Mining dust monitoring system and method based on multi-wavelength fusion and dynamic compensation

Abstract

According to the mining dust monitoring system and method based on multi-wavelength fusion and dynamic compensation, a light chamber is arranged in an inner cavity of a shell, an inlet of the light chamber is connected with a main gas inlet of the shell through a gas inlet pipeline, and an outlet of the light chamber is connected with a main gas outlet of the shell through a gas exhaust pipeline; the laser detection module is mounted in the light chamber and comprises a lens group, a laser source I, a laser source II, a light scattering signal receiving module, a light transmission signal receiving module and a light trap; the first laser source and the second laser source form a dual-wavelength laser source; the micro fan is mounted in the main air outlet; the environment sensing module is mounted in the exhaust pipeline; the method comprises the following steps: constructing an environment drift prediction model; testing the dust concentration in the environment to be tested, and calculating the dust concentration I Csct and the dust concentration II Ct based on the scattered light electric signal and the transmission photoelectric signal; and carrying out environment drift prediction, and obtaining corrected dust concentration data through a fusion calculation mode. According to the system and the method, an accurate dust concentration monitoring result can be obtained.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent dust monitoring, and specifically relates to a mine dust monitoring system and method based on multi-wavelength fusion and dynamic compensation. Background Art

[0002] With the increase of mine depth and the complication of mining technology, the safety of the mine operation environment has been widely concerned. Among them, the mine dust concentration is an important parameter related to the health and safety of miners. The dust in the mine mainly comes from operations such as coal drilling, blasting, conveying, and processing. These dusts are suspended in the air for a long time and are easily inhaled by miners, which can easily lead to the occurrence of various occupational diseases, such as pneumoconiosis. Long-term exposure to a high dust concentration environment will cause irreversible damage to the respiratory system, visual system, and even the cardiovascular system of miners. In addition, under certain conditions, a mixture of high-concentration coal dust and air may also cause dust explosions in the mine, which will threaten the safety production work of the entire mine. Therefore, the real-time and accurate monitoring of the mine dust concentration can not only timely detect the over-standard phenomenon of dust, but also help to take corresponding treatment measures in critical moments, and thus can provide strong technical support and safety guarantee for the safety production work of the mine.

[0003] At present, the monitoring technologies for dust concentration at home and abroad mainly include the gravimetric method with filter membrane, β-ray method, light scattering method, charge induction method, and micro oscillating balance method, etc. These methods are applied in the fields of respirable dust monitoring, pneumoconiosis prevention, particulate matter emission control, etc., and have important significance for improving the dust monitoring level. However, there are still some deficiencies in the existing technologies for monitoring mine dust concentration, specifically as follows: Although the gravimetric method with filter membrane has relatively accurate monitoring data, its real-time performance is poor and the maintenance process is relatively complex; the β-ray method is limited by the ray penetration ability, and its measurement range and measurement accuracy are limited; the light scattering method is easily affected by the change of particle size distribution. When the median particle size of the dust deviates from the calibrated range, the measurement error will exceed ±20%, and it is easily contaminated and the maintenance process is frequent; the light transmission method has a sudden drop in sensitivity due to the saturation of light intensity attenuation in the high-concentration range (>500mg / m 3 );The charge induction method's accuracy is easily affected by factors such as dust particle size, shape, and concentration, and it is difficult to further improve the measurement accuracy; the micro oscillating balance method is not suitable for use in high-dust-concentration environments due to the small dust capacity of the filter membrane. To sum up, the existing dust concentration monitoring technologies have obvious deficiencies in particle size sensitivity, high-concentration measurement, environmental adaptability, and optical window cleaning. There is an urgent need for a new monitoring device and method to overcome these problems to improve the accuracy and reliability of mine dust concentration monitoring, and thus provide stronger safety guarantee for the safety production work in coal mines. Summary of the Invention

[0004] In view of the problems existing in the above-mentioned prior art, the present invention provides a mine dust monitoring system and method based on multi-wavelength fusion and dynamic compensation. The system has a simple structure and a high degree of intelligence. It can obtain accurate dust concentration monitoring results and provide more powerful safety guarantees for the safe production work in coal mines. The method has a high degree of intelligence and a convenient measurement process. It can accurately and efficiently obtain the dust concentration in the environment. Through the multi-wavelength fusion technology and the dynamic compensation mechanism, it can effectively improve the overall performance and effect of mine dust concentration monitoring.

[0005] To achieve the above object, the present invention provides a mine dust monitoring system based on multi-wavelength fusion and dynamic compensation, including a housing, an air intake filter, an optical chamber, a laser detection module, a micro fan, an environmental perception module and a control module;

[0006] A main air intake and a main air outlet are respectively provided at the top and bottom of the housing; the air intake filter is installed in the main air intake;

[0007] The optical chamber is arranged in the inner cavity of the housing. An optical chamber inlet and an optical chamber outlet are oppositely provided at the top and bottom thereof, and an optical detection channel is formed inside between the optical chamber inlet and the optical chamber outlet. At the same time, the optical chamber inlet is connected to the main air intake through an air inlet pipeline, and the optical chamber outlet is connected to the main air outlet through an exhaust pipeline;

[0008] The laser detection module is installed inside the optical chamber. It includes a lens group, a first laser source, a second laser source, a light scattering signal receiving module, a light transmission signal receiving module and a light trap. The lens group is fixedly supported on the left side of the right side wall of the optical chamber, and its central axis passes through the optical detection channel, and its focal point is located at the optical detection channel. The first laser source and the second laser source are symmetrically distributed up and down about the central axis of the lens group and are fixedly installed on the right side wall of the optical chamber to form a dual-wavelength laser source. The light scattering signal receiving module is located at the rear side of the optical detection channel and is fixedly installed on the rear side wall of the optical chamber, and the axis of its receiving surface is at a 90° angle to the central axis of the lens group. The light transmission signal receiving module is fixedly installed on the left side wall of the optical chamber, and the axis of its receiving surface coincides with the central axis of the lens group. The light trap is located at the front side of the optical detection channel and is installed on the front side wall of the optical chamber;

[0009] The micro fan is installed in the main air outlet; the environmental perception module is installed in the exhaust pipeline and is used for collecting the temperature signal, humidity signal and vibration acceleration signal of the dust-containing airflow passing through the exhaust pipeline;

[0010] The control module is respectively connected to the environmental perception module, the laser detection module and the micro fan.

[0011] Furthermore, in order to facilitate the cleaning treatment of the optical chamber, an ultrasonic cleaning module is further included, and the ultrasonic cleaning module is installed on the outer side wall surface of the optical chamber.

