Low-cost dual-drive self-calibration dust concentration online monitoring system and monitoring method

Abstract

The invention discloses a low-cost dual-drive self-calibration dust concentration on-line monitoring system and a monitoring method. According to the system, a dust concentration monitoring assembly comprises an oscillating balance dust continuous tester and a plurality of low-cost dust monitors; the coal mine information processing center is respectively connected with the oscillating balance dust continuous tester and the low-cost dust monitor; the visual client is connected with the central processor through a third communication module. The method comprises the following steps: carrying out on-line monitoring operation by utilizing multiple groups of dust concentration monitoring assemblies; performing self-calibration on dust concentration data of the low-cost dust monitor by using a physically-driven prediction model; performing secondary real-time calibration on the dust concentration data of the low-cost dust monitor by using a data-driven machine learning calibration model; and generating a mine whole area dust distribution cloud picture by using the mine whole area dust distribution cloud picture unit, and sending the mine whole area dust distribution cloud picture to the visual client for real-time display. According to the system and the method, high-precision monitoring of the dust concentration can be realized on the basis of greatly reducing the monitoring cost.

Description

Technical Field

[0001] The invention belongs to the technical field of dust monitoring, and in particular relates to a low-cost dual-drive self-calibration dust concentration online monitoring system and a monitoring method. Background Art

[0002] Monitoring of underground dust concentration is extremely important for safe production in coal mines. Dust refers to fine solid particles suspended and dispersed in the gas. If the dust concentration is too high, workers may suffer from occupational diseases such as pneumoconiosis, which will pose a serious threat to their health. In addition, dust may also cause explosions under certain conditions, resulting in significant casualties and economic losses. In order to effectively prevent such accidents and ensure safe production operations in coal mines, it is necessary to monitor dust concentration in real time so that effective countermeasures can be taken in time when danger occurs.

[0003] At present, most dust monitoring equipment used in industrial and mining enterprises is based on the principle of light scattering, although light scattering dust monitoring equipment is widely used in industrial and mining enterprises. However, its monitoring results are easily affected by environmental factors such as humidity, temperature and air pressure, and the traceability of the measurement value is poor. In addition, during long-term continuous monitoring, it is easy to produce problems such as optical window contamination and instrument drift, which can easily cause a decrease in measurement accuracy. In addition, since there is a need for dust reduction in some working environments, there are often a large number of atomized droplets around the spray dust reduction site or in the wind flow. The atomized droplets will seriously affect the test results, and thus the accuracy of the test results cannot be guaranteed. For this reason, the detection equipment needs to be calibrated regularly, which increases the frequency of maintenance and reduces the detection efficiency. In addition, the oscillating balance dust monitoring method (TEOM) is a direct-reading monitoring method that does not require pretreatment and can directly monitor dust in the air. This method can achieve rapid and continuous monitoring, has the characteristics of high temporal and spatial resolution, and can provide more timely and accurate monitoring data. However, its cost is relatively high. In actual use, in order to ensure monitoring accuracy, it is often necessary to deploy multiple oscillating balance dust continuous measurement devices in the same monitoring area, which greatly increases the cost of dust monitoring and brings a huge burden to the long-term real-time monitoring of dust in a large number of monitoring areas underground. In summary, traditional dust monitoring methods have many limitations, such as low detection accuracy, expensive equipment costs, and high maintenance frequency. The existence of these problems limits the popularization and application of dust monitoring technology.

[0004] With the advancement of science and technology and the increase in industrial demand, higher requirements have been put on dust detection technology, including real-time detection, accuracy, stability and cost-effectiveness. In order to meet the requirements of dust monitoring technology at this stage, and to effectively improve the monitoring accuracy of dust in different environments, and to significantly reduce the monitoring cost and effectively reduce the maintenance frequency to effectively ensure the monitoring efficiency, it is urgent to provide a low-cost, high-efficiency, high-precision and easy-to-maintain dust concentration monitoring system and monitoring method. Summary of the invention

[0005] In view of the problems existing in the above-mentioned prior art, the present invention provides a low-cost dual-drive self-calibration dust concentration online monitoring system and a monitoring method. The system has low investment cost and high monitoring accuracy, and can realize high-precision monitoring of dust concentration in the monitored place. The method uses physical-driven and data-driven models to perform multiple self-calibrations on low-cost dust monitors. A large number of low-cost dust monitors can be used to replace a large number of high-cost oscillating balance dust continuous measuring instruments, which can significantly reduce the economic cost while improving the accuracy of dust monitoring.

[0006] In order to achieve the above-mentioned object, the present invention provides a low-cost dual-drive self-calibration dust concentration online monitoring system, including multiple groups of dust concentration monitoring components, a coal mine information processing center and a visual client; The dust concentration monitoring assembly includes an oscillating balance dust continuous measuring instrument and a plurality of low-cost dust monitors; The low-cost dust monitor includes a housing, a light scattering chamber, a laser, a lens group, a light trap, a laser detection matrix, a filter component, a temperature and humidity sensor, a micro fan, a positioning module 1, a power module, a communication module 1 and a data processing module 1; The housing forms a receiving chamber inside, and a main air inlet and a main air outlet communicating with the receiving chamber are provided at the front center and the rear center opposite to each other; The light scattering chamber is located in the accommodating chamber, and a measuring chamber is formed inside the chamber. A measuring inlet and a measuring outlet are respectively provided at the front and rear ends of the chamber at positions corresponding to the central area of ​​the measuring chamber, and the measuring inlet is connected to the main air inlet through an air inlet pipeline, and the measuring outlet is connected to the main air outlet through an air outlet pipeline; The laser is installed at the center of the bottom side wall of the housing, and its output end is located inside the measuring chamber. At the same time, the laser beam emitted by the laser passes through the central area of ​​the measuring chamber and is perpendicular to the axis of the measuring chamber. The lens group is fixedly installed between the central area of ​​the measurement chamber and the output end of the laser, and is used to focus the laser beam output by the laser on the central area of ​​the measurement chamber; The optical trap is located in the measurement chamber and is mounted on the top side wall of the housing opposite to the laser; The laser detection matrix is ​​located in the measurement chamber and is installed between the central area of ​​the measurement chamber and the light trap; the laser detection matrix is ​​composed of three groups of silicon photodiodes, wherein the first group of silicon photodiodes is distributed relative to the laser, and its detection surface is located on the same straight line as the output direction of the laser, and the second group of silicon photodiodes and the third group of silicon photodiodes are symmetrically distributed on both sides of the first group of silicon photodiodes, and their detection surfaces are all facing the central area of ​​the measurement chamber; the three silicon photodiodes in the laser detection matrix are used to receive multi-angle scattered light signals generated by particles in the dust-laden airflow from different angles; The filter assembly is installed at the inlet end of the air intake pipeline to filter out large-size particles in the air; The temperature and humidity sensor is installed in the air outlet pipeline and is used to collect the temperature signal and humidity signal at the location in real time; The micro fan is installed in the main air outlet; The positioning module 1, power module, communication module 1 and data processing module 1 are all installed in the accommodating cavity; the positioning module 1 is used to collect the position information of the location; the power module is respectively connected with the laser, the laser detection matrix, the temperature and humidity sensor, the micro fan, the positioning module 1, the data processing module 1 and the communication module 1, and the data processing module 1 includes a dust concentration calculation unit, a physical drive self-calibration unit and a database 1; the data processing module 1 is respectively connected with the laser, the laser detection matrix, the temperature and humidity sensor, the micro fan, the positioning module 1 and the communication module 1; the dust concentration calculation unit is used to calculate the dust concentration data; the physical drive self-calibration unit is used to perform a calibration process of the dust concentration data; The oscillating balance dust continuous measuring instrument is built with a positioning module 2, a data processing module 2 and a communication module 2; the positioning module 2 is used to collect the location information of the location; the data processing module 2 is connected to the positioning module 2 and the communication module 2 respectively; the oscillating balance dust continuous measuring instrument is used to monitor the dust concentration at the location in real time and obtain dust concentration data; The coal mine information processing center includes a central processor and a communication module three, wherein the central processor includes a machine learning self-calibration unit, a dust distribution cloud map unit for the entire mine area, and a database two; the central processor is connected to the communication module three; the communication module three is connected to the communication module one, and is used to establish a communication link between the coal mine information processing center and the low-cost dust monitor; the communication module three is also connected to the communication module two, and is used to establish a communication link between the coal mine information processing center and the oscillating balance dust continuous measuring instrument; the machine learning self-calibration unit is used to perform secondary calibration processing on the dust concentration data emitted by the low-cost dust monitor based on the dust concentration data of the oscillating balance dust continuous measuring instrument; the dust distribution cloud map unit for the entire mine area is used to generate a dust distribution cloud map for the entire mine area; The visual client is connected to the central processor via communication module three, and is used to receive and display in real time the dust distribution cloud map of the entire mine area.

