In this embodiment, Xingyun satellite is used to provide a communication link. The communication between the terminal and the satellite depends on the satellite communication module, and the communication method is as Figure 4 shown. There is a control software running inside the satellite communication module. At the control center, the system plans the communication time window between the satellite and the terminal and a series of instructions for controlling the terminal. When the satellite passes over the terminal, if it is within the communication time window, the control software controls the communication module to establish a connection with the satellite and transmits the data to the satellite. After the connection is established, if there are control instructions on the satellite, they will also be downloaded to the terminal simultaneously for updating the software in the terminal and the schedule for the module in the terminal to establish a connection with the satellite. After the communication time window ends, the terminal closes the communication module to save power.

[0031] This embodiment also provides a smoke recognition model, which is pre-trained in the system background. First, various pictures of factory-emitted smoke are collected on the Internet, and the smoke and non-smoke are labeled. The SVM algorithm is used to perform image segmentation on the pictures to extract the texture information and spatio-temporal energy characteristics of the smoke. According to the color and motion characteristics of different PM smokes, the RGB color space is converted into the CIELab color space, and then the background subtraction and K-means clustering algorithms are used to segment the smoke pixels, effectively clustering the flare soot and chimney soot. The trained model is pre-set in the terminal processor for the terminal to use in actual scenarios. Since the types of each factory are different, the emitted smoke and the substances contained therein are different, and the training models will also be different. After the image data collected by the terminal in the actual scenario is returned to the background system via satellite, it will be used as new sample data to iteratively optimize the model, making the model more capable of identifying specific factories. The process is as Figure 5 shown.

[0032] This embodiment also provides the processing of satellite data. The back-end system receives the recognition result data pushed by the gateway station, and needs to verify, parse, and store the data to provide data support for the statistical analysis of the factory's atmospheric pollution emissions. The received data format is hexadecimal data, and the components are "0X + protocol header (1 byte) + timestamp (4-byte UNIX timestamp) + data length + pollutant data + key seed hash (32-byte SHA-256) + check bit (4-byte CRC)".

[0033] The back-end system receives the pollution emission recognition result data of the gateway station, and the processes of data verification, parsing, and storage are as Figure 6 shown. If the verification passes, the data bits are parsed; if the CRC verification fails, the data processing process ends. Based on the data that passes the verification, the data is split according to the data protocol bits, and the protocol header, timestamp, and pollutant data are parsed. Among them, the pollutant data is compressed by the terminal, and the system decompresses and stores the decompressed pollutant data. At the same time, after the back-end system receives the data pushed by the gateway station, it stores a copy of the source data, generates a data reception success signal, and proceeds with the next task planning.

[0034] The format of the atmospheric pollutant data is composed of the pollutant name and value, such as PM2.5 200. The data compression uses a compression algorithm that combines RLE (Run-Length Encoding) and differential encoding. First, we group the data according to the pollutant type, such as PM2.5, PM10, etc. Then, for the data within each group, we use RLE for preliminary compression, representing consecutive identical PM values with a value and the number of repetitions. To further optimize the compression effect, we introduce special markers to replace sequences with long consecutive repeated values, further reducing the data volume. In addition, for time series data, we use differential encoding and only transmit the change amount of the data, that is, the first data point uses the original value, and subsequent data points transmit the difference from the previous data point. Finally, all compressed data is represented and transmitted in hexadecimal format. The decompression process is the opposite of the compression process. First, we identify and parse the data groups of different pollutant types. Then, for the data within each group, we decompress it according to the compression rules of RLE to restore the original continuous data sequence. For the continuous repeated value sequences replaced by special markers, we restore the corresponding data according to the indication of the markers. Next, we process the differential encoding part of the time series data and reconstruct the original continuous data values by accumulating the differences. Finally, we obtain the decompressed atmospheric pollutant data composed of the pollutant name and value. The whole process ensures that the decompressed data is exactly the same as the original data.

[0035] The foregoing has shown and described the basic principles, main features and advantages of the present invention. For a person skilled in the art, it is obvious that the present invention is not limited to the details of the above-mentioned exemplary embodiments, and without departing from the spirit or basic features of the present invention, the present invention can be implemented in other specific forms.

