On-line detection device and method for ash content of coal slime flotation substance

Abstract

The invention provides an online detection device and method for ash content of a coal slime flotation object, and belongs to the technical field of ash content detection of the coal slime flotation object. The technical problem to be solved is to provide the hardware structure of the coal slime flotation material ash content online detection device and the improvement of the detection method. According to the technical scheme, a pulse laser is arranged in a device shell, the output end of the pulse laser right faces a quartz convex lens, and the quartz convex lens is movably installed on the bottom face of the device shell through a lens support; a conveying belt is arranged on the bottom face of the device shell in a penetrating mode and can input coal slime mixture samples conveyed by the flotation machine in a bypass mode into the device shell. The pulse laser and the quartz convex lens are arranged on one side of the conveying belt, a photoelectric detector array is further arranged on the other side of the conveying belt, and the signal input end of the photoelectric detector array directly faces the pulse laser; the method is applied to coal slime flotation material ash content detection.

Description

Technical Field

[0001] The invention provides an online detection device and method for ash content of coal slime flotation material, belonging to the technical field of ash content detection of coal slime flotation material. Background Art

[0002] Flotation coal preparation is currently a major coal preparation method. It utilizes the difference in the surface chemical properties of coal and gangue particles. The surface of coal is hydrophobic and the mineral impurities are water-absorbent, which causes the coal to adhere to the bubbles dispersed in the water and float to the water surface. After being scraped out by a scraper, it is dehydrated to obtain clean coal. The mineral impurities absorb water and sink to be discharged as tailings. The non-combustible inorganic matter remaining after the coal is completely burned is called coal ash, and the coal ash content is an important indicator to measure the inorganic content in coal. The larger the ash content, the more inorganic content in the coal and the worse the quality of the coal. How to more effectively reduce the coal ash content to improve the combustion quality of coal is an important issue that needs to be solved in the production and utilization of coal.

[0003] For the detection of coal ash content, the currently used burning detection method has a long detection cycle and low efficiency. There is also an ash detection method based on image analysis and processing, but this method generally only considers the mapping relationship between the pixel value of the coal image and the ash value, while the actual coal ash content is often affected by many factors. The use of this method often leads to low accuracy in coal quality detection. Therefore, in order to effectively improve the detection efficiency and accuracy of flotation coal ash, it is necessary to improve the existing online detection device and method for ash content in coal slime flotation. Summary of the invention

[0004] In order to overcome the deficiencies in the prior art, the present invention aims to solve the technical problem of providing an improvement in the hardware structure and detection method of an online detection device for ash content in coal slime flotation.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: an online detection device for ash content of coal slime flotation material, comprising a device housing, a pulse laser is arranged inside the device housing, the output end of the pulse laser is arranged directly opposite to a quartz convex lens, and the quartz convex lens is movably mounted on the bottom surface of the device housing through a lens bracket;

[0006] A conveyor belt is also provided through the bottom surface of the device housing, and the conveyor belt can input the coal slime mixture sample transported by the flotation machine bypass into the device housing;

[0007] The pulse laser and the quartz convex lens are arranged on one side of the conveyor belt, and a photodetector array is also arranged on the other side of the conveyor belt. The signal input end of the photodetector array is arranged directly opposite to the pulse laser, so that after the laser pulse emitted by the pulse laser passes through the quartz convex lens and is focused on the coal-slime mixture sample, the material on the surface of the sample is ionized to generate plasma, and after the plasma is cooled, a light signal with elemental characteristic spectrum lines is released, which is received by the photodetector array;

[0008] A control unit is also fixed to the side wall of the device shell by a bracket, and a central controller, a data acquisition module, a data communication module, a data storage module, and a power management module are arranged inside the control unit. The signal output end of the photoelectric detector array is connected to the data acquisition module through a wire, and the central controller is connected to the control ends of the pulse laser, the conveyor belt, the data communication module, and the data storage module through wires, and the power input end of the central controller is connected to the power management module.

[0009] The device shell is specifically installed on a pair of supporting brackets, and the outer sides of the supporting brackets are specifically connected and fixed to the device shell through connecting brackets. The setting width between the supporting brackets is adapted to the width of the conveyor belt.