[0007] As a preference, the filter component is a filter net.

[0008] As a preferred embodiment, both the positioning module 1 and the positioning module 2 are low-power positioning modules.

[0009] As a preference, the data processing module 1 is a PLC controller.

[0010] As a preferred embodiment, the shell is made of explosion-proof material, a power interface is provided on the shell, and the power interface is connected to the power module.

[0011] As a preference, the communication module 1 and the communication module 2 are connected wirelessly, and the communication module 2 and the communication module 3 are connected wiredly.

[0012] As a preferred embodiment, the visual client is designed for multi-platform intercommunication, and has the function of intercommunication with IOS platform, Android platform and Windows platform.

[0013] Furthermore, in order to prevent the inner surface of the light scattering chamber from adsorbing dust particles and to ensure the detection accuracy of the light scattering chamber, the inner surface of the light scattering chamber is coated with an anti-adsorption coating throughout.

[0014] In the present invention, for a low-cost dust monitor, by installing a filter assembly in the air intake pipeline, large-size particles in the air can be filtered out, so that large-size particles can be prevented from entering the light scattering chamber, thereby preventing large-size particles from polluting or blocking the light scattering chamber, and at the same time, preventing large-size particles from affecting the accuracy of monitoring data, so that the accuracy and stability of the light scattering chamber monitoring can be fully ensured. By arranging a micro fan in the main air outlet, the micro fan can provide negative pressure for the air outlet pipeline and the light scattering chamber, and then the negative pressure can be used to introduce the gas to be detected in the external environment into the light scattering chamber through the main air inlet and the air inlet pipeline, and discharged through the air outlet pipeline and the main air outlet. The continuously provided negative pressure can maintain the fluidity of the air inside the low-cost dust monitor, effectively reducing the residence time of the air and the dust therein inside the low-cost dust monitor, and can control the speed of the air flow by controlling the speed of the micro fan, which is conducive to improving the response speed and detection accuracy of the low-cost dust monitor to dust detection. At the same time, this can also help to remove the heat and moisture inside the low-cost dust monitor, ensuring that the low-cost dust monitor can always operate stably and reliably in a suitable environment. By setting a temperature and humidity sensor in the outlet pipe, it is convenient to collect the temperature signal and humidity signal of the air flowing through the outlet pipe in real time. Therefore, the data processing module 1 can obtain temperature data and humidity data based on the temperature signal and humidity signal, and then can be calibrated in combination with the temperature data and humidity data in the process of calculating the dust concentration data, so as to obtain more accurate dust concentration data. Through the setting of the positioning module 1, it is convenient to collect the position information of the location, so that the coal mine information processing center can obtain the position information of each low-cost dust monitor based on the position information. In the measurement chamber, a lens group is set between the laser and the central area of ​​the measurement chamber, so that the lens group can be used to focus the laser beam emitted by the laser on the central area of ​​the measurement chamber. In this way, the light intensity of the detection beam acting on the central area of ​​the measurement chamber can be greatly improved, so that when the dust-containing aerosol flows through the center of the measurement chamber, a higher intensity scattered light signal can be obtained, which is conducive to improving the accuracy of the detection data. A laser detection matrix is ​​set on one side of the center of the measurement chamber, and the laser detection matrix is ​​composed of three groups of silicon photodiodes distributed in sequence along the direction of airflow, and the detection surfaces of the three groups of silicon photodiodes are all facing the central area of ​​the measurement chamber. In this way, when the laser is generated by the laser and the lens group and focused on the dust-laden airflow flowing through the center of the chamber, multi-angle scattered light signals can be generated, and then the three groups of silicon photodiodes are used to receive the multi-angle scattered light signals from different angles, so that more accurate dust concentration data can be obtained based on the multi-angle information of the laser.Setting a light trap outside the laser detection matrix can fully capture and absorb the laser irradiated on the top side wall, thereby avoiding the interference of the light beam reflected from the top side wall on each silicon photodiode, effectively improving the clarity of the collected signal, and helping to fully ensure the measurement accuracy. By setting a physical drive self-calibration unit and a dust concentration calculation unit in the data processing module one, not only can the dust concentration calculation unit be used to calculate the dust concentration data, but also the physical drive self-calibration unit can be used for a calibration process of the calculated dust concentration data, thereby ensuring the accuracy of the data calculated by the dust concentration calculation unit. The low-cost dust monitor has a simple structure, low maintenance frequency, low manufacturing cost, and high detection accuracy, and it can meet the real-time monitoring needs of local dust concentration. The oscillating balance dust concentration continuous monitor is provided with a positioning module 2, which can facilitate the collection of location information at the location, so that the coal mine information processing center can obtain the location information of each low-cost dust monitor based on the location information. An oscillating balance dust concentration continuous monitor and multiple low-cost dust monitors are arranged in the same group of dust concentration monitoring components, and the oscillating balance dust concentration continuous monitor and the low-cost dust monitor are connected to the coal mine information processing center. A machine learning self-calibration unit is set in the central processor in the coal mine information processing center, so that the machine learning self-calibration unit can perform secondary calibration processing on the dust concentration data from the low-cost dust monitor based on the dust concentration data of the oscillating balance dust concentration continuous measuring instrument in the same group of dust concentration monitoring components, thereby effectively reducing the overall monitoring cost of the same monitoring area and fully ensuring the monitoring accuracy of dust at different positions in the same monitoring area, and then significantly reducing the overall monitoring cost when there are a large number of monitoring areas, and ensuring the overall monitoring accuracy. The coal mine information processing center is provided with a dust distribution cloud map unit for the overall area of ​​the mine, which can be convenient for generating a dust distribution cloud map for the overall area of ​​the mine based on the location information of the low-cost dust monitor and the dust concentration data after secondary calibration, as well as the location information and dust concentration data of the oscillating balance dust continuous measuring instrument. By using a visual client to display the dust distribution cloud map of the entire mine area in real time, relevant personnel can observe the dust concentration in the entire mine area in real time, and then take necessary response measures to abnormal areas in a timely manner.

[0015] For the dust data monitored by the low-cost dust monitor, the system, on the one hand, realizes a calibration of the dust concentration data during its first calculation process, and on the other hand, uses the machine learning self-calibration unit at the coal mine information processing center to perform a second real-time calibration based on the monitoring data of the oscillating balance dust continuous measuring instrument in the same monitoring area. In this way, the accuracy and stability of the monitoring data from the low-cost dust monitor can be fully improved by combining the secondary calibration, which can significantly reduce the monitoring error of the low-cost dust monitor and fully improve the measurement accuracy and reliability of the low-cost dust monitor under variable environmental conditions. In this way, in a larger monitoring area, only one high-cost oscillating balance dust continuous measuring instrument and multiple low-cost dust monitors need to be arranged, so that the synchronous monitoring operation of multiple monitoring points can be carried out by replacing the high-cost with the low-cost method, and then the data can be calibrated by the secondary calibration method, which can not only significantly reduce the monitoring cost, but also effectively ensure the accuracy of the monitoring data in the entire monitoring area. The system can achieve high-precision monitoring of dust concentration and ensure the accuracy and reliability of monitoring data through multiple calibration mechanisms, thereby accurately and intuitively generating dust distribution cloud maps for the entire mine area. At the same time, it takes into account the ease of use, ease of maintenance, cost-effectiveness and convenience of the observation process of the equipment, making it more suitable for monitoring needs in dust-polluted environments such as coal mines.

