Two-parameter calibration-free pulverized coal temperature measurement method based on acoustic tomography

Abstract

The invention relates to a two-parameter calibration-free pulverized coal temperature measurement method based on acoustic tomography, and the method comprises the following steps: 1, arranging a plurality of loudspeaker and microphone combination modules around a measurement field domain for transmitting and receiving acoustic signals, and forming an annular acoustic sensor array; step 2, multi-frequency sound transit time and amplitude attenuation measurement; 3, constructing a tomography inversion problem model of sound velocity and sound attenuation distribution; 4, respectively realizing tomography inversion of sound velocity and sound attenuation distribution through an improved joint iterative reconstruction method SIRT, and solving pulverized coal accumulation slowness and sound attenuation coefficient distribution under each frequency; 5, calculating a corresponding multi-frequency complex wave number through the multi-frequency slowness and the sound attenuation coefficient in each grid; step 6, constructing a temperature decoupling model of a complex wave number; and 7, regarding the temperature decoupling model of the complex wave number as a nonlinear objective function optimization problem, establishing a nonlinear objective function, and performing parameter decoupling and temperature calculation.

Description

Technical Field

[0001] The invention belongs to the technical field of temperature measurement, and relates to a dual-parameter calibration-free coal powder temperature measurement method based on acoustic tomography, which can be used for visual monitoring of stored coal powder temperature and spontaneous combustion early warning. Background Art

[0002] Coal resources are mainly transported, stored and burned in the form of piled coal powder. In this process, the piled coal powder is exposed to an oxidizing environment and undergoes an exothermic oxidation reaction with the air, forming a local high-temperature area. When the temperature rises above the auto-ignition point, it spontaneously ignites, causing production safety accidents. Therefore, accurately measuring the temperature of coal powder and identifying the local high-temperature area in the piled coal powder are crucial to detecting and preventing coal spontaneous combustion.

[0003] At present, the means of measuring coal powder temperature can be divided into two categories: invasive and non-invasive. The invasive measurement technology using thermocouples and thermistors places the sensor at a specified position in the coal powder, utilizes the thermoelectric effect, and obtains the temperature change at the specified position by measuring the thermoelectric potential. It has the advantages of simple structure, low cost and wide measurement range. However, the internal heat conduction of the accumulated coal powder is poor. The single-point temperature measurement technology represented by thermocouples is easy to miss the local high-temperature area in the accumulated coal powder, and cannot achieve effective measurement of the global temperature distribution. The non-invasive temperature measurement technology represented by infrared thermal imaging relies on the infrared radiation of the object to determine the temperature. It can only measure the surface temperature of the accumulated coal powder and cannot provide the temperature changes inside the coal powder.

[0004] Acoustic tomography temperature measurement has the advantages of being non-invasive, radiation-free, low equipment cost, and a large measurement range. Compared with invasive measurement technologies such as thermocouples, acoustic tomography can invert the temperature distribution image in the entire sensing area and detect and identify local high-temperature areas inside the accumulated coal powder. Compared with non-invasive measurement technologies such as infrared thermal imaging, the acoustic method can penetrate the accumulated coal powder at a depth of several meters through low-frequency, especially sound wave excitation below 10kHz, and accurately measure its internal temperature changes. Therefore, acoustic temperature measurement has great potential in coal powder temperature measurement and has received widespread attention from researchers.

[0005] Acoustic temperature measurement mainly relies on measuring the change of sound velocity in coal powder to obtain temperature information. Constructing a response model of sound velocity in coal powder to temperature change is the key to acoustic tomography. Pile coal powder is a porous medium material. Sound waves propagate through the pores between coal powder particles. Its equivalent sound velocity is determined by the adiabatic sound velocity of the gas in the pores and the microstructural parameters of the porous medium (such as porosity, tortuosity and pore size distribution). In 1994, researchers proposed a tortuosity factor model for sound propagation pipelines, that is, the equivalent sound velocity in coal powder is proportional to the adiabatic sound velocity of the gas. Based on this model, Yan Hua proposed a characteristic factor calibration method in 2012. The temperature sound velocity characteristic factor is obtained through static measurement calibration, thereby converting the coal powder sound velocity in actual measurement into adiabatic sound velocity, and calculating the corresponding temperature through the adiabatic sound velocity. Hu Yonghui proposed a temperature adaptive characteristic factor measurement method in 2019 to further improve the measurement accuracy. However, these acoustic-based temperature measurement techniques require precise calibration of the characteristic factors under reference temperature conditions. For other coal samples with different particle sizes, porosities and microstructures, or when the particle size distribution structure of coal powder changes, the calibrated characteristic factors may be inaccurate, thus affecting the temperature measurement accuracy. In addition to the speed of sound, acoustic attenuation can also be used for coal powder measurement, and is relatively sensitive to changes in coal powder particle size and stacking structure. Utilizing this feature, the present invention proposes a dual-parameter calibration-free coal powder temperature measurement method based on acoustic tomography, which measures the changes in sound velocity and acoustic attenuation in coal powder, and simultaneously reconstructs the temperature and other porous media structural parameters. Summary of the invention