[0010] The photoelectric detector array includes a plurality of photoelectric detectors, each of which is embedded with a filter for filtering light signals of different wavelengths, and the type of the filter is determined by the spectral line characteristics of the elements contained in the coal slime mixture sample.

[0011] A method for detecting ash content of coal slime flotation material using an online detection device comprises the following detection steps:

[0012] S10: installing the detection device next to the flotation machine, connecting the conveyor belt of the detection device to the transport bypass of the flotation machine, and transporting the bypass sample to the detection device through the conveyor belt;

[0013] S20: Controlling the pulsed laser to emit laser pulses through the central controller. The laser pulses are focused on the coal slime mixture sample conveyed by the conveyor belt through a quartz convex lens, so that the material on the surface of the sample is ionized to generate plasma. After the plasma cools down, light with characteristic spectral lines of the elements is released. At this time, the photodetector array is controlled to receive the light information filtered from a specific filter. The photodetector array converts the light information into an electrical signal and sends it to the control unit. The data acquisition module processes the signal to obtain the characteristic light intensity signal of the element corresponding to the filter, and finally sends it to the central controller.

[0014] S30: The central controller receives light intensity data transmitted by multiple photodetectors, which respectively represent the light intensity of characteristic spectrum lines of each element. Each light intensity signal is input as input layer data into a regression model of light intensity data and ash content built into the central controller for training a multivariate linear regression model.

[0015] S40: input the data to be detected into the trained regression model, calculate the predicted coal ash content data, store and output it, and transmit it to the host computer through the data communication module for display.

[0016] The step S20 specifically includes:

[0017] S21: The central controller specifically controls the triggering of the pulsed laser, and the energy of the emitted laser is higher than the breakdown threshold of the surface material of the coal slime sample, and is focused on the sample surface through a quartz convex lens, and the breakdown threshold I of the emitted laser is controlled to be:

[0018]

[0019] In the above formula: V e is the volume of the area illuminated by the focused laser, in m 3 ;

[0020] n b is the density of neutral atoms, in m / V;

[0021] c g It is a parameter related to gas;

[0022] ω m is the emission frequency of the laser, in Hz;

[0023] k ba is the collision probability of neutral atoms, in s -1 ;

[0024] S22: Molecules and atoms on the surface of coal slime continuously absorb energy and generate ionization, forming a large number of electrons. Under the continuous action of laser, free electrons are accelerated and bombard atoms, causing the number of electrons to double in a short time, generating plasma;

[0025] S23: The electrons of ionized atoms in the plasma jump from high energy level to low energy level, radiating information about the components and component concentrations of the target objects. The intensity of the radiation spectrum is: I = aC b ;

[0026] In the above formula: I represents the spectral line intensity of the radiation, unit J / cm 3 s;

[0027] a and b are two constants;

[0028] C is the concentration of the element in the coal slime sample, in mol;

[0029] S24: Light of corresponding wavelength is filtered out by a narrow-band filter of a specific type, received by the photodetector array, converted into an electrical signal and transmitted to the data acquisition module, and then transmitted to the central controller after the acquisition is completed.

[0030] The step S30 specifically includes:

[0031] S31: The expression of the multivariate linear regression model to be trained is:

[0032] f(x)=k T x+b;

[0033] In the above formula: f(x) is the output of the model, specifically the predicted ash content;

[0034] k T is the feature weight of the model;

[0035] x is the input vector, which is specifically the light intensity data collected by all data acquisition modules;

[0036] b is the bias of the model;

[0037] Introducing mean square error The performance of the model is measured as follows:

[0038]

[0039] In the above formula, y i is the actual ash content;

[0040] is the predicted value;

[0041] The optimization goal is to minimize the mean square error. The least square method is introduced to minimize the k in the mean square error function. T Take the derivative and set it equal to zero to get the optimized solution, and finally get the multivariate linear regression equation, which is expressed as:

[0042]

[0043] In the above formula: X is the feature matrix, which contains the feature data of all training samples; Y is the actual observed value of coal ash content;

[0044] The model is trained to the optimal result through least square estimation to obtain the actual multivariate linear regression equation;

[0045] S32: Import the trained multivariate linear regression equation into the central controller, receive the light intensity data transmitted by multiple detectors as the x value of the regression equation, and finally calculate the predicted y value of the coal ash content.