[0016] The present invention also provides a low-cost dual-drive self-calibration dust concentration online monitoring method, which uses a low-cost dual-drive self-calibration dust concentration online monitoring system; Step 1: Divide the site to be monitored into multiple monitoring areas, and arrange multiple groups of dust concentration monitoring components in multiple monitoring areas respectively. At the same time, multiple low-cost dust monitors and an oscillating balance dust continuous measuring instrument in the same group of dust concentration monitoring components are evenly distributed in the monitoring areas; Step 2: Establish a communication connection between multiple low-cost dust monitors in multiple dust concentration monitoring components and the coal mine information processing center; establish a communication connection between the oscillating balance dust continuous measuring instrument in multiple dust concentration monitoring components and the coal mine information processing center; establish a communication connection between the coal mine information processing center and the visual client; Step 3: Use multiple sets of dust concentration monitoring components to perform online dust concentration monitoring operations on multiple monitoring areas; S31: Use low-cost dust monitors to monitor dust at the location; S311: Control the micro fan to start working at the set power through the data processing module 1, and use the micro fan to provide negative pressure to allow the air to be detected in the environment to enter the low-cost dust monitor; at the same time, control the laser to emit a laser beam of a specific wavelength through the data processing module 1, and use the lens group to focus the laser beam and irradiate it on the dust-containing airflow flowing through the central area of ​​the measurement chamber. At the same time, use the three silicon photodiodes in the laser detection matrix to receive the multi-angle scattered light signals generated by the particles in the dust-containing airflow from different angles, and then send the received multi-angle scattered light signals to the data processing module 1; At the same time, the location information of the low-cost dust monitor is collected by the positioning module 1 and sent to the data processing module 1; the temperature and humidity sensors are used to collect the temperature and humidity signals of the air flowing out of the light scattering chamber in real time and send them to the data processing module 1; S312: The data processing module 1 obtains temperature data and humidity data according to the received temperature signal and humidity signal, obtains laser multi-angle information according to the received multi-angle scattered light signal, and sends the laser light source parameters, laser multi-angle information, temperature data and humidity data to the comprehensive data storage area 1 in the database 1 for storage. At the same time, the dust concentration calculation unit calculates the dust concentration data CA1 based on the laser light source parameters and laser multi-angle information, and in combination with the temperature data and humidity data of the gas to be measured; at the same time, in each monitoring cycle, the Xth n In the calculation process, the obtained Xth n The dust concentration data CA1 is sent to the physical drive self-calibration unit, and the physical drive self-calibration unit converts the Xth n The dust concentration data CA1 is compared with the dust concentration information data set in the calibration data storage area 1 in the database 1, wherein the calibration data storage area 1 includes dust concentration information data sets calculated by comprehensive factors such as different light source parameters, different temperature and humidity information, and different laser multi-angle information. If the comparison deviation is less than the set threshold, no action is taken. If the comparison deviation is greater than the set threshold, the physical-driven prediction model is used to analyze the relationship between the light source parameters, laser multi-angle information, temperature data and humidity data and the output of the dust concentration calculation unit to predict the expected behavior characteristics of the dust concentration calculation unit. At the same time, the calculation factors that need to be modified are calculated by inversion calculation and fed back to the dust concentration calculation unit. Then, through repeated iteration and optimization, the difference between the output value of the dust concentration calculation unit and the predicted value of the physical-driven model is controlled within the set error range. The data processing module 1 sends the location information, light source parameters, laser multi-angle information, temperature data, humidity data and dust concentration data CA1 of the low-cost dust monitor to the coal mine information processing center unit; S32: Monitor the dust at the location using an oscillating balance dust continuous measuring instrument; The location information of the oscillating balance dust continuous measuring instrument is collected by the positioning module 2, and sent to the data processing module 2; the dust concentration data CB at the location is obtained by the data processing module 2 in the oscillating balance dust continuous measuring instrument; the data processing module 2 sends the location information of the oscillating balance dust continuous measuring instrument and the dust concentration data CB to the coal mine information processing center; Step 4: Draw a dust distribution cloud map of the entire mine area; S41: After receiving the location information, light source parameters, laser multi-angle information, temperature data, humidity data and dust concentration data CA1 of the low-cost dust monitor, and the location information and dust concentration data CB of the oscillating balance dust continuous measuring instrument, the central data processor stores them in the comprehensive data storage area 2 in the database 2; S42: The central data processor cleans and standardizes the dust concentration data CA1 and the dust concentration data CB in the integrated data storage area 2 to eliminate the noise and outliers therein. In the process of eliminating outliers, for the dust concentration data from the same group of dust concentration monitoring components, the dust concentration data CB of the oscillating balance dust continuous measuring instrument is used as the reference data, and then the dust concentration data CA1 of the remaining multiple low-cost dust monitors are compared with the reference data. If the comparison difference is greater than the set difference range, it is confirmed to be an outlier and is removed; the screened high-quality and consistent data that meet the requirements are stored in the calibration data storage area 2 in the database 2, and used as a data set for building a machine learning calibration model; S43: The machine learning self-calibration unit divides the data set in the calibration data storage area 2 into a training set and a validation set according to a set ratio; uses light source parameters, laser multi-angle information, temperature data and humidity data as feature data, and adopts a support vector machine machine learning algorithm with higher prediction accuracy and generalization ability in dust concentration monitoring to build a machine learning calibration model; then uses the training set to train the machine learning calibration model to obtain a trained machine learning calibration model; then, minimizes the prediction error by adjusting the model parameters to obtain an optimized machine learning calibration model; finally, uses the validation set to verify the optimized machine learning calibration model, and further adjusts and optimizes the machine learning calibration model parameters according to the verification results to obtain a data-driven machine learning calibration model; S44: using the dust concentration data CA1 received in real time from the low-cost dust monitor as input data, inputting it into the machine learning calibration model, performing machine learning self-calibration on the input data through the machine learning calibration model, and outputting the calibrated dust concentration data CA; S45: The central data processor stores the location information and calibrated dust concentration data CA of the low-cost dust monitor, and the location information and dust concentration data CB of the oscillating balance dust continuous measuring instrument in the dust cloud data storage area in the second database; S46: The dust distribution cloud map unit of the whole mine area draws the dust distribution cloud map of the whole mine area based on the location information and calibrated dust concentration data CA of the low-cost dust monitor, and the location information and dust concentration data CB of the oscillating balance dust continuous measuring instrument, and sends the dust distribution cloud map of the whole mine area to the visual client; wherein, the spatial dust concentration prediction network model is established by means of a massive position combination experiment, and the specific process is as follows: S461: Use an oscillating balance dust continuous measuring instrument to monitor dust concentration at different locations, collect and record dust concentration data and corresponding location information through a large number of measurement processes, and use a large amount of dust concentration data and corresponding location information as a sample data set; S462: Divide the sample data set into a training set and a validation set according to a set ratio; S463: Establishing a spatial dust concentration prediction network model, and training it using a training set to obtain a trained spatial dust concentration prediction network model; S464: Adjust and optimize model parameters to improve the prediction accuracy of the spatial dust concentration prediction network model; S465: Use the validation set to validate the spatial dust concentration prediction network model. At the same time, the accuracy and reliability of the model are evaluated by comparing the dust concentration data predicted by the spatial dust concentration prediction network model with the actual observed values. The model parameters are further adjusted and optimized according to the evaluation results to obtain the final spatial dust concentration prediction network model. Step 5: The visual client displays the received dust distribution cloud map of the entire mine area in real time.

[0017] Furthermore, in order to ensure the accuracy of calibration, in step three S312, the physically driven prediction model is established based on Monte Carlo simulation and Mie scattering theory.