[0006] The present invention proposes a dual-parameter calibration-free coal powder temperature measurement method. The present invention utilizes the dispersion and attenuation phenomenon of sound propagation in coal powder. On the basis of traditional sound velocity measurement, the acoustic attenuation measurement is introduced to construct a sound velocity and acoustic attenuation fusion model. The sound velocity and acoustic attenuation coefficient distribution images of sound waves in the coal powder medium at multiple frequencies are obtained through acoustic tomography technology, and finally the temperature distribution and the distribution of porous medium structural parameters such as coal powder porosity, pore size median and standard deviation are decoupled, realizing a dual-parameter calibration-free coal powder temperature measurement method based on acoustic tomography. The technical solution is as follows:

[0007] A dual-parameter calibration-free coal powder temperature measurement method based on acoustic tomography comprises the following steps:

[0008] Step 1: Build the sensor array

[0009] Arrange multiple speaker and microphone combination modules around the measurement field to transmit and receive acoustic signals, forming a ring-shaped acoustic sensor array;

[0010] Step 2: Multi-frequency sound transit time and amplitude attenuation measurement

[0011] H different excitation frequencies are selected, and each speaker is excited in turn to emit a multi-frequency sound signal; when each speaker is excited, the microphones at other positions simultaneously collect and receive the multi-frequency sound signal waveform, and extract the transit time and amplitude attenuation;

[0012] Step 3: Construct a tomographic inversion problem model of sound velocity and sound attenuation distribution, and its objective function is as follows:

[0013]

[0014] where F is the second-order Laplacian matrix of the four-connected domain, which is used for the smoothing regularization of tomography inversion; λ s and λ α represent the regularization coefficients in the inversion of slowness and attenuation coefficient, respectively; is the tomographic projection matrix, the matrix elements L i,j represents the length of the intercept formed by the j-th path passing through the i-th grid; represents the transit time and amplitude attenuation measurement vector; N is the number of tomographic imaging grids, and M is the number of projection paths; It represents the slowness and acoustic attenuation coefficient vector of the accumulated coal powder in the measured field.

[0015] Step 4: The tomographic inversion of the sound velocity and sound attenuation distribution is realized by improving the joint iterative reconstruction method SIRT, and the distribution of the slowness and sound attenuation coefficient of the accumulated coal powder at each frequency is solved;

[0016] Step 5: Calculate the corresponding multi-frequency complex wave number through the multi-frequency slowness s and the acoustic attenuation coefficient α in each grid:

[0017] k=ωs-iα (4)

[0018] Where ω is the angular frequency of the sound wave;

[0019] Step 6: Calculate the complex wave number and the acoustic angular frequency ω, the coal powder temperature T, and the median of the coal powder pore size and the pore size standard deviation σ s and porosity φ, to construct a temperature decoupling model of complex wavenumber;

[0020] Step 7: Treat the temperature decoupling model of complex wavenumber as a nonlinear objective function optimization problem, establish a nonlinear objective function, and perform parameter decoupling and temperature calculation.

[0021] Furthermore, in step 1, the frequency range of the sound waves emitted and received by the sensor should cover 100 Hz to 6 kHz, and the radius of the circular array should not be less than 1 m.

[0022] Furthermore, in step 4, the iterative process of solving the tomographic inversion of the sound velocity and the sound attenuation distribution is realized by improving the joint iterative reconstruction method SIRT as follows:

[0023]

[0024] Where t is the number of iterations, μ represents the iteration step size;

[0025] Matrix P and matrix W represent the preprocessing matrix and geometric weight matrix respectively, which are used to improve the convergence performance and anti-noise performance. The calculation formula is as follows:

[0026]

[0027] in The distribution of the slowness and acoustic attenuation coefficient of accumulated coal powder at each frequency is solved respectively.