[0046] The beneficial effects of the present invention relative to the prior art are as follows: the present invention sets a laser pulse generating device in the detection device, so that the emitted laser pulse is focused by a quartz convex lens and then penetrates the surface material of the coal slime, causing it to ionize and generate plasma, and when the excited state atom returns to the ground state, the spectral information with the surface material of the coal slime is released, and each element releases a specific wavelength spectrum line, and after filtering by a specific type of filter, the detector receives the light intensity data of the specific wavelength range, and transmits it to the data acquisition module as the input data of the multivariate linear regression model, which can more efficiently and accurately extract the mineral impurity spectral data in the coal slime sample, and improve the accuracy and stability of the subsequent multivariate linear regression model. The present invention adopts a multivariate linear regression model to predict the coal ash content, and the loss function selects the mean square error function, and the least squares estimation is used to approach the optimal value, and the trained model is deployed in the main controller. After receiving the input data from the main controller, the predicted coal ash content can be calculated, and the cumbersome steps of transmitting the spectral data to the host computer for detection are omitted, and the real-time performance of the coal ash content detection is improved, and the data transmission is reduced. The host computer is only used to display the predicted coal ash content, and the stability of the online detection system is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The present invention will be further described below in conjunction with the accompanying drawings:

[0048] Figure 1 It is a schematic diagram of the structure of the detection device of the present invention;

[0049] Figure 2 is a flow chart of the steps of the detection method of the present invention;

[0050] Figure 3 A structural guide diagram of a multivariate linear regression model used in the detection method of the present invention;

[0051] In the figure: 1 is the device housing, 2 is the pulse laser, 3 is the quartz convex lens, 4 is the lens bracket, 5 is the conveyor belt, 6 is the photoelectric detector array, 10 is the control unit, 11 is the central controller, 12 is the data acquisition module, 13 is the data communication module, 14 is the data storage module, 15 is the power management module, 20 is the supporting bracket, and 21 is the connecting bracket. DETAILED DESCRIPTION

[0052] like Figure 1As shown, in order to overcome the shortcomings of the prior art, the present invention specifically proposes a device for detecting ash content of coal slime flotation based on LIBS, and uses Laser-Induced Breakdown Spectroscopy (LIBS) to achieve the effect of quickly detecting the ash content of coal slime flotation; it mainly includes a detection device and a host computer. When installed, the detection device is placed in a black box and connected to a bypass of the flotation tailings tank. The detection device includes a stainless steel device housing, a pulsed laser, a quartz convex lens, and a plurality of photoelectric detectors, each detector is embedded with a filter for filtering different wavelengths, and also includes a data acquisition module, a central controller, a data communication module, a data storage module, and a power management module;

[0053] During detection, laser pulses are mainly used to irradiate the surface of coal slime. When the laser energy emitted by the pulsed laser is greater than the breakdown threshold of the coal slime surface material, the coal slime surface material will produce high-temperature plasma. After the plasma cools, it releases light with a specific wavelength, which is received by the detector and converted into spectral data, which can map the content of inorganic matter inside it. Then, a neural network is trained to establish a regression model of spectral data and ash content, and the trained model is deployed in the main controller. The collected data is passed to the trained multivariate linear regression model in the main control to perform ash regression prediction. Finally, the test results are displayed in real time by the host computer to complete the online detection of the ash content of the coal slime flotation material, which can effectively improve the accuracy of ash content detection and shorten the detection time.

[0054] In the embodiment of the present invention, a conveyor belt is specifically provided for conveying the coal slime sample conveyed out of the flotation machine bypass, and a valve is provided on the flotation machine bypass to control the opening and closing of the bypass; a pulse laser is fixed by a laser bracket, and the laser bracket is connected to a central controller, and the opening and closing of the laser is adjusted by the central controller; at the same time, a plurality of photoelectric detectors are installed and fixed on the bottom of the device, facing the pulse laser, and connected to the data acquisition module in the control unit; the control unit is installed on the side of the device shell by a fixed bracket, and includes a data acquisition module, a data communication module, a data storage module, and a central controller; the central controller is respectively connected to the data acquisition module, the data communication module, the data storage module and the power management module; before use, the entire device shell is placed on the support brackets on both sides of the conveyor belt and fixed by connecting brackets.