[0018] The present invention sets a high-cost oscillating balance dust continuous measuring instrument and multiple low-cost dust monitors in a monitoring area, and adopts a "double-drive" calibration model. Specifically, the low-cost dust monitor part uses a physically driven prediction model to regularly analyze the relationship between light source parameters, laser multi-angle information, temperature data and humidity data and the output of the dust concentration calculation unit, and then regularly performs physical self-calibration on the calculation factors in the dust concentration calculation unit to achieve the initial calibration of the dust concentration data, so as to fully improve the monitoring accuracy of its own dust concentration data. At the same time, at the coal mine information processing center, a data-driven machine learning calibration model is used to perform a secondary real-time calibration on the dust concentration data monitored by the low-cost dust monitor each time, which significantly improves the efficiency and accuracy of coal mine safety monitoring and reduces economic costs. Through the effective combination of the two calibration methods, the measurement accuracy and stability of the monitoring data are significantly improved, and the error is reduced. In this way, in a larger monitoring area, only one high-cost oscillating balance dust continuous measuring instrument and multiple low-cost dust monitors need to be arranged to realize the synchronous monitoring operation of multiple monitoring points, and then multiple low-cost dust monitors can be used to perform alternative monitoring operations of high-cost monitoring equipment, which can not only significantly reduce the overall monitoring cost, but also effectively ensure the accuracy of monitoring data. Finally, the dust distribution cloud map unit of the overall area of ​​the mine is used to generate the overall area dust distribution cloud map of the mine based on the location information of the low-cost dust monitor and the dust concentration data after secondary calibration, as well as the location information and dust concentration data of the oscillating balance dust continuous measuring instrument, and send it to the visual client, and display it intuitively through the visual client, which can facilitate the relevant personnel to understand the dust concentration in different monitoring areas in the entire monitoring site in real time, and then facilitate the coordinated realization of real-time and reliable monitoring of the dust concentration of the entire monitoring site.

[0019] The present invention adopts a real-time "dual-drive" correction method, applies physical-driven and data-driven models to realize self-calibration of low-cost dust monitors, and utilizes a large number of low-cost dust monitors to replace a large number of high-cost oscillating balance dust continuous measuring instruments through substitution. While improving the accuracy of dust monitoring, it can significantly reduce economic costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a principle block diagram of the present invention; Figure 2 is a schematic diagram of the present invention arranged in a monitoring location; Figure 3 It is a structural schematic diagram of the low-cost dust monitor of the present invention; Figure 4 It is a principle block diagram of the data processing module 1 in the present invention; Figure 5It is a principle block diagram of the central processing unit in the present invention.

[0021] In the figure: 1. shell, 2. light scattering chamber, 3. laser, 4. lens group, 5. light trap, 6. laser detection matrix, 7. temperature and humidity sensor, 8. micro fan, 9. positioning module one, 10. data processing module one, 11. communication module one, 12. power module, 13. filter component, 14. main air inlet, 15. main air outlet, 16. air inlet pipeline, 17. air outlet pipeline. DETAILED DESCRIPTION

[0022] The present invention will be further described below in conjunction with the accompanying drawings.

[0023] like Figures 1 to 5 As shown, the present invention provides a low-cost dual-drive self-calibration dust concentration online monitoring system, including multiple groups of dust concentration monitoring components, a coal mine information processing center and a visual client; The dust concentration monitoring assembly includes an oscillating balance dust continuous measuring instrument and a plurality of low-cost dust monitors; The low-cost dust monitor includes a housing 1, a light scattering chamber 2, a laser 3, a lens group 4, a light trap 5, a laser detection matrix 6, a filter component 13, a temperature and humidity sensor 7, a micro fan 8, a positioning module 9, a power module 12, a communication module 11 and a data processing module 10; The housing 1 forms a receiving chamber inside, and a main air inlet 14 and a main air outlet 15 communicating with the receiving chamber are provided at the front center and the rear center opposite to each other. The light scattering chamber 2 is located in the accommodating chamber, and a measuring chamber is formed inside the chamber. A measuring inlet and a measuring outlet are respectively provided at the front and rear ends of the chamber at positions corresponding to the central area of ​​the measuring chamber, and the measuring inlet is connected to the main air inlet 14 through an air inlet pipeline 16, and the measuring outlet is connected to the main air outlet 15 through an air outlet pipeline 17; As a preferred embodiment, the main air inlet 14, the measuring inlet, the measuring outlet and the main air outlet 15 are all located in the same horizontal plane and are coaxially distributed; The laser 3 is installed at the center of the bottom side wall of the housing 1, and its output end is located inside the measuring chamber. At the same time, the laser beam emitted by the laser 3 passes through the central area of ​​the measuring chamber and is perpendicular to the axis of the measuring chamber. The lens group 4 is fixedly installed between the central area of ​​the measurement chamber and the output end of the laser 3, and is used to focus the laser beam output by the laser 3 on the central area of ​​the measurement chamber; The light trap 5 is located in the measurement chamber and is mounted on the top side wall of the housing 1 opposite to the laser 3; The laser detection matrix 6 is located in the measurement chamber and is installed between the central area of ​​the measurement chamber and the light trap 5; the laser detection matrix 6 is composed of three groups of silicon photodiodes, wherein the first group of silicon photodiodes is distributed relative to the laser 3, and its detection surface is located on the same straight line as the output direction of the laser 3, and the second group of silicon photodiodes and the third group of silicon photodiodes are symmetrically distributed on both sides of the first group of silicon photodiodes, and their detection surfaces are all facing the central area of ​​the measurement chamber; the three silicon photodiodes in the laser detection matrix 6 are used to receive multi-angle scattered light signals generated by particles in the dust-laden airflow from different angles; The filter assembly 13 is installed at the inlet end of the air intake pipe 16 to filter out large-size particles in the air; The temperature and humidity sensor 7 is installed in the air outlet pipeline 17 and is used to collect the temperature signal and humidity signal at the location in real time; The micro fan 8 is installed in the main air outlet 14; The positioning module 19, the power module 12, the communication module 11 and the data processing module 10 are all installed in the accommodating cavity; the positioning module 19 is used to collect the position information of the location, and as a preferred embodiment, the positioning module 19 is a low-power positioning module; the power module 12 is respectively connected to the laser 3, the laser detection matrix 6, the temperature and humidity sensor 7, the micro fan 8, the positioning module 19, the data processing module 10 and the communication module 11, and the data processing module 10 includes a dust concentration calculation unit, a physical drive self-calibration unit and a database 1; the data processing module 10 is respectively connected to the laser 3, the laser detection matrix 6, the temperature and humidity sensor 7, the micro fan 8, the positioning module 19 and the communication module 11; the dust concentration calculation unit is used to calculate the dust concentration data; the physical drive self-calibration unit is used to perform a primary calibration process of the dust concentration data; The oscillating balance dust continuous measuring instrument is built with a positioning module 2, a data processing module 2 and a communication module 2; the positioning module 2 is used to collect the location information of the location, and as a preferred embodiment, the positioning module 2 is a low-power positioning module; the data processing module 2 is connected to the positioning module 2 and the communication module 2 respectively; the oscillating balance dust continuous measuring instrument is a high-precision dust measuring instrument that already exists in the prior art, and preferably a direct-reading oscillating balance dust continuous measuring instrument is used to monitor the dust concentration at the location in real time and output dust concentration data; for the oscillating balance dust continuous monitor, a replaceable filter membrane is installed at the oscillating end of its oscillating hollow conical tube. When the sampling airflow passes through the filter membrane, the dust therein will be deposited on the filter membrane, and the mass of the filter membrane will change, resulting in a change in the oscillation frequency. The mass concentration of the dust deposited on the filter membrane is calculated by the change in the oscillation frequency, and then the dust concentration data CB is obtained. The data processing module 2 sends the obtained dust concentration data CB and the received location information to the coal mine information processing center; The coal mine information processing center includes a central processor and a communication module three, wherein the central processor includes a machine learning self-calibration unit, a dust distribution cloud map unit for the entire mine area, and a database two; the central processor is connected to the communication module three; the communication module three is connected to the communication module one, and is used to establish a communication link between the coal mine information processing center and the low-cost dust monitor; the communication module three is also connected to the communication module two, and is used to establish a communication link between the coal mine information processing center and the oscillating balance dust continuous measuring instrument; the machine learning self-calibration unit is used to perform secondary calibration processing on the dust concentration data emitted by the low-cost dust monitor based on the dust concentration data of the oscillating balance dust continuous measuring instrument; the dust distribution cloud map unit for the entire mine area is used to generate a dust distribution cloud map for the entire mine area; The visual client is connected to the central processor via communication module three, and is used to receive and display in real time the dust distribution cloud map of the entire mine area.