[0012] As a further preference, the control module includes a dual-wave time-division driving module, a data acquisition module, a central processing module, and a data storage module;

[0013] The dual-wave time-division driving module is respectively connected to the first laser source and the second laser source; the data acquisition module is respectively connected to the light scattering signal receiving module, the light transmission signal receiving module, and the environmental perception module; the central processing module is respectively connected to the dual-wave time-division driving module, the data storage module, the micro fan, and the ultrasonic cleaning module.

[0014] Further, to ensure the accuracy of the measurement results, the light chamber is made of aluminum alloy material, and black light-absorbing materials are coated on the inner and outer wall surfaces; an anti-adsorption coating is also coated on the inner wall surface of the light chamber.

[0015] As a preference, a power supply module is further included. A power supply interface is also provided on one side of the housing, and the power supply module is connected to the power supply interface.

[0016] Further, to avoid the influence of dust deposition on the transmittance and ensure the accuracy of the monitoring data, a fused quartz protection sheet is installed at the receiving surface position of the light scattering signal receiving module.

[0017] Further, to facilitate the real-time display of the measurement results of the dust concentration, a display screen is further included. The display screen is embedded on the outer surface of the housing and is connected to the central processing module.

[0018] In the present invention, by installing an intake filter in the main intake port, it is convenient to prevent large particle impurities from entering the interior of the housing, thereby improving the anti-pollution ability of the monitoring system. By installing a micro fan in the main outlet port, a continuous negative pressure can be provided for the detection process. Thus, the dusty air flow in the environment can continuously enter the light detection channel of the light chamber through the main intake port and the intake pipeline, and then be discharged to the environment through the exhaust pipeline and the main outlet port. By simultaneously setting laser sources one and two, under the condition that they emit different wavelengths, signal acquisition of scattered light signals and transmitted light signals can be respectively performed, which is conducive to decoupling particle size interference according to the dual-wavelength orthogonal scattering matrix, greatly improving the accuracy of the measurement results. By setting the lens group, the light beam emitted by the laser source can be focused on the light detection channel, which can effectively act on the dusty air flow passing through the light detection channel. In this way, a scattered light signal with better intensity can be obtained, and at the same time, a transmitted light signal with an obvious change in intensity can also be obtained. By setting an environmental perception module in the exhaust pipeline, it is convenient to collect temperature signals, humidity signals, and vibration acceleration signals in the dusty air flow in real time during the detection process. In this way, it is convenient for the control module to obtain temperature data, humidity data, and vibration acceleration data in the dusty air flow, and further, an environmental drift amount can be obtained based on these data. Finally, the measurement error caused by environmental changes can be compensated by using the environmental drift amount, which is conducive to ensuring the accuracy of the measurement results. By setting the control module, it is conducive to realizing full automation in the entire monitoring process and can realize automatic calibration of the measurement results.

[0019] The system has a simple structure and a high degree of intelligence. It can obtain accurate dust concentration monitoring results and provide more powerful safety guarantees for the safe production work in coal mines.

[0020] The present invention also provides a mine dust monitoring method based on multi-wavelength fusion and dynamic compensation. Using a mine dust monitoring system based on multi-wavelength fusion and dynamic compensation, it includes the following steps;

[0021] Step 1: Construct an environmental drift prediction model;

[0022] A1: Construct a sample data set and store it in the data storage module; at the same time, construct the scattering ratio R(d) corresponding to dust particles of different particle sizes and store it in the data storage module;

[0023] A2: Divide the sample data set into a training set and a test set according to a set ratio;

[0024] A3: Build a prediction model based on a long short-term memory network, train the prediction model using the training set, and at the same time, test it using the test set. Finally, an environmental drift prediction model is obtained;

[0025] Step 2: Arrange the mine dust monitoring system based on multi-wavelength fusion and dynamic compensation in the environment to be measured. Use the control module to control the start of the micro-fan to work continuously, providing negative pressure for the optical chamber. Under the action of the negative pressure, the dust-containing air flow in the environment passes through the intake filter in the main intake port and the intake pipeline into the optical detection channel of the optical chamber in sequence, and is discharged into the atmosphere through the exhaust pipeline and the main outlet.

[0026] Meanwhile, in each sub-cycle within a single sampling period, the dual-wavelength time-division driving module is used to control the laser source 1 and the laser source 2 to start working for a set time in sequence. During the working period of the laser source 1, the beam emitted by it is focused on the dust-containing air flow in the optical detection channel through the lens group, and acts on the dust particles to generate a scattered light signal. Synchronously, the optical scattering receiving module is used to collect the scattered light signal, convert it into a scattered photoelectric signal, and then send it to the data acquisition module. During the working period of the laser source 2, the beam emitted by it is focused on the dust-containing air flow in the optical detection channel through the lens group, and irradiates on the left end of the optical chamber through the dust-containing air flow. Synchronously, the optical transmission receiving module is used to collect the transmitted light signal, convert it into a transmitted photoelectric signal, and then send it to the data acquisition module.

[0027] Meanwhile, the environmental perception module is used to collect the temperature signal, humidity signal, and vibration acceleration signal of the dust-containing air flow in real time, and send them to the data acquisition module.

[0028] Meanwhile, the data acquisition module sends the received scattered photoelectric signal, transmitted photoelectric signal, temperature signal, humidity signal, and vibration acceleration signal to the central processing module.

[0029] Step 3: The central processing module calculates the dust concentration based on the received scattered photoelectric signal, transmitted photoelectric signal, temperature signal, humidity signal, and vibration acceleration signal.

[0030] B1: Calculate the dust concentration C1 under scattering conditions sct ;

[0031] B11: Use the Mie scattering matrix decomposition algorithm to obtain the calculation formula of the scattered light intensity distribution I(λ,θ) of the particulate matter, as shown in formula (1); sct In the formula, S1(θ) is the scattering amplitude function; d is the particle size; λ is the wavelength; C

[0032]

[0033] is the light scattering concentration; N is the number concentration of particulate matter; θ is the scattering angle, representing the angle between the incident light and the scattered light; r is the distance from the scatterer to the acquisition point; sca

[0034] ​B12: Based on Mie scattering theory, construct a particle size insensitive term to obtain the calculation formula for the scattering ratio R(d) of different particle sizes, as shown in formula (2);

[0035]

[0036] In the formula, λ1 is the wavelength of the laser beam emitted by laser source one, and λ2 is the wavelength of the laser beam emitted by laser source two;

[0037] B13: Calculate the dust concentration C under light scattering conditions in real time according to formula (3) sct ;

[0038]