[0055] The control unit of the present invention needs to realize the acquisition of light intensity data of different wavelengths, data storage between the data storage module, data transmission between the data storage module and the host computer, and the deployment of the multiple linear regression model; the filter embedded in each photoelectric detector can be selected according to the characteristic spectrum of the main elements in the coal ash, for example: Si256.16nm, Fe275.57nm, Mg279.55nm, Al308.22nm, etc. The data obtained by each filter corresponding to the detector is stored in the data storage module after the data acquisition module as the input data of the multiple linear regression model.

[0056] In order to facilitate the implantation of the multivariate linear regression model, the present invention selects the ZYNQ7000 series of Xilinx Company that supports embedded systems as the central controller, the data acquisition module selects AD9280 as the analog-to-digital converter, lora as the data communication module, and the data storage module selects SDRAM of Micron Company. The pulse laser is initialized and configured and controlled to start and stop by the central controller. The laser pulse emitted by the laser is focused on the coal slime sample through a quartz convex lens. The generated spectral information is received by the optical ground detector embedded with a filter, and the data acquisition module performs raw data acquisition after being converted into an electrical signal, and is transmitted to SDRAM. The host computer receives data through the serial port to train the regression model, and the multivariate linear regression equation is obtained and then embedded into the model in ZYNQ. After reading the input data from the SDRAM, linear regression is performed to obtain the predicted coal ash content, and finally the prediction result is sent to the host computer through the lora wireless network.

[0057] Furthermore, the present invention also provides an online detection method for ash content in coal slime flotation, which is mainly implemented based on the above detection device. Figure 2 The specific process step diagram of the on-line detection method of ash content of coal slime flotation material of the present invention specifically includes the following steps:

[0058] S10: The device is installed next to the flotation machine and is connected to the flotation machine through a bypass. The bypass sample is transported to the device, and the switch of the laser is controlled by the host computer.

[0059] S20: The laser emits high-intensity laser pulses, which are focused on the coal-slime mixture sample through a quartz convex lens. The material on the sample surface is ionized to produce plasma, and after the plasma cools, it releases light with elemental characteristic spectral lines. The detector receives the light information filtered from a specific filter, converts it into an electrical signal, and obtains the characteristic light intensity of the element corresponding to the filter, which is transmitted to the main controller as the original data set. The specific steps of this scheme are:

[0060] S21: The upper computer controls the laser triggering, and the emitted laser energy is higher than the breakdown threshold of the surface material of the coal slime sample, and is focused on the sample surface through the quartz convex lens. The breakdown threshold I of the emitted laser is:

[0061]

[0062] In the above formula: V e is the volume of the area irradiated by the focused laser (unit: m 3 );

[0063] n b is the density of neutral atoms (unit: m / V);

[0064] c g It is a parameter related to gas;

[0065] ω m is the emission frequency of the laser (unit: Hz);

[0066] k ba is the collision probability of neutral atoms (unit: s -1 ).

[0067] S22: The molecules and atoms on the surface of the coal slime continuously absorb energy to produce ionization, forming a large number of electrons; under the continuous action of high-intensity lasers, the free electrons are accelerated and bombard the atoms, causing the number of electrons to double in a short period of time, generating plasma.

[0068] S23: The electrons of ionized atoms in the plasma transition from high energy level to low energy level, radiating information about the components and component concentrations of the target. The intensity of the radiation spectrum is:

[0069] I=aC b ;

[0070] In the above formula: I represents the spectral line intensity of the radiation (unit: J / (cm 3 s));

[0071] a and b are two constants;

[0072] C is the concentration of the element in the coal slime sample (unit: mol).