[0024] As a preferred embodiment, the filter component 13 is a filter net.

[0025] As a preference, the data processing module 10 is a PLC controller.

[0026] As a preferred embodiment, the shell 1 is made of explosion-proof material, a power interface is provided on the shell 1, and the power interface is connected to the power module 12.

[0027] As a preference, the communication module 1 11 and the communication module 2 are connected wirelessly, and the communication module 2 and the communication module 3 are connected wiredly.

[0028] As a preferred embodiment, the visual client is designed for multi-platform intercommunication, and has the function of intercommunication with IOS platform, Android platform and Windows platform.

[0029] In order to prevent the inner surface of the light scattering chamber from adsorbing dust particles and to ensure the detection accuracy of the light scattering chamber, the inner surface of the light scattering chamber 2 is coated with an anti-adsorption coating.

[0030] In the present invention, for a low-cost dust monitor, by installing a filter assembly in the air intake pipeline, large-size particles in the air can be filtered out, so that large-size particles can be prevented from entering the light scattering chamber, thereby preventing large-size particles from polluting or blocking the light scattering chamber, and at the same time, preventing large-size particles from affecting the accuracy of monitoring data, so that the accuracy and stability of the light scattering chamber monitoring can be fully ensured. By arranging a micro fan in the main air outlet, the micro fan can provide negative pressure for the air outlet pipeline and the light scattering chamber, and then the negative pressure can be used to introduce the gas to be detected in the external environment into the light scattering chamber through the main air inlet and the air inlet pipeline, and discharged through the air outlet pipeline and the main air outlet. The continuously provided negative pressure can maintain the fluidity of the air inside the low-cost dust monitor, effectively reducing the residence time of the air and the dust therein inside the low-cost dust monitor, and can control the speed of the air flow by controlling the speed of the micro fan, which is conducive to improving the response speed and detection accuracy of the low-cost dust monitor to dust detection. At the same time, this can also help to remove the heat and moisture inside the low-cost dust monitor, ensuring that the low-cost dust monitor can always operate stably and reliably in a suitable environment. By setting a temperature and humidity sensor in the outlet pipe, it is convenient to collect the temperature signal and humidity signal of the air flowing through the outlet pipe in real time. Therefore, the data processing module 1 can obtain temperature data and humidity data based on the temperature signal and humidity signal, and then can be calibrated in combination with the temperature data and humidity data in the process of calculating the dust concentration data, so as to obtain more accurate dust concentration data. Through the setting of the positioning module 1, it is convenient to collect the position information of the location, so that the coal mine information processing center can obtain the position information of each low-cost dust monitor based on the position information. In the measurement chamber, a lens group is set between the laser and the central area of ​​the measurement chamber, so that the lens group can be used to focus the laser beam emitted by the laser on the central area of ​​the measurement chamber. In this way, the light intensity of the detection beam acting on the central area of ​​the measurement chamber can be greatly improved, so that when the dust-containing aerosol flows through the center of the measurement chamber, a higher intensity scattered light signal can be obtained, which is conducive to improving the accuracy of the detection data. A laser detection matrix is ​​set on one side of the center of the measurement chamber, and the laser detection matrix is ​​composed of three groups of silicon photodiodes distributed in sequence along the direction of airflow, and the detection surfaces of the three groups of silicon photodiodes are all facing the central area of ​​the measurement chamber. In this way, when the laser is generated by the laser and the lens group and focused on the dust-laden airflow flowing through the center of the chamber, multi-angle scattered light signals can be generated, and then the three groups of silicon photodiodes are used to receive the multi-angle scattered light signals from different angles, so that more accurate dust concentration data can be obtained based on the multi-angle information of the laser.Setting a light trap outside the laser detection matrix can fully capture and absorb the laser irradiated on the top side wall, thereby avoiding the interference of the light beam reflected from the top side wall on each silicon photodiode, effectively improving the clarity of the collected signal, and helping to fully ensure the measurement accuracy. By setting a physical drive self-calibration unit and a dust concentration calculation unit in the data processing module one, not only can the dust concentration calculation unit be used to calculate the dust concentration data, but also the physical drive self-calibration unit can be used for a calibration process of the calculated dust concentration data, thereby ensuring the accuracy of the data calculated by the dust concentration calculation unit. The low-cost dust monitor has a simple structure, low maintenance frequency, low manufacturing cost, and high detection accuracy, and it can meet the real-time monitoring needs of local dust concentration. The oscillating balance dust concentration continuous monitor is provided with a positioning module 2, which can facilitate the collection of location information at the location, so that the coal mine information processing center can obtain the location information of each low-cost dust monitor based on the location information. An oscillating balance dust concentration continuous monitor and multiple low-cost dust monitors are arranged in the same group of dust concentration monitoring components, and the oscillating balance dust concentration continuous monitor and the low-cost dust monitor are connected to the coal mine information processing center. A machine learning self-calibration unit is set in the central processor in the coal mine information processing center, so that the machine learning self-calibration unit can perform secondary calibration processing on the dust concentration data from the low-cost dust monitor based on the dust concentration data of the oscillating balance dust concentration continuous measuring instrument in the same group of dust concentration monitoring components, thereby effectively reducing the overall monitoring cost of the same monitoring area and fully ensuring the monitoring accuracy of dust at different positions in the same monitoring area, and then significantly reducing the overall monitoring cost when there are a large number of monitoring areas, and ensuring the overall monitoring accuracy. The coal mine information processing center is provided with a dust distribution cloud map unit for the overall area of ​​the mine, which can be convenient for generating a dust distribution cloud map for the overall area of ​​the mine based on the location information of the low-cost dust monitor and the dust concentration data after secondary calibration, as well as the location information and dust concentration data of the oscillating balance dust continuous measuring instrument. By using a visual client to display the dust distribution cloud map of the entire mine area in real time, relevant personnel can observe the dust concentration in the entire mine area in real time, and then take necessary response measures to abnormal areas in a timely manner.

[0031] For the dust data monitored by the low-cost dust monitor, the system, on the one hand, realizes a calibration of the dust concentration data during its first calculation process, and on the other hand, uses the machine learning self-calibration unit at the coal mine information processing center to perform a second real-time calibration based on the monitoring data of the oscillating balance dust continuous measuring instrument in the same monitoring area. In this way, the accuracy and stability of the monitoring data from the low-cost dust monitor can be fully improved by combining the secondary calibration, which can significantly reduce the monitoring error of the low-cost dust monitor and fully improve the measurement accuracy and reliability of the low-cost dust monitor under variable environmental conditions. In this way, in a larger monitoring area, only one high-cost oscillating balance dust continuous measuring instrument and multiple low-cost dust monitors need to be arranged, so that the synchronous monitoring operation of multiple monitoring points can be carried out by replacing the high-cost with the low-cost method, and then the data can be calibrated by the secondary calibration method, which can not only significantly reduce the monitoring cost, but also effectively ensure the accuracy of the monitoring data in the entire monitoring area. The system can achieve high-precision monitoring of dust concentration and ensure the accuracy and reliability of monitoring data through multiple calibration mechanisms, thereby accurately and intuitively generating dust distribution cloud maps for the entire mine area. At the same time, it takes into account the ease of use, ease of maintenance, cost-effectiveness and convenience of the observation process of the equipment, making it more suitable for monitoring needs in dust-polluted environments such as coal mines.