[0039] In the formula, I sct (λ1) is the scattered light intensity under the condition of the laser beam emitted by laser source one, which is obtained by the central processing module according to the received scattered photoelectric signal. K1 is the calibration constant for converting the scattered light intensity into a concentration unit;

[0040] B2: Conduct the dust concentration C under transmission conditions t ;

[0041] B21: Based on Lambert-Beer's law, perform non-linear correction under high dust concentration conditions to obtain the calculation formula for the transmitted light intensity, as shown in formula (4);

[0042] I t = I0·e -βCL (4);

[0043] In the formula, I t is the transmitted light intensity after the laser beam emitted by laser source two irradiates the dust; I0 is the reference light intensity without dust; β is the extinction coefficient; C is the light transmission concentration; L is the path length of the light passing through the dust stream;

[0044] B22: Perform a second-order Taylor expansion approximation on formula (4) to obtain formula (5);

[0045]

[0046] B23: Calculate the dust concentration C under light transmission conditions in real time according to formula (6) t ;

[0047]

[0048] In the formula, I t (λ2) is the transmitted light intensity under the condition of the laser beam emitted by laser source two, which is obtained by the central processing module according to the received transmitted photoelectric signal; K2 is the calibration constant for converting the transmitted light intensity into a concentration unit;

[0049] B3: Calculate the corrected dust concentration data through drift compensation and dynamic fusion;

[0050] B31: The central processing module obtains temperature data, humidity data, and vibration acceleration data based on the received temperature signal, humidity signal, and vibration acceleration signal, and uses the temperature data, humidity data, and vibration acceleration data as input data, inputting them into the environmental drift prediction model for prediction, and outputting the environmental drift amount ΔC env ;

[0051] B32: Obtain the dynamic weight distribution function according to formula (7);

[0052]

[0053] Where C prev is the concentration value of the previous iteration, and the initial concentration value is 0;

[0054] B33: Calculate the corrected dust concentration data by iteration according to formula (8);

[0055] C current = w(C prev )·C sct +[1 - w(C prev )]·C t +ΔC env (8).

[0056] As an optimization, in B3 of step three, during the operation of the mine dust monitoring system based on multi-wavelength fusion and dynamic compensation, the environmental drift prediction model is continuously updated in an online manner, fine-tuning the network weights in an incremental learning manner, and dynamically adjusting the learning rate according to historical errors. When the output variance exceeds 5 mg / m 3 , an abnormal alarm is triggered, and it rolls back to the previous stable state and saves this stable state.

[0057] As an optimization, in step one, several sample data sets are constructed in the following manner:

[0058] The mining dust monitoring system based on multi-wavelength fusion and dynamic compensation is arranged in an experimental environment. Under a large number of combined environmental conditions of different temperatures, different humidity, and different vibration accelerations, an airflow with a known dust concentration containing PTFE standard particles and coal mine dust particles in the range of 0.1-10μm is used as a sample airflow. Laser source one and laser source two are used to emit laser beams to irradiate the sample airflow, respectively, to obtain corresponding scattering intensity and transmission intensity, and then the temperature, humidity, vibration acceleration, particle size, scattering intensity, transmission intensity and dust concentration information are correspondingly formed into a group of sample data, and several groups of sample data are collected as sample data sets.

[0059] The present invention discloses a mining dust concentration monitoring method based on multi-wavelength fusion and dynamic compensation, which effectively improves the problems of particle size sensitivity, high concentration nonlinearity and environmental interference coupling of traditional optical methods. By simultaneously setting laser source one and laser source two to form a dual-wavelength laser source, and collecting scattered signals and transmission signals respectively, a dual-wavelength orthogonal scattering matrix can be effectively constructed to decouple particle size interference, and then a particle size insensitive measurement model can be established through the collaborative design of dual wavelengths. A large number of experiments have proved that the orthogonal matrix constructed by the two-wavelength scattering signal ratio can eliminate 90% of the particle size correlation, which is conducive to obtaining more accurate monitoring results. When calculating the final dust concentration data, after obtaining the dust concentration under scattering conditions, the transmission logarithmic transformation and dynamic fusion function are established, which can effectively design a transmission channel adaptive compensation mechanism for the high-concentration scattering signal saturation problem, and combine the Sigmoid weight function to achieve smooth switching of scattering-transmission mode. Based on the long short-term memory network, an environmental drift prediction model is constructed, which can easily capture the long-term dependence of temperature, humidity and vibration acceleration data, and then it is beneficial to collect the temperature, humidity and vibration acceleration data of the dust-laden airflow in real time in the future. After the temperature, humidity and vibration acceleration data are input into the environmental drift prediction model, it can effectively learn the nonlinear drift law, greatly reduce the temperature and humidity drift, and still maintain high stability under vibration conditions, and finally quickly and accurately predict the environmental drift amount, and then obtain a monitoring result with higher accuracy than the traditional compensation method by drift compensation. The present invention uses scattering and transmission to measure dust concentration through the effective cooperation of dual-wavelength laser source and scattering receiving and transmission receiving modules, and combines the environmental drift prediction model established based on LSTM neural network to dynamically compensate for the measurement error caused by environmental changes, which can achieve high-precision and wide-range measurement of dust concentration, has strong environmental adaptability, and can be applied to dust monitoring conditions in complex environments such as coal mines.

[0060] This method has a high degree of intelligence and a convenient measurement process. It can accurately and efficiently obtain dust concentration data in the environment. Through multi-wavelength fusion technology and dynamic compensation mechanism, it effectively solves the defects of the existing technology and improves the overall performance and effect of the mine dust concentration monitoring technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 is a schematic structural diagram of the present invention;

[0062] Figure 2 is an assembly schematic diagram of the laser detection module in the present invention;

[0063] Figure 3 is a flowchart of the calculation of dust concentration data in the present invention;

[0064] Figure 4 is a principle block diagram of the control part in the present invention.

[0065] In the figure: 1, housing; 2, optical chamber; S3, laser detection module; 4, ultrasonic cleaning module; 5, micro fan; 6, intake air filter; 7, environmental perception module; 8, control module; 9, dual-wave time-division drive module; 10, data acquisition module; 11, central processing module; 12, data storage module; 13, power supply module; 3.1, laser source 1; 3.2, laser source 2; 3.3, lens group; 3.4, light scattering signal receiving module; 3.5, light transmission signal receiving module; 3.6, light trap; 14, intake air pipeline; 15, exhaust air pipeline. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0066] The present invention will be further described below with reference to the accompanying drawings.