[0073] S24: A narrowband filter of a specific type filters out light of corresponding wavelength, which is received by the detector and converted into an electrical signal and transmitted to the data acquisition module. After the acquisition is completed, it is transmitted to the main controller.

[0074] S30: The main controller receives the light intensity data transmitted by multiple detectors, which respectively represent the light intensity of the characteristic spectrum line of each element, and transmits it into the regression model of light intensity data and ash content as the input data of the input layer, which is used to train the neural network model to improve the accuracy of detection. The model parameters are extracted from the trained model and integrated into the main controller. The specific steps of this scheme are as follows:

[0075] S31: Figure 3As shown, the expression of the multiple linear regression model is:

[0076] f(x)=k T x+b;

[0077] In the above formula: f(x) is the output of the model, specifically the predicted ash content;

[0078] k T is the feature weight of the model;

[0079] x is the input vector, which is specifically the light intensity data collected by all data acquisition modules;

[0080] b is the bias of the model;

[0081] Introducing mean square error Measuring the performance of the model:

[0082]

[0083] In the above formula, y i is the actual ash content;

[0084] is the predicted value;

[0085] The optimization goal is to minimize the mean square error. The least square method is introduced to minimize the k in the mean square error function. T Take the derivative and set it equal to zero to get the optimal solution, and finally get the multivariate linear regression equation;

[0086]

[0087] In the above formula, X is the feature matrix, which contains the feature data of all training samples;

[0088] Y is the actual observed value of coal ash content;

[0089] The model is trained to the optimal result through least square estimation to obtain the actual multivariate linear regression equation.

[0090] S32: The trained multivariate linear regression equation is imported into the main controller, and the light intensity data transmitted by multiple detectors is received as x of the regression equation, and the predicted coal ash content y is calculated.

[0091] S40: The trained model outputs the predicted coal ash content, which is transmitted to the host computer through the data communication module for display, and the current coal ash content is obtained, so as to more intuitively evaluate the current coal quality and dynamically adjust the dosage of the flotation agent.

[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An online detection device for ash content of coal slime flotation material, comprising a device housing (1), characterized in that: A pulse laser (2) is arranged inside the device housing (1); the output end of the pulse laser (2) is arranged directly opposite to the quartz convex lens (3); and the quartz convex lens (3) is movably mounted on the bottom surface of the device housing (1) via a lens bracket (4); A conveyor belt (5) is also provided through the bottom surface of the device housing (1), and the conveyor belt (5) is capable of inputting the coal slime mixture sample transported by the flotation machine bypass into the device housing (1); The pulse laser (2) and the quartz convex lens (3) are arranged on one side of the conveyor belt (5), and a photodetector array (6) is also arranged on the other side of the conveyor belt (5). The signal input end of the photodetector array (6) is arranged directly opposite to the pulse laser (2), so that after the laser pulse emitted by the pulse laser (2) passes through the quartz convex lens (3) and is focused on the coal-slime mixture sample, the material on the surface of the sample is ionized to generate plasma, and after the plasma is cooled, a light signal with elemental characteristic spectrum lines is released, which is received by the photodetector array (6); A control unit (10) is also fixed to the side wall of the device housing (1) via a bracket, and a central controller (11), a data acquisition module (12), a data communication module (13), a data storage module (14), and a power management module (15) are arranged inside the control unit (10); the signal output end of the photodetector array (6) is connected to the data acquisition module (12) via a wire, the central controller (11) is connected to the control ends of the pulse laser (2), the conveyor belt (5), the data communication module (13), and the data storage module (14) via wires, and the power input end of the central controller (11) is connected to the power management module (15).

2. The on-line detection device for ash content of coal slime flotation material according to claim 1 is characterized in that: The device housing (1) is specifically installed on a pair of support brackets (20), and the outer sides of the support brackets (20) are specifically connected and fixed to the device housing (1) via connecting brackets (21), and the setting width between the support brackets (20) is adapted to the width of the conveyor belt (5).

3. The on-line detection device for ash content of coal slime flotation material according to claim 2 is characterized in that: The photoelectric detector array (6) comprises a plurality of photoelectric detectors, each of which is embedded with a filter for filtering light signals of different wavelengths, and the type of the filter is determined by the spectral line characteristics of the elements contained in the coal slime mixture sample.