[0032] The present invention also provides a low-cost dual-drive self-calibration dust concentration online monitoring method, which uses a low-cost dual-drive self-calibration dust concentration online monitoring system; Step 1: Divide the site to be monitored into multiple monitoring areas, and arrange multiple groups of dust concentration monitoring components in multiple monitoring areas respectively. At the same time, multiple low-cost dust monitors and an oscillating balance dust continuous measuring instrument in the same group of dust concentration monitoring components are evenly distributed in the monitoring areas; Step 2: Establish a communication connection between multiple low-cost dust monitors in multiple dust concentration monitoring components and the coal mine information processing center; establish a communication connection between the oscillating balance dust continuous measuring instrument in multiple dust concentration monitoring components and the coal mine information processing center; establish a communication connection between the coal mine information processing center and the visual client; Step 3: Use multiple sets of dust concentration monitoring components to perform online dust concentration monitoring operations on multiple monitoring areas; S31: Use low-cost dust monitors to monitor dust at the location; S311: Control the micro fan 8 to start working at the set power through the data processing module 10, and use the micro fan 8 to provide negative pressure so that the air to be detected in the environment enters the low-cost dust monitor; at the same time, control the laser 3 to emit a laser beam of a specific wavelength through the data processing module 10, and use the lens group 4 to focus the laser beam and irradiate it on the dust-containing airflow flowing through the central area of ​​the measurement chamber. At the same time, use the three silicon photodiodes in the laser detection matrix 6 to receive the multi-angle scattered light signals generated by the particles in the dust-containing airflow from different angles, and then send the received multi-angle scattered light signals to the data processing module 10; At the same time, the location information of the low-cost dust monitor is collected by the positioning module 9 and sent to the data processing module 10; the temperature and humidity sensor 7 is used to collect the temperature signal and humidity signal of the air flowing out of the light scattering chamber 2 in real time and send them to the data processing module 10; S312: The data processing module 10 obtains temperature data and humidity data according to the received temperature signal and humidity signal, obtains laser multi-angle information according to the received multi-angle scattered light signal, and sends the laser 3 light source parameters, laser multi-angle information, temperature data and humidity data to the comprehensive data storage area 1 in the database 1 for storage. At the same time, the dust concentration calculation unit calculates the dust concentration data CA1 based on the light source parameters and laser multi-angle information of the laser 3 and the temperature data and humidity data of the gas to be measured; at the same time, in each monitoring cycle, the Xth n (X n is a positive integer, which can be set and adjusted according to user needs) in the calculation process, the obtained Xth n The dust concentration data CA1 is sent to the physical drive self-calibration unit, and the physical drive self-calibration unit converts the Xth n The dust concentration data CA1 is compared with the dust concentration information data set in the calibration data storage area 1 in the database 1, wherein the calibration data storage area 1 includes dust concentration information data sets calculated by comprehensive factors such as different light source parameters, different temperature and humidity information, and different laser multi-angle information. If the comparison deviation is less than the set threshold, no action is taken. If the comparison deviation is greater than the set threshold, the physical-driven prediction model is used to analyze the relationship between the light source parameters (wavelength, refractive index, etc.), laser multi-angle information, temperature data and humidity data and the output of the dust concentration calculation unit to predict the expected behavior characteristics of the dust concentration calculation unit. At the same time, the calculation factors that need to be modified are calculated by inversion calculation and fed back to the dust concentration calculation unit. Then, through repeated iteration and optimization, the difference between the output value of the dust concentration calculation unit and the predicted value of the physical-driven model is controlled within the set error range. The data processing module 10 sends the location information, light source parameters, laser multi-angle information, temperature data, humidity data and dust concentration data CA1 of the low-cost dust monitor to the coal mine information processing center unit; S32: Monitor the dust at the location using an oscillating balance dust continuous measuring instrument; The location information of the oscillating balance dust continuous measuring instrument is collected by the positioning module 2, and sent to the data processing module 2; the dust concentration data CB at the location is obtained by the data processing module 2 in the oscillating balance dust continuous measuring instrument; the data processing module 2 sends the location information of the oscillating balance dust continuous measuring instrument and the dust concentration data CB to the coal mine information processing center; Step 4: Draw a dust distribution cloud map of the entire mine area; S41: After receiving the location information, light source parameters, laser multi-angle information, temperature data, humidity data and dust concentration data CA1 of the low-cost dust monitor, and the location information and dust concentration data CB of the oscillating balance dust continuous measuring instrument, the central data processor stores them in the comprehensive data storage area 2 in the database 2; S42: The central data processor cleans and standardizes the dust concentration data CA1 and the dust concentration data CB in the integrated data storage area 2 to eliminate the noise and outliers therein. In the process of eliminating outliers, for the dust concentration data from the same group of dust concentration monitoring components, the dust concentration data CB of the oscillating balance dust continuous measuring instrument is used as the reference data, and then the dust concentration data CA1 of the remaining multiple low-cost dust monitors are compared with the reference data. If the comparison difference is greater than the set difference range, it is confirmed to be an outlier and is removed; the screened high-quality and consistent data that meet the requirements are stored in the calibration data storage area 2 in the database 2, and used as a data set for building a machine learning calibration model; S43: The machine learning self-calibration unit divides the data set in the calibration data storage area 2 into a training set and a validation set according to a set ratio; uses light source parameters, laser multi-angle information, temperature data and humidity data as feature data, and adopts a support vector machine (SVM) machine learning algorithm with higher prediction accuracy and generalization ability in dust concentration monitoring to build a machine learning calibration model; then uses the training set to train the machine learning calibration model to obtain a trained machine learning calibration model; then, minimizes the prediction error by adjusting the model parameters to obtain an optimized machine learning calibration model; finally, uses the validation set to verify the optimized machine learning calibration model to evaluate the performance of the model, and further adjusts and optimizes the machine learning calibration model parameters according to the verification results to improve the accuracy of the model, so as to obtain a data-driven machine learning calibration model; S44: using the dust concentration data CA1 received in real time from the low-cost dust monitor as input data, inputting it into the machine learning calibration model, performing machine learning self-calibration on the input data through the machine learning calibration model, and outputting the calibrated dust concentration data CA; S45: The central data processor stores the location information and calibrated dust concentration data CA of the low-cost dust monitor, and the location information and dust concentration data CB of the oscillating balance dust continuous measuring instrument in the dust cloud data storage area in the second database; S46: The dust distribution cloud map unit of the whole mine area draws the dust distribution cloud map of the whole mine area based on the location information and calibrated dust concentration data CA of the low-cost dust monitor, and the location information and dust concentration data CB of the oscillating balance dust continuous measuring instrument, and sends the dust distribution cloud map of the whole mine area to the visual client; wherein, the spatial dust concentration prediction network model is established by means of a massive position combination experiment, and the specific process is as follows: S461: Use an oscillating balance dust continuous measuring instrument to monitor dust concentration at different locations, collect and record dust concentration data and corresponding location information through a large number of measurement processes, and use a large amount of dust concentration data and corresponding location information as a sample data set; S462: Divide the sample data set into a training set and a validation set according to a set ratio; S463: Establishing a spatial dust concentration prediction network model, and training it using a training set to obtain a trained spatial dust concentration prediction network model; S464: Adjust and optimize model parameters to improve the prediction accuracy of the spatial dust concentration prediction network model; S465: Use the validation set to validate the spatial dust concentration prediction network model. At the same time, the accuracy and reliability of the model are evaluated by comparing the dust concentration data predicted by the spatial dust concentration prediction network model with the actual observed values. The model parameters are further adjusted and optimized according to the evaluation results to obtain the final spatial dust concentration prediction network model. Step 5: The visual client displays the received dust distribution cloud map of the entire mine area in real time.

[0033] In order to ensure the accuracy of calibration, in step 3 S312, the physical-driven prediction model is established based on Monte Carlo simulation and Mie scattering theory. The physical-driven prediction model can comprehensively analyze the light source parameters (wavelength, refractive index, etc.) and laser multi-angle information, and at the same time combine the influence of multiple variables such as external environmental factors (temperature, relative humidity, etc.) on the response of the dust concentration calculation unit. And according to the relationship between the above variables and the output of the dust concentration calculation unit, the expected behavior characteristics of the particle dust concentration calculation unit can be predicted, thereby constructing a calibration framework that can reflect the actual monitoring environment.