[0067] As Figures 1 to 4 shown, the present invention provides a mine dust monitoring system based on multi-wavelength fusion and dynamic compensation, including a housing 1, an intake air filter 6, an optical chamber 2, a laser detection module S3, a micro fan 5, an environmental perception module 7, and a control module 8;

[0068] A main intake port and a main exhaust port are respectively opened at the top and bottom of the housing 1. The main intake port and the main exhaust port of the housing 1 are both used to communicate with the test environment; the intake air filter 6 is installed in the main intake port, and the intake air filter 6 is used to prevent large particle impurities from entering the interior of the housing 1 to improve the anti-pollution ability of the monitoring system;

[0069] The optical chamber 2 is arranged in the inner cavity of the housing 1. A light chamber inlet and a light chamber outlet are oppositely opened at the top and bottom thereof, and a light detection channel is formed inside between the light chamber inlet and the light chamber outlet. At the same time, the light chamber inlet is connected to the main intake port through the intake air pipeline 14, and the light chamber outlet is connected to the main exhaust port through the exhaust air pipeline 15;

[0070] The laser detection module S3 is installed inside the optical chamber 2, and includes a lens group 3.3, a laser source 1 3.1, a laser source 2 3.2, a light scattering signal receiving module 3.4, a light transmission signal receiving module 3.5 and a light trap 3.6; the lens group 3.3 is fixedly supported on the left side of the right side wall of the optical chamber 2, with its axis passing through the light detection channel and its focus located at the light detection channel; the laser source 1 3.1 and the laser source 2 3.2 are symmetrically distributed about the central axis of the lens group 3.3, and are ... The light scattering signal receiving module 3.4 is located at the rear side of the light detection channel and is fixedly installed on the rear side wall of the light chamber 2, and the axis of its receiving surface is 90° with the central axis of the lens group 3.3; the light transmission signal receiving module 3.5 is fixedly installed on the left side wall of the light chamber 2, and the axis of its receiving surface coincides with the central axis of the lens group 3.3; the light trap 3.6 is located at the front side of the light detection channel and is installed on the front side wall of the light chamber 2;

[0071] Among them, the light scattering signal receiving module 3.4 is used to collect the scattered light signal generated by the laser irradiating the dust particles and convert it into a scattered photoelectric signal; the light transmission signal receiving module 3.5 is used to collect the transmitted light signal generated by the laser passing through the dust particles and convert it into a transmitted photoelectric signal; the light trap 3.6 is used to avoid cross-interference between different photodiodes in the same measurement light chamber to improve the signal-to-noise ratio;

[0072] Among them, the wavelengths of the laser beams emitted by the laser source 1 3.1 and the laser source 2 3.2 are different. As a preferred embodiment, the wavelength of the laser beam emitted by the laser source 1 3.1 is shorter than the wavelength of the laser beam emitted by the laser source 2 3.2, the laser beam emitted by the laser source 1 3.1 is in the Mie scattering sensitive area, the laser beam emitted by the laser source 2 3.2 is less affected by dust scattering, has a strong transmission ability, and can still maintain a linear response at high dust concentrations, and the scattering response matrix number of the laser beams emitted by the laser source 1 3.1 and the laser source 2 3.2 is low; therefore, the light beam emitted by the laser source 1 3.1 irradiates the dust-laden airflow passing through the light detection channel, forming a light scattering phenomenon, and the light beam emitted by the laser source 2 3.2 irradiates the dust-laden airflow passing through the light detection channel, forming a light transmission phenomenon;

[0073] The micro fan 5 is installed in the main air outlet; the environment sensing module 7 is installed in the exhaust pipe. Preferably, the environment sensing module 7 is provided with a temperature sensor, a humidity sensor and a vibration sensor for collecting the temperature signal, humidity signal and vibration acceleration signal of the dust-laden airflow passing through the exhaust pipe 15;

[0074] The control module 8 is connected to the environment sensing module 7 , the laser detection module S3 and the micro fan 5 respectively.

[0075] To facilitate the cleaning operation of the optical chamber 2, an ultrasonic cleaning module 4 is further included, and the ultrasonic cleaning module 4 is installed on the outer wall surface of the optical chamber 2. As a preference, a piezoelectric ceramic ultrasonic vibrator is configured inside the ultrasonic cleaning module 4, which can trigger pulsed cleaning when the pollution in the optical chamber 2 exceeds the standard. The inner wall of the optical chamber 2 is cleaned using ultrasonic pulses of 1.7 MHz. The duty cycle of the pulse is 50%, and the duration is 2 seconds. By promptly removing the dust deposition on the optical window, the accuracy of the measurement signal can be effectively ensured. As a further preference, the ultrasonic cleaning module 4 is installed on the central axis of the optical chamber 2 to effectively improve the ultrasonic cleaning effect and efficiency;

[0076] The control module 8 includes a dual-wave time-sharing driving module 9, a data acquisition module 10, a central processing module 11, and a data storage module 12;

[0077] The dual-wave time-sharing driving module 9 is respectively connected to the laser source 1 3.1 and the laser source 2 3.2; Preferably, the dual-wave time-sharing driving module 9 is used to control the laser source 1 3.1 to work for 10 ms in a pulsed driving manner in a sub-cycle, then turn off the laser source 1 3.1, and then control the laser source 2 3.2 to work for 10 ms in a pulsed driving manner, and then turn off the laser source 2 3.2. That is, the laser source 2 3.2 is delayed by 10 ms in each sub-cycle compared to the laser source 2 3.2. By the time-sharing working method, crosstalk between the lasers emitted by the two can be avoided; Preferably, when it lasts for 480 - 500 ms, the dual light sources (laser source 1 3.1 and laser source 2 3.2) are turned off, and the total sampling period is preferably 500 ms;

[0078] The data acquisition module 10 is respectively connected to the light scattering signal receiving module 3.4, the light transmission signal receiving module 3.5, and the environmental perception module 7; The central processing module 11 is respectively connected to the dual-wave time-sharing driving module 9, the data acquisition module 10, the data storage module 12, the micro fan 5, and the ultrasonic cleaning module 4. The data acquisition module 10 is used to process the light scattering electrical signal and the light transmission electrical signal emitted from the light scattering signal receiving module 3.4 and the light transmission signal receiving module 3.5, and then send the processed light scattering electrical signal and light transmission electrical signal to the central processing module 11. The central processing module 11 can also evaluate the pollution degree by real-time monitoring of the background noise of the light transmission channel during 480 - 500 ms in each total sampling period. When the transmission background noise > 50 mV, it is determined that the pollution level ≥ 3. At this time, a cleaning instruction is generated to trigger the ultrasonic cleaning module 4 to start working to facilitate the cleaning operation of the optical chamber 2;

[0079] As a preference, an environmental drift prediction model is built in the central processing module 11. It can predict the impact of environmental changes on the measurement results, compensate for environmental drift according to the changes in temperature, humidity and vibration acceleration during the test process, dynamically adjust the calculation result of the dust concentration, effectively eliminate the drift caused by environmental changes, and improve the accuracy of measurement data. As a preference, the environmental drift prediction model uses the time series data of temperature, humidity and vibration acceleration within a 60-second window as input data. After being processed by a 2-layer LSTM network (32 units in each layer) and a Dropout layer (dropout rate of 0.2), the environmental drift amount ΔC is output env , so as to effectively compensate for the measurement error caused by environmental changes.