4. The method for detecting ash content of coal slime flotation material by an online detection device according to claim 3, characterized in that: The detection steps include the following: S10: installing the detection device next to the flotation machine, connecting the conveyor belt (5) of the detection device to the transport bypass of the flotation machine, and transporting the bypass sample to the detection device through the conveyor belt (5); S20: The central controller (11) controls the pulse laser (2) to emit laser pulses, and the laser pulses are focused on the coal-slime mixture sample conveyed by the conveyor belt (5) through the quartz convex lens (3), so that the material on the surface of the sample is ionized to generate plasma. After the plasma cools down, light with characteristic spectral lines of the elements is released. At this time, the photodetector array (6) is controlled to receive light information filtered from a specific filter. The photodetector array (6) converts the light information into an electrical signal and sends it to the control unit (10). The data acquisition module (12) processes the signal to obtain a characteristic light intensity signal of the element corresponding to the filter, and finally sends the signal to the central controller (11); S30: The central controller (11) receives light intensity data transmitted by multiple photodetectors, which respectively represent the light intensity of characteristic spectrum lines of each element. Each light intensity signal is input as input layer data into a regression model of light intensity data and ash content built into the central controller (11) for training a multivariate linear regression model. S40: inputting the data to be detected into the trained regression model, calculating the predicted coal ash content data, storing and outputting the data, and transmitting the data to the host computer for display through the data communication module (13).

5. The method for detecting ash content of coal slime flotation material by an online detection device according to claim 4, characterized in that: The step S20 specifically includes: S21: The central controller (11) specifically controls the triggering of the pulse laser (2), and emits laser energy higher than the breakdown threshold of the surface material of the coal slime sample, and focuses on the sample surface through the quartz convex lens (3), and controls the breakdown threshold I of the emitted laser to be: In the above formula: V e is the volume of the area irradiated by the focused laser, in m 3 ; nb is the density of neutral atoms, in m / V; c g It is a parameter related to gas; ω m is the emission frequency of the laser, in Hz; k ba is the collision probability of neutral atoms, in s -1 ; S22: Molecules and atoms on the surface of coal slime continuously absorb energy and generate ionization, forming a large number of electrons. Under the continuous action of laser, free electrons are accelerated and bombard atoms, causing the number of electrons to double in a short time, generating plasma; S23: The electrons of ionized atoms in the plasma jump from high energy level to low energy level, radiating information about the components and component concentrations of the target objects. The intensity of the radiation spectrum is: I = aC b ; In the above formula: I represents the spectral line intensity of the radiation, unit J / cm 3 s; a and b are two constants; C is the concentration of the element in the coal slime sample, in mol; S24: Light of corresponding wavelength is filtered out by a narrow-band filter of a specific type, received by the photodetector array (6), converted into an electrical signal and transmitted to the data acquisition module (12), and then transmitted to the central controller (11) after the acquisition is completed.

6. The method for detecting ash content of coal slime flotation material by an online detection device according to claim 5, characterized in that: The step S30 specifically includes: S31: The expression of the multivariate linear regression model to be trained is: f(x)=k T x+b; In the above formula: f(x) is the output of the model, specifically the predicted ash content; k T is the feature weight of the model; x is the input vector, which is specifically the light intensity data collected by all data acquisition modules; b is the bias of the model; Introducing mean square error The performance of the model is measured as follows: In the above formula, y i is the actual ash content; is the predicted value; The optimization goal is to minimize the mean square error. The least square method is introduced to minimize the k in the mean square error function. T Take the derivative and set it equal to zero to get the optimized solution, and finally get the multivariate linear regression equation, which is expressed as: In the above formula: X is the feature matrix, which contains the feature data of all training samples; Y is the actual observed value of coal ash content; The model is trained to the optimal result through least square estimation to obtain the actual multivariate linear regression equation; S32: Import the trained multivariate linear regression equation into the central controller (11), receive the light intensity data transmitted by multiple detectors as the x value of the regression equation, and finally calculate the predicted coal ash content y value.

Images