[0034] The present invention sets a high-cost oscillating balance dust continuous measuring instrument and multiple low-cost dust monitors in a monitoring area, and adopts a "double-drive" calibration model. Specifically, the low-cost dust monitor part uses a physically driven prediction model to regularly analyze the relationship between light source parameters, laser multi-angle information, temperature data and humidity data and the output of the dust concentration calculation unit, and then regularly performs physical self-calibration on the calculation factors in the dust concentration calculation unit to achieve the initial calibration of the dust concentration data, so as to fully improve the monitoring accuracy of its own dust concentration data. At the same time, at the coal mine information processing center, a data-driven machine learning calibration model is used to perform a secondary real-time calibration on the dust concentration data monitored by the low-cost dust monitor each time, which significantly improves the efficiency and accuracy of coal mine safety monitoring and reduces economic costs. Through the effective combination of the two calibration methods, the measurement accuracy and stability of the monitoring data are significantly improved, and the error is reduced. In this way, in a larger monitoring area, only one high-cost oscillating balance dust continuous measuring instrument and multiple low-cost dust monitors need to be arranged to realize the synchronous monitoring operation of multiple monitoring points, and then multiple low-cost dust monitors can be used to perform alternative monitoring operations of high-cost monitoring equipment, which can not only significantly reduce the overall monitoring cost, but also effectively ensure the accuracy of monitoring data. Finally, the dust distribution cloud map unit of the overall area of ​​the mine is used to generate the overall area dust distribution cloud map of the mine based on the location information of the low-cost dust monitor and the dust concentration data after secondary calibration, as well as the location information and dust concentration data of the oscillating balance dust continuous measuring instrument, and send it to the visual client, and display it intuitively through the visual client, which can facilitate the relevant personnel to understand the dust concentration in different monitoring areas in the entire monitoring site in real time, and then facilitate the coordinated realization of real-time and reliable monitoring of the dust concentration of the entire monitoring site.

[0035] The present invention adopts a real-time "dual-drive" correction method, applies physical-driven and data-driven models to realize self-calibration of low-cost dust monitors, and utilizes a large number of low-cost dust monitors to replace a large number of high-cost oscillating balance dust continuous measuring instruments through substitution. While improving the accuracy of dust monitoring, it can significantly reduce economic costs.

Claims

1. A low-cost dual-drive self-calibration dust concentration online monitoring system, comprising multiple groups of dust concentration monitoring components, characterized in that: It also includes coal mine information processing center and visual client; The dust concentration monitoring assembly includes an oscillating balance dust continuous measuring instrument and a plurality of low-cost dust monitors; The low-cost dust monitor comprises a housing (1), a light scattering chamber (2), a laser (3), a lens group (4), a light trap (5), a laser detection matrix (6), a filter component (13), a temperature and humidity sensor (7), a micro fan (8), a positioning module (9), a power module (12), a communication module (11) and a data processing module (10); The housing (1) has a receiving chamber formed inside, and a main air inlet (14) and a main air outlet (15) communicating with the receiving chamber are provided at the front center and the rear center opposite to each other; The light scattering chamber (2) is located in the accommodating chamber, and a measuring chamber is formed inside the chamber. A measuring inlet and a measuring outlet are respectively provided at the front and rear ends of the chamber at positions corresponding to the central area of ​​the measuring chamber, and the measuring inlet is connected to the main air inlet (14) via an air inlet pipeline (16), and the measuring outlet is connected to the main air outlet (15) via an air outlet pipeline (17); The laser (3) is mounted at the center of the bottom side wall of the housing (1), and its output end is located inside the measuring chamber. At the same time, the laser beam emitted by the laser (3) passes through the center area of ​​the measuring chamber and is perpendicular to the axis of the measuring chamber. The lens group (4) is fixedly mounted between the central area of ​​the measurement chamber and the output end of the laser (3), and is used to focus the laser beam output by the laser (3) onto the central area of ​​the measurement chamber; The light trap (5) is located in the measurement chamber and is mounted on the top side wall of the housing (1) opposite to the laser (3); The laser detection matrix (6) is located in the measurement chamber and is installed between the central area of ​​the measurement chamber and the light trap (5); the laser detection matrix (6) is composed of three groups of silicon photodiodes, wherein the first group of silicon photodiodes is distributed relative to the laser (3), and its detection surface is located on the same straight line as the output direction of the laser (3); the second group of silicon photodiodes and the third group of silicon photodiodes are symmetrically distributed on both sides of the first group of silicon photodiodes, and their detection surfaces are all facing the central area of ​​the measurement chamber; the three silicon photodiodes in the laser detection matrix (6) are used to receive multi-angle scattered light signals generated by particles in the dust-laden airflow from different angles; The filter assembly (13) is installed at the inlet end of the air intake pipeline (16) and is used to filter out large-size particles in the air; The temperature and humidity sensor (7) is installed in the air outlet pipeline (17) and is used to collect temperature signals and humidity signals at the location in real time; The micro fan (8) is installed in the main air outlet (14); The positioning module (9), power module (12), communication module (11) and data processing module (10) are all installed in the accommodating cavity; the positioning module (9) is used to collect position information at the location; the power module (12) is respectively connected to the laser (3), the laser detection matrix (6), the temperature and humidity sensor (7), the micro fan (8), the positioning module (9), the data processing module (10) and the communication module (11); the data processing module (10) comprises a dust concentration calculation unit, a physical drive self-calibration unit and a database; the data processing module (10) is respectively connected to the laser (3), the laser detection matrix (6), the temperature and humidity sensor (7), the micro fan (8), the positioning module (9) and the communication module (11); the dust concentration calculation unit is used to calculate dust concentration data; the physical drive self-calibration unit is used to perform a primary calibration process on the dust concentration data; The oscillating balance dust continuous measuring instrument is built with a positioning module 2, a data processing module 2 and a communication module 2; the positioning module 2 is used to collect the location information of the location; the data processing module 2 is connected to the positioning module 2 and the communication module 2 respectively; the oscillating balance dust continuous measuring instrument is used to monitor the dust concentration at the location in real time and obtain dust concentration data; The coal mine information processing center includes a central processor and a communication module three, wherein the central processor includes a machine learning self-calibration unit, a dust distribution cloud map unit for the entire mine area, and a database two; the central processor is connected to the communication module three; the communication module three is connected to the communication module one, and is used to establish a communication link between the coal mine information processing center and the low-cost dust monitor; the communication module three is also connected to the communication module two, and is used to establish a communication link between the coal mine information processing center and the oscillating balance dust continuous measuring instrument; the machine learning self-calibration unit is used to perform secondary calibration processing on the dust concentration data emitted by the low-cost dust monitor based on the dust concentration data of the oscillating balance dust continuous measuring instrument; the dust distribution cloud map unit for the entire mine area is used to generate a dust distribution cloud map for the entire mine area; The visual client is connected to the central processor via communication module three, and is used to receive and display in real time the dust distribution cloud map of the entire mine area.

2. According to claim 1, a low-cost dual-drive self-calibration dust concentration online monitoring system is characterized in that: The filter component (13) is a filter screen.

3. A low-cost dual-drive self-calibration dust concentration online monitoring system according to claim 2, characterized in that: The positioning module 1 (9) and the positioning module 2 are both low-power positioning modules.

4. According to claim 3, a low-cost dual-drive self-calibration dust concentration online monitoring system is characterized in that: The data processing module 1 (10) is a PLC controller.

5. A low-cost dual-drive self-calibration dust concentration online monitoring system according to claim 1, characterized in that: The housing (1) is made of explosion-proof material, and a power interface is provided on the housing (1), and the power interface is connected to the power module (12).

6. A low-cost dual-drive self-calibration dust concentration online monitoring system according to claim 1, characterized in that: The communication module 1 (11) is connected to the communication module 2 in a wireless manner, and the communication module 2 is connected to the communication module 3 in a wired manner.

7. A low-cost dual-drive self-calibration dust concentration online monitoring system according to claim 1, characterized in that: The visual client is designed for multi-platform intercommunication, and has the function of intercommunication with IOS platform, Android platform and Windows platform.