[0080] As a preference, the data storage module 12 stores several sets of sample data sets obtained through a large number of experiments under a large number of combined environmental conditions of PTFE standard particles and coal mine site dust particles in the range of 0.1 - 10μm at different temperatures, different humidities and different vibration accelerations. Each set of sample data sets includes information such as temperature, humidity, vibration acceleration, particle size, scattering intensity, transmission intensity and dust concentration. Thus, it is convenient to realize the inversion calculation of dust concentration data. At the same time, the data storage module 12 also stores the scattering ratio R(d) corresponding to dust particles of different particle sizes;

[0081] As a further preference, the control module 8 further includes a communication module. The communication module is connected to the central processing module 11 and is used to establish a communication connection between the central processing module 11 and external devices;

[0082] To ensure the accuracy of the measurement results, the optical chamber 2 is made of aluminum alloy material, and black light-absorbing materials are coated on the inner and outer wall surfaces. An anti-adsorption coating is also coated on the inner side wall surface of the optical chamber 2, which can effectively prevent dust particles from adsorbing on the inner wall of the optical chamber 2, and then have an adverse impact on the optical detection unit, affecting the accuracy of the monitoring data.

[0083] As a preference, a power supply module 13 is further included. A power supply interface is also provided on one side of the housing 1, and the power supply module 13 is connected to the power supply interface. The power supply module 13 is connected to an external power supply device through the power supply interface and is also connected to the laser source 1 3.1, the laser source 2 3.2, the light scattering signal receiving module 3.4, the light transmission signal receiving module 3.5, the ultrasonic cleaning module 4, the micro fan 5, the environmental perception module 7 and the control module 8 respectively to supply power to each electrical device.

[0084] To avoid the influence of dust deposition on the transmittance and ensure the accuracy of monitoring data, a fused silica protection sheet is installed at the receiving surface position of the light scattering signal receiving module 3.4.

[0085] To facilitate the real-time display of the measurement results of the dust concentration, a display screen is further included. The display screen is embedded in the outer surface of the housing 1 and is connected to the central processing module 11.

[0086] In the present invention, by installing an intake air filter in the main intake port, it is convenient to prevent large particle impurities from entering the interior of the housing, thereby improving the anti-pollution ability of the monitoring system. By installing a micro fan in the main exhaust port, a continuous negative pressure can be provided for the detection process. Thus, the dust-containing air flow in the environment can continuously enter the light detection channel of the light chamber through the main intake port and the intake pipeline, and then be discharged to the environment through the exhaust pipeline and the main exhaust port. By simultaneously setting laser sources one and two, under the condition that they emit different wavelengths, signal acquisition of scattered light signals and transmitted light signals can be carried out respectively, which is beneficial to decouple the particle size interference according to the dual-wavelength orthogonal scattering matrix, greatly improving the accuracy of the measurement results. Through the setting of the lens group, the light beam emitted by the laser source can be focused on the light detection channel, and then effectively act on the dust-containing air flow passing through the light detection channel. In this way, a scattered light signal with better intensity can be obtained, and at the same time, a transmitted light signal with an obvious change in intensity can also be obtained. By setting an environmental perception module in the exhaust pipeline, it is convenient to collect the temperature signal, humidity signal and vibration acceleration signal in the dust-containing air flow in real time during the detection process. In this way, it is convenient for the control module to obtain the temperature data, humidity data and vibration acceleration data in the dust-containing air flow, and further, the environmental drift amount can be obtained based on these data. Finally, the measurement error caused by environmental changes can be compensated by using the environmental drift amount, which is beneficial to ensuring the accuracy of the measurement results. Through the setting of the control module, it is beneficial to realize the full automation in the whole monitoring process and achieve the automatic calibration of the measurement results.

[0087] The system has a simple structure and a high degree of intelligence. It can obtain accurate dust concentration monitoring results and provide more powerful safety guarantees for the safe production work in coal mines.

[0088] The present invention also provides a mine dust monitoring method based on multi-wavelength fusion and dynamic compensation, using a mine dust monitoring system based on multi-wavelength fusion and dynamic compensation, including the following steps;

[0089] Step 1: Construct an environmental drift prediction model;

[0090] A1: Construct a sample data set and store it in the data storage module 12; at the same time, construct the scattering ratio R(d) corresponding to dust particles of different particle sizes and store it in the data storage module 12;

[0091] A2: Divide the sample data set into a training set and a test set according to a set ratio;

[0092] A3: Build a prediction model based on the long short - term memory network (LSTM), and use the training set to train the prediction model. At the same time, use the test set for testing, and finally obtain the environmental drift prediction model;

[0093] Step 2: Arrange the mine dust monitoring system based on multi - wavelength fusion and dynamic compensation in the environment to be measured. Use the control module 8 to control the micro - fan 5 to start working, continuously provide negative pressure for the optical chamber 2. Due to the effect of negative pressure, the dust - containing air flow in the environment passes through the intake filter screen 6 in the main intake port and the intake pipeline 14 in sequence and enters the optical detection channel of the optical chamber 2, and is discharged to the atmosphere through the exhaust pipeline 15 and the main exhaust port;

[0094] At the same time, in each sub - cycle within a single sampling period, the dual - wavelength time - division driving module 9 controls the laser source 1 3.1 and the laser source 2 3.2 to start working for a set time in sequence. During the working period of the laser source 1 3.1, the beam emitted by it passes through the lens group 3.3 and is focused on the dust - containing air flow in the optical detection channel, and acts on the dust particles to generate a scattered light signal. At the same time, use the light scattering receiving module 3.4 to collect the scattered light signal, convert it into a scattered photoelectric signal, and then send it to the data acquisition module 10; During the working period of the laser source 2 3.2, the beam emitted by it passes through the lens group 3.3 and is focused on the dust - containing air flow in the optical detection channel, and passes through the dust - containing air flow and irradiates on the left end of the optical chamber 2. At the same time, use the light transmission receiving module 3.5 to collect the transmitted light signal, convert it into a transmitted photoelectric signal, and then send it to the data acquisition module 10;