[0036] In addition, it should be understood that although this specification is described in terms of embodiments, not every embodiment only contains an independent technical solution. This narrative manner of the specification is only for clarity. Those skilled in the art should regard the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A monitoring system for air pollution emissions from a remote factory based on a space-based Internet of Things, characterized in that: The monitoring system is used to manage the terminal and train the model. The monitoring system is trained with a PM detection model, which is used to identify smoke and non-smoke. After the system generates an instruction to control the operation of the terminal, it sends the instruction data and model data to the ground station, and then sends them to the terminal via satellite. The monitoring system is also used to receive the terminal data transmitted back by the satellite.

2. The air pollution emission monitoring system for a remote factory based on the space-based Internet of Things according to claim 1, wherein: After the monitoring system converts the satellite data into the data required by the large model, it further analyzes the mass concentration of PM, and displays the smoke image data captured by the terminal through a visual front-end interface, and uses charts to show various indicators of air pollution.

3. A monitoring terminal for air pollution emissions in a remote factory based on a space-based Internet of Things, characterized in that: The monitoring terminal includes a terminal processor, a timing module, a satellite module, a battery module, an infrared sensor, a camera module, and a storage module. After the terminal is deployed near the factory, the camera is aimed at the factory emission area for shooting. The monitoring terminal combines with an image recognition module to export the key frames of the collected video images into images. When smoke is recognized, a smoke image is generated, and the image is compressed and stored using a picture compression algorithm.

4. The air pollution emission monitoring terminal for a remote factory based on a space-based Internet of Things according to claim 3, characterized in that: The monitoring terminal analyzes the smoke image to obtain the analysis result of the PM mass concentration. When the satellite passes by the terminal, it uploads the processed analysis result through the satellite module. At the same time, the monitoring system plans the picture data upload window. The picture data is used for the system to train the smoke recognition model, which is beneficial to model iteration and optimization.

5. The air pollution emission monitoring terminal for a remote factory based on a space-based Internet of Things according to claim 4, wherein: The terminal processor is built-in with a smoke recognition module, a PM analysis module, and a sample library processing module. The terminal processor communicates with the infrared sensor through an RS-485 interface. After the infrared sensor detects a heat source, it sends a signal to start the camera to shoot video. The camera module communicates with the processor using a USB interface. The storage module communicates with the processor through an SDIO interface and is used to store the captured video.

6. A method for monitoring air pollution emissions from a remote factory based on a space-based Internet of Things, characterized in that, It includes a method for reducing the power consumption of the terminal based on the infrared sensor: using the infrared sensor to detect temperature changes. When the factory emits smoke, accompanied by heat generation through combustion or machinery, the infrared sensor starts the camera to shoot when it recognizes the heat source, and the camera is in a sleep state in other cases.

7. The air pollution emission monitoring method for a remote factory based on a space-based Internet of Things according to claim 6, wherein It also includes a method for reducing transmission costs: compressing the pictures through the monitoring terminal to reduce the occupancy of communication bandwidth. Secondly, redundant data processing is performed on the pictures to reduce the number of samples and the total duration of transmitting pictures.

8. The method for monitoring air pollution emissions from a remote factory based on a space-based Internet of Things according to claim 7, characterized in that: When the system receives the picture data, it gives feedback to notify the terminal to destroy the transmitted pictures, saving storage space.

9. The method for monitoring air pollution emissions from a remote factory based on a space-based Internet of Things according to claim 6, characterized in that, It also includes a method for realizing continuous power supply based on solar panels: charging and storing energy for the terminal battery through the built solar panels, which is used to support the long-term use of the terminal in remote areas.

10. A method for monitoring air pollution emissions from a remote factory based on a space-based Internet of Things according to claim 6, characterized in that, It also includes a communication method: using Xingyun satellites to provide communication links. The communication between the terminal and the satellite depends on the satellite module. At the control center, the communication time window between the satellite and the terminal and a series of instructions for controlling the terminal are planned by the system. When the satellite passes over the module, if it is within the communication time window, the control software controls the communication module to establish a connection with the satellite and transmit data to the satellite. After the connection is established, if there are control instructions on the satellite, they will also be downlinked to the terminal simultaneously, which are used to update the software in the terminal and update the schedule for the module in the terminal to establish a connection with the satellite. After the communication time window ends, the terminal closes the communication module.

Images