8. A low-cost dual-drive self-calibration dust concentration online monitoring system according to claim 7, characterized in that: The inner surface of the light scattering chamber (2) is coated with an anti-adsorption coating throughout.

9. A low-cost dual-drive self-calibration dust concentration online monitoring method, using a low-cost dual-drive self-calibration dust concentration online monitoring system as claimed in any one of claims 1 to 8, characterized in that ； Step 1: Divide the site to be monitored into multiple monitoring areas, and arrange multiple groups of dust concentration monitoring components in multiple monitoring areas respectively. At the same time, multiple low-cost dust monitors and an oscillating balance dust continuous measuring instrument in the same group of dust concentration monitoring components are evenly distributed in the monitoring areas; Step 2: Establish a communication connection between multiple low-cost dust monitors in multiple dust concentration monitoring components and the coal mine information processing center; establish a communication connection between the oscillating balance dust continuous measuring instrument in multiple dust concentration monitoring components and the coal mine information processing center; establish a communication connection between the coal mine information processing center and the visual client; Step 3: Use multiple sets of dust concentration monitoring components to perform online dust concentration monitoring operations on multiple monitoring areas; S31: Use low-cost dust monitors to monitor dust at the location; S311: Controlling the micro fan (8) to start working at a set power through the data processing module (10), and using the micro fan (8) to provide negative pressure so that the air to be detected in the environment enters the low-cost dust monitor; at the same time, controlling the laser (3) to emit a laser beam of a specific wavelength through the data processing module (10), and using the lens group (4) to focus the laser beam and irradiate it on the dust-containing airflow flowing through the central area of ​​the measurement chamber, and at the same time, using the three silicon photodiodes in the laser detection matrix (6) to receive multi-angle scattered light signals generated by particles in the dust-containing airflow from different angles, and then sending the received multi-angle scattered light signals to the data processing module (10); At the same time, the location information of the low-cost dust monitor is collected by using the positioning module 1 (9) and sent to the data processing module 1 (10); the temperature signal and humidity signal of the air flowing out of the light scattering chamber (2) are collected in real time by using the temperature and humidity sensor (7) and sent to the data processing module 1 (10); S312: The data processing module 1 (10) obtains temperature data and humidity data based on the received temperature signal and humidity signal, obtains laser multi-angle information based on the received multi-angle scattered light signal, and sends the laser (3) light source parameters, laser multi-angle information, temperature data and humidity data to the comprehensive data storage area 1 in the database 1 for storage. At the same time, the dust concentration calculation unit calculates the dust concentration data CA1 based on the laser (3) light source parameters and laser multi-angle information, and in combination with the temperature data and humidity data of the gas to be measured; at the same time, in each monitoring cycle, the xth n In the calculation process, the obtained Xth n The dust concentration data CA1 is sent to the physical drive self-calibration unit, and the physical drive self-calibration unit converts the Xth n The dust concentration data CA1 is compared with the dust concentration information data set in the calibration data storage area 1 in the database 1, wherein the calibration data storage area 1 includes dust concentration information data sets calculated by comprehensive factors such as different light source parameters, different temperature and humidity information, and different laser multi-angle information. If the comparison deviation is less than the set threshold, no action is taken. If the comparison deviation is greater than the set threshold, the physical-driven prediction model is used to analyze the relationship between the light source parameters, laser multi-angle information, temperature data and humidity data and the output of the dust concentration calculation unit to predict the expected behavior characteristics of the dust concentration calculation unit. At the same time, the calculation factors that need to be modified are calculated by inversion calculation and fed back to the dust concentration calculation unit. Then, through repeated iteration and optimization, the difference between the output value of the dust concentration calculation unit and the predicted value of the physical-driven model is controlled within the set error range. The data processing module 1 (10) sends the location information, light source parameters, laser multi-angle information, temperature data, humidity data and dust concentration data CA1 of the low-cost dust monitor to the coal mine information processing center unit; S32: Monitor the dust at the location using an oscillating balance dust continuous measuring instrument; The location information of the oscillating balance dust continuous measuring instrument is collected by the positioning module 2, and sent to the data processing module 2; the dust concentration data CB at the location is obtained by the data processing module 2 in the oscillating balance dust continuous measuring instrument; the data processing module 2 sends the location information of the oscillating balance dust continuous measuring instrument and the dust concentration data CB to the coal mine information processing center; Step 4: Draw a dust distribution cloud map of the entire mine area; S41: After receiving the location information, light source parameters, laser multi-angle information, temperature data, humidity data and dust concentration data CA1 of the low-cost dust monitor, and the location information and dust concentration data CB of the oscillating balance dust continuous measuring instrument, the central data processor stores them in the comprehensive data storage area 2 in the database 2; S42: The central data processor cleans and standardizes the dust concentration data CA1 and the dust concentration data CB in the integrated data storage area 2 to eliminate the noise and outliers therein. In the process of eliminating outliers, for the dust concentration data from the same group of dust concentration monitoring components, the dust concentration data CB of the oscillating balance dust continuous measuring instrument is used as the reference data, and then the dust concentration data CA1 of the remaining multiple low-cost dust monitors are compared with the reference data. If the comparison difference is greater than the set difference range, it is confirmed to be an outlier and is removed; the screened high-quality and consistent data that meet the requirements are stored in the calibration data storage area 2 in the database 2, and used as a data set for building a machine learning calibration model; S43: The machine learning self-calibration unit divides the data set in the calibration data storage area 2 into a training set and a validation set according to a set ratio; uses light source parameters, laser multi-angle information, temperature data and humidity data as feature data, and adopts a support vector machine machine learning algorithm with higher prediction accuracy and generalization ability in dust concentration monitoring to build a machine learning calibration model; then uses the training set to train the machine learning calibration model to obtain a trained machine learning calibration model; then, minimizes the prediction error by adjusting the model parameters to obtain an optimized machine learning calibration model; finally, uses the validation set to verify the optimized machine learning calibration model, and further adjusts and optimizes the machine learning calibration model parameters according to the verification results to obtain a data-driven machine learning calibration model; S44: using the dust concentration data CA1 received in real time from the low-cost dust monitor as input data, inputting it into the machine learning calibration model, performing machine learning self-calibration on the input data through the machine learning calibration model, and outputting the calibrated dust concentration data CA; S45: The central data processor stores the location information and calibrated dust concentration data CA of the low-cost dust monitor, and the location information and dust concentration data CB of the oscillating balance dust continuous measuring instrument in the dust cloud data storage area in the second database; S46: The dust distribution cloud map unit of the whole mine area uses the spatial dust concentration prediction network model to draw the dust distribution cloud map of the whole mine area based on the location information of the low-cost dust monitor and the calibrated dust concentration data CA, as well as the location information and dust concentration data CB of the oscillating balance dust continuous measuring instrument, and sends the dust distribution cloud map of the whole mine area to the visual client; wherein, the spatial dust concentration prediction network model is established by means of a massive position combination experiment, and the specific process is as follows: S461: Use an oscillating balance dust continuous measuring instrument to monitor dust concentration at different locations, collect and record dust concentration data and corresponding location information through a large number of measurement processes, and use a large amount of dust concentration data and corresponding location information as a sample data set; S462: Divide the sample data set into a training set and a validation set according to a set ratio; S463: Establishing a spatial dust concentration prediction network model, and training it using a training set to obtain a trained spatial dust concentration prediction network model; S464: Adjust and optimize model parameters to improve the prediction accuracy of the spatial dust concentration prediction network model; S465: Use the validation set to validate the spatial dust concentration prediction network model. At the same time, the accuracy and reliability of the model are evaluated by comparing the dust concentration data predicted by the spatial dust concentration prediction network model with the actual observed values. The model parameters are further adjusted and optimized according to the evaluation results to obtain the final spatial dust concentration prediction network model. Step 5: The visual client displays the received dust distribution cloud map of the entire mine area in real time.

10. A low-cost dual-drive self-calibration dust concentration online monitoring method according to claim 9, characterized in that: In step three S312, the physically driven prediction model is established based on Monte Carlo simulation and Mie scattering theory.

Images