[0095] At the same time, use the environmental perception module 7 to collect the temperature signal, humidity signal and vibration acceleration signal of the dust - containing air flow in real - time, and send them to the data acquisition module 10;

[0096] At the same time, the data acquisition module 10 sends the received scattered photoelectric signal, transmitted photoelectric signal, temperature signal, humidity signal and vibration acceleration signal to the central processing module 11;

[0097] Step 3: The central processing module 11 calculates the dust concentration based on the received scattered photoelectric signal, transmitted photoelectric signal, temperature signal, humidity signal and vibration acceleration signal;

[0098] B1: Calculate the dust concentration 1 C under scattering conditions sct ;

[0099] B11: Adopt the Mie scattering matrix decomposition algorithm to obtain the scattered light intensity distribution I of the particulate matter sct(λ,θ) calculation formula to achieve the measurement of particle size distribution invariance, as shown in formula (1);

[0100]

[0101] In the formula, S1(θ) is the scattering amplitude function; d is the particle size; λ is the wavelength; C sca is the light scattering concentration; N is the number concentration of particulate matter; θ is the scattering angle, representing the angle between the incident light and the scattered light; r is the distance from the scatterer to the collection point;

[0102] B12: Based on the Mie scattering theory, construct a particle size insensitive term to obtain the calculation formula of the scattering ratio R(d) for different particle sizes, as shown in formula (2);

[0103]

[0104] In the formula, λ1 is the wavelength of the laser beam emitted by laser source 3.1, and λ2 is the wavelength of the laser beam emitted by laser source 3.2;

[0105] B13: Calculate the dust concentration C under light scattering conditions in real time according to formula (3) sct ;

[0106]

[0107] In the formula, I sct (λ1) is the scattered light intensity under the condition of the laser beam emitted by laser source 3.1, obtained by the central processing module 11 according to the received scattered light electrical signal, and K1 is the calibration constant for converting the scattered light intensity into a concentration unit. Preferably, it can be obtained by calibrating with PTFE standard particles;

[0108] B2: Perform the dust concentration C under transmission conditions t ;

[0109] B21: Since the light transmission method follows the Lambert-Beer law, based on the Lambert-Beer law, a non-linear correction is performed under high dust concentration conditions to obtain the calculation formula of the transmitted light intensity, as shown in formula (4);

[0110] I t = I0·e -βCL (4);

[0111] In the formula, I t is the transmitted light intensity after the laser beam emitted by laser source 3.2 irradiates the dust; I0 is the reference light intensity without dust; β is the extinction coefficient, which is related to the dust particle size, shape, etc.; C is the light transmission concentration; L is the path length of the light passing through the dust stream;

[0112] B22: The logarithmic transformation is adopted to overcome the high-concentration non-linearity, normalize the measured transmitted light intensity, and solve the problem of transmitted light intensity saturation at high concentrations. Specifically, the second-order Taylor expansion approximation is performed on formula (4) to obtain formula (5);

[0113]

[0114] B23: Calculate the dust concentration C under the light transmission condition in real time according to formula (6) t ;

[0115]

[0116] wherein, I t (λ2) is the transmitted light intensity under the laser beam emitted by the laser source 3.2, which is obtained by the central processing module 11 according to the received transmitted photoelectric signal; K2 is the calibration constant for converting the transmitted light intensity into a concentration unit. Preferably, it can be obtained by calibrating with PTFE standard particles;

[0117] B3: Calculate the corrected dust concentration data through drift compensation and dynamic fusion;

[0118] B31: The central processing module 11 obtains temperature data, humidity data, and vibration acceleration data according to the received temperature signal, humidity signal, and vibration acceleration signal, and uses the temperature data, humidity data, and vibration acceleration data as input data and inputs them into the environmental drift prediction model. The environmental drift prediction model is used for prediction, and the environmental drift amount ΔC is output env ;

[0119] B32: Adaptively adjust the weights of the scattering and transmission components according to the concentration range. Specifically, obtain the dynamic weight distribution function according to formula (7); this function is a Sigmoid function. When the concentration C is higher than 400 mg / m 3 , the weight w(C) approaches 1, otherwise it approaches 0, which is used to smoothly switch between the two measurement methods of scattering and transmission.

[0120]

[0121] wherein, C prev is the concentration value of the previous iteration, and the initial concentration value is 0;

[0122] B33: Calculate the corrected dust concentration data through iteration according to formula (8). As a preference, repeat the iteration 2 - 3 times until convergence. When C current < 0, take C current = max(C sct , 0). When C current > 1000 mg / m3 When taking w = 0, it completely depends on the transmitted signal;

[0123] C current = w(C prev )·C sct + [1 - w(C prev )]·C t + ΔC env (8).

[0124] As a preference, in B3 of step three, during the operation of the mine dust monitoring system based on multi - wavelength fusion and dynamic compensation, the environmental drift prediction model is continuously updated in an online manner, fine - tunes the network weights in an incremental learning manner, and dynamically adjusts the learning rate according to historical errors. When the output variance exceeds 5 mg / m 3 it triggers an abnormal alarm, rolls back to the previous stable state, and saves this stable state, so that the central processing module 11 can dynamically adjust the measurement result of the dust concentration to effectively eliminate the drift caused by environmental changes.

[0125] As a preference, in step one, several sample data sets are constructed in the following way:

[0126] Arrange the mine dust monitoring system based on multi - wavelength fusion and dynamic compensation in an experimental environment. Under a large number of combined environmental conditions of different temperatures, different humidities, and different vibration accelerations, use a known dust - concentration air flow containing PTFE standard particles and coal - mine on - site dust particles in the range of 0.1 - 10 μm as the sample air flow. Respectively use laser source 1 3.1 and laser source 2 3.2 to emit laser beams to irradiate the sample air flow, obtain the corresponding scattering intensity and transmission intensity, and then form a set of sample data corresponding to temperature, humidity, vibration acceleration, particle size, scattering intensity, transmission intensity, and dust concentration information. Collect several groups of sample data as the sample data set.

[0127] The present invention discloses a method for monitoring the concentration of mine dust based on multi-wavelength fusion and dynamic compensation, which effectively improves the problems of particle size sensitivity, high-concentration non-linearity, and environmental interference coupling in the traditional optical method. By simultaneously setting laser source one and laser source two to form a dual-wavelength laser source, and respectively collecting the scattering signal and the transmission signal, it can effectively construct a dual-wavelength orthogonal scattering matrix to decouple the particle size interference. Furthermore, a particle size-insensitive measurement model can be established through the collaborative design of the two wavelengths. Through a large number of experiments, it is proved that the orthogonal matrix constructed by the ratio of the scattering signals of the two wavelengths can eliminate 90% of the particle size correlation, which is conducive to obtaining more accurate monitoring results. When calculating the final dust concentration data, after obtaining the dust concentration under the scattering condition, a transmission logarithmic transformation and a dynamic fusion function are established, which can effectively design a transmission channel adaptive compensation mechanism for the problem of high-concentration scattering signal saturation, and realize the smooth switching of the scattering-transmission mode in combination with the Sigmoid weight function. An environmental drift prediction model is constructed based on the long short-term memory network, which is convenient for capturing the long-term dependence relationship of temperature, humidity, and vibration acceleration data. Furthermore, after the temperature, humidity, and vibration acceleration data of the dust-containing air flow are collected in real time subsequently, when the temperature, humidity, and vibration acceleration data are input into the environmental drift prediction model, it can effectively learn the non-linear drift law, greatly reduce the temperature and humidity drift, and still maintain high stability under vibration conditions. Finally, it can quickly and accurately predict the environmental drift amount, and then a monitoring result with higher accuracy than the traditional compensation method can be obtained through the drift compensation method. Through the effective cooperation of the dual-wavelength laser source, the scattering receiving module, and the transmission receiving module, the present invention uses scattering and transmission to measure the dust concentration, and combines the environmental drift prediction model established based on the LSTM neural network to dynamically compensate the measurement error caused by environmental changes, can realize the high-precision and wide-range measurement of the dust concentration, has strong environmental adaptability, and can be applied to the dust monitoring working conditions in complex environments such as coal mines.

[0128] This method has a high degree of intelligence and a convenient measurement process. It can accurately and efficiently obtain the dust concentration data in the environment. Through the multi-wavelength fusion technology and the dynamic compensation mechanism, it effectively solves the defects of the existing technology and improves the overall performance and effect of the mine dust concentration monitoring technology.

Claims

1. A mine dust monitoring system based on multi-wavelength fusion and dynamic compensation, comprising a housing (1) and an intake air filter screen (6). A main intake port and a main outlet port are respectively formed in the top and bottom of the housing (1); the intake air filter screen (6) is installed in the main intake port; and it is characterized in that, It further includes an optical chamber (2), a laser detection module (S3), a micro fan (5), an environmental perception module (7) and a control module (8); The optical chamber (2) is arranged in the inner cavity of the housing (1). An optical chamber inlet and an optical chamber outlet are oppositely formed at the top and bottom thereof, and an optical detection channel is formed inside between the optical chamber inlet and the optical chamber outlet. Meanwhile, the optical chamber inlet is connected to the main air inlet through an air inlet pipeline (14), and the optical chamber outlet is connected to the main air outlet through an exhaust pipeline (15); The laser detection module (S3) is installed inside the optical chamber (2), and includes a lens group (3.3), a first laser source (3.1), a second laser source (3.2), a light scattering signal receiving module (3.4), a light transmission signal receiving module (3.5) and a light trap (3.6); The lens group (3.3) is fixedly supported on the left side of the right side wall of the optical chamber (2), its central axis passes through the optical detection channel, and its focal point is located at the optical detection channel; The first laser source (3.1) and the second laser source (3.2) are symmetrically distributed up and down with respect to the central axis of the lens group (3.3), and are fixedly installed on the right side wall of the optical chamber (2) to form a dual-wavelength laser source; The light scattering signal receiving module (3.4) is located at the rear side of the optical detection channel and is fixedly installed on the rear side wall of the optical chamber (2), and the axis of its receiving surface is at a 90° angle to the central axis of the lens group (3.3); The light transmission signal receiving module (3.5) is fixedly installed on the left side wall of the optical chamber (2), and the axis line of its receiving surface coincides with the central axis of the lens group (3.3); The light trap (3.6) is located at the front side of the optical detection channel and is installed on the front side wall of the optical chamber (2); The micro fan (5) is installed in the main air outlet; The environmental perception module (7) is installed in the exhaust pipeline and is used for collecting the temperature signal, humidity signal and vibration acceleration signal of the dust-containing airflow passing through the exhaust pipeline (15); The control module (8) is respectively connected to the environmental perception module (7), the laser detection module (S3) and the micro fan (5).

2. The mine dust monitoring system based on multi-wavelength fusion and dynamic compensation according to claim 1, characterized in that, It further includes an ultrasonic cleaning module (4), and the ultrasonic cleaning module (4) is installed on the outer side wall surface of the optical chamber (2).

3. The mine dust monitoring system based on multi-wavelength fusion and dynamic compensation according to claim 2, characterized in that The control module (8) includes a dual-wave time-division driving module (9), a data acquisition module (10), a central processing module (11), and a data storage module (12); The dual-wave time-division driving module (9) is respectively connected to the first laser source (3.1) and the second laser source (3.2); The data acquisition module (10) is respectively connected to the light scattering signal receiving module (3.4), the light transmission signal receiving module (3.5), and the environmental perception module (7); The central processing module (11) is respectively connected to the dual-wave time-division driving module (9), the data acquisition module (10), the data storage module (12), the micro fan (5) and the ultrasonic cleaning module (4).

4. The mine dust monitoring system based on multi-wavelength fusion and dynamic compensation according to claim 3, wherein, The optical chamber (2) is made of aluminum alloy material, and black light-absorbing materials are coated on the inner and outer wall surfaces; An anti-adsorption coating is also coated on the inner side wall surface of the optical chamber (2).

5. The mine dust monitoring system based on multi-wavelength fusion and dynamic compensation according to claim 4, characterized in that, It further includes a power supply module (13). A power interface is also provided on one side of the housing (1), and the power supply module (13) is connected to the power interface.

6. The mine dust monitoring system based on multi-wavelength fusion and dynamic compensation according to claim 5, wherein, A fused silica protection sheet is installed at the position of the receiving surface of the light scattering signal receiving module (3.4).

7. The mine dust monitoring system based on multi-wavelength fusion and dynamic compensation according to claim 6, wherein It further includes a display screen, which is embedded in the outer surface of the housing (1) and connected to the central processing module (11).

8. A mine dust monitoring method based on multi-wavelength fusion and dynamic compensation, which uses a mine dust monitoring system based on multi-wavelength fusion and dynamic compensation as described in any one of claims 1 to 7, characterized in that, It includes the following steps; Step 1: Construct an environmental drift prediction model; A1: Construct a sample data set and store it in the data storage module (12); meanwhile, construct the scattering ratio R(d) corresponding to dust particles of different particle sizes and store it in the data storage module (12); A2: Divide the sample data set into a training set and a test set according to a set ratio; A3: Build a prediction model based on a long short-term memory network, train the prediction model using the training set, and at the same time, test it using the test set to finally obtain an environmental drift prediction model; Step 2: Arrange the mine dust monitoring system based on multi-wavelength fusion and dynamic compensation in the environment to be measured. Use the control module (8) to control the micro fan (5) to start working, continuously provide negative pressure for the optical chamber (2), and use the effect of negative pressure to make the dust-containing air flow in the environment pass through the intake filter screen (6) in the main intake port and the intake pipeline (14) in sequence and enter the optical detection channel of the optical chamber (2), and be discharged to the atmosphere through the exhaust pipeline (15) and the main exhaust port; Meanwhile, in each sub-cycle within a single sampling period, the dual-wavelength time-division driving module (9) sequentially controls the laser source one (3.1) and the laser source two (3.2) to start working for a set time. During the working period of the laser source one (3.1), the beam emitted by it passes through the lens group (3.3) and is focused on the dust-containing air flow in the optical detection channel, and acts on the dust particles to generate a scattered light signal. Synchronously, the light scattering receiving module (3.4) is used to collect the scattered light signal, convert it into a scattered photoelectric signal, and then send it to the data acquisition module (10); during the working period of the laser source two (3.2), the beam emitted by it passes through the lens group (3.3) and is focused on the dust-containing air flow in the optical detection channel, and passes through the dust-containing air flow and irradiates on the left end of the optical chamber (2). Synchronously, the light transmission receiving module (3.5) is used to collect the transmitted light signal, convert it into a transmitted photoelectric signal, and then send it to the data acquisition module (10); Meanwhile, the environmental perception module (7) is used to collect the temperature signal, humidity signal, and vibration acceleration signal of the dust-containing air flow in real time and send them to the data acquisition module (10); Meanwhile, the data acquisition module (10) sends the received scattered photoelectric signal, transmitted photoelectric signal, temperature signal, humidity signal, and vibration acceleration signal to the central processing module (11); Step 3: The central processing module (11) calculates the dust concentration based on the received scattered photoelectric signal, transmitted photoelectric signal, temperature signal, humidity signal, and vibration acceleration signal; B1: Conduct dust concentration C under scattering conditions sct ; B11: Using the Mie scattering matrix decomposition algorithm, the scattering light intensity distribution I sct (λ,θ) calculation formula is shown in formula (1); In the formula, S1(θ) is the scattering amplitude function; d is the particle size; λ is the wavelength; C sca is the light scattering concentration; N is the particle number concentration; θ is the scattering angle, representing the angle between the incident light and the scattered light; r is the distance from the scatterer to the collection point; B12: Construct a particle size-insensitive term based on the Mie scattering theory to obtain the calculation formula for the scattering ratio R(d) of different particle sizes, as shown in formula (2); Where, λ1 is the wavelength of the laser beam emitted by laser source 1 (3.1), and λ2 is the wavelength of the laser beam emitted by laser source 2 (3.2); B13: Calculate the dust concentration C under light scattering conditions in real time according to formula (3). sct ; Where, I sct (λ1) is the scattered light intensity under the condition of the laser beam emitted by the first laser source (3.1), which is obtained by the central processing module (11) according to the received scattered photoelectric signal, and K1 is a calibration constant for converting the scattered light intensity into a concentration unit; B2: Conduct the dust concentration two C under the transmission condition t ; B21: Based on the Lambert-Beer law, a calculation formula for the transmitted light intensity obtained by non-linear correction under high dust concentration conditions is as shown in formula (4); I t = I0·e -βCL (4); Where, I t is the transmitted light intensity after the laser beam emitted by the second laser source (3.2) irradiates the dust; I0 is the reference light intensity without dust; β is the extinction coefficient; C is the light transmission concentration; L is the path length of the light passing through the dust stream; B22: Perform a second-order Taylor expansion approximation on formula (4) to obtain formula (5); B23: Calculate the dust concentration C under the light transmission condition in real time according to formula (6). t ; Where, I t (λ2) is the transmitted light intensity under the condition of the laser beam emitted by the second laser source (3.2), which is obtained by the central processing module (11) according to the received transmitted photoelectric signal; K2 is the calibration constant for converting the transmitted light intensity into a concentration unit; B3: Calculate the corrected dust concentration data through drift compensation and dynamic fusion; B31: The central processing module (11) obtains temperature data, humidity data, and vibration acceleration data based on the received temperature signal, humidity signal, and vibration acceleration signal, and uses the temperature data, humidity data, and vibration acceleration data as input data to input into the environmental drift prediction model. The environmental drift prediction model is used for prediction, and the environmental drift amount ΔC is output env ; B32: Obtain the dynamic weight distribution function according to formula (7); where C prev is the concentration value of the previous iteration, and the initial concentration value is 0; B33: Calculate the corrected dust concentration data by iteration according to formula (8); C current = w(C prev )·C sct + [1 - w(C prev )]·C t + ΔC env (8).

9. The method for monitoring mine dust based on multi-wavelength fusion and dynamic compensation according to claim 8, wherein In B3 of Step 3, during the operation of the mine dust monitoring system based on multi-wavelength fusion and dynamic compensation, the environmental drift prediction model is continuously updated in an online manner, fine-tuning the network weights in an incremental learning manner, and dynamically adjusting the learning rate according to historical errors. When the output variance exceeds 5 mg / m 3 it triggers an abnormal alarm, rolls back to the previous stable state, and saves this stable state.

10. A mine dust monitoring method based on multi-wavelength fusion and dynamic compensation according to claim 8, characterized in that, In step one, several sample data sets are constructed in the following manner: Arrange the mine dust monitoring system based on multi-wavelength fusion and dynamic compensation in the experimental environment. Under a large number of combined environmental conditions of different temperatures, different humidities, and different vibration accelerations, use a known dust concentration gas flow containing PTFE standard particles and coal mine on-site dust particles in the range of 0.1 - 10 μm as the sample gas flow. Use the laser beams emitted by laser source 1 (3.1) and laser source 2 (3.2) to irradiate the sample gas flow respectively to obtain the corresponding scattering intensity and transmitted intensity. Then, form a set of sample data by corresponding the temperature, humidity, vibration acceleration, particle size, scattering intensity, transmitted intensity, and dust concentration information. Collect several groups of sample data as the sample data set.

Images