[0028] Furthermore, the method of step six is ​​as follows:

[0029] Based on the equivalent fluid assumption, the complex wave number k in the piled coal powder is given by the complex volume compressibility coefficient and dynamic bulk density The only thing that is certain is:

[0030]

[0031] Based on the three-parameter JCAL model, the complex volume compressibility and dynamic bulk density Calculated by the following formula:

[0032]

[0033] in and is the Padé approximation parameter, calculated by the following formula:

[0034]

[0035] The multiple intermediate variables in the formula are defined as follows:

[0036]

[0037] where ρ 0 , η and N pr represent the air density, dynamic viscosity coefficient and Prandtl number in the pores of the accumulated coal powder respectively; P 0 , γ and ω represent the gas static pressure, specific heat ratio, and sound wave angular frequency respectively; φ, and σ s represent porosity, median pore size of pulverized coal, and standard deviation of pore size, respectively; is the torsion coefficient of porous media;

[0038] According to formula (5), the complex wave number and the acoustic angular frequency ω, the pulverized coal temperature T, and the median pore size of the pulverized coal are calculated. and the pore size standard deviation σ s And the porosity φ, construct the temperature decoupling model of the complex wave number:

[0039]

[0040] Where K is the transfer function based on the three-parameter JCAL model, which describes the relationship between the complex wave number and angular frequency, temperature, coal powder porosity, and the median and standard deviation of pore size in coal powder.

[0041] Furthermore, in step 7, the temperature decoupling model of the complex wave number is regarded as a nonlinear objective function optimization problem, and a nonlinear objective function is established:

[0042]

[0043] in is the parameter vector to be solved, which is related to temperature, coal powder porosity, pore size median and standard deviation.

[0044] Furthermore, in step seven, the nonlinear objective function optimization is solved by a hybrid particle swarm optimization algorithm; the solution process is divided into two steps, including a standard particle swarm optimization calculation for approximating the global optimal solution, and an interior point method optimization based on this for accurate solution.

[0045] The beneficial effects of the present invention are as follows:

[0046] ● Realize visual measurement of the temperature distribution inside the piled coal powder, which can be used for local high-temperature zone detection and spontaneous combustion warning of the coal powder.

[0047] The acoustic temperature measurement model does not require calibration and can measure the temperature of coal powder with different particle size distribution and different stacking structures. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The following drawings describe selected embodiments of the present invention, which are exemplary and not exhaustive or limiting, wherein:

[0049] Figure 1 Flow chart of the calibration-free coal powder temperature measurement method based on dual-parameter acoustic tomography of the present invention;

[0050] Figure 2 Schematic diagram of tomography sensor array;

[0051] Figure 3 Three stacked coal powder temperature simulation models are selected for the examples of the present invention: (a) is the temperature distribution of the coal powder pre-combustion process at room temperature; (b) is the temperature distribution of the local high temperature area produced after the coal powder spontaneously combusts; and (c) is the complex temperature distribution formed after multiple coal powder spontaneously combusts.

[0052] Figure 4 and Figure 5 This is the inversion result diagram of acoustic attenuation coefficient and slowness tomography.

[0053] Figure 6 This is the temperature distribution diagram after decoupling.

[0054] Figure 7 This is the measurement error diagram of the temperature peak point in the local high temperature area under different measurement signal-to-noise ratios. DETAILED DESCRIPTION

[0055] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0056] The measurement method of the present invention includes acoustic sensor array excitation and acquisition, dual-parameter acoustic tomography inversion, and parameter decoupling and coal powder temperature calculation based on complex wave number. The calibration-free coal powder temperature measurement method of the present invention includes the following specific steps:

[0057] Step 1: Build the sensor array

[0058] 16 speaker and microphone combination modules are evenly arranged around the measurement field to transmit and receive sound signals, forming a circular acoustic sensor array. The sensor should transmit and receive sound waves in the frequency range of 100Hz to 6kHz, the circular array radius should be no less than 1m, and the sensor array forms a total of 16*15 / 2=120 effective acoustic measurement paths.

[0059] Step 2: Multi-frequency sound transit time and amplitude attenuation measurement

[0060] 16 speakers are excited in turn, emitting multi-frequency sound signals with six frequencies: 200Hz, 600Hz, 1kHz, 1.4kHz, 1.8kHz, and 2.2kHz. When each speaker is excited, 15 microphones at other positions simultaneously collect and receive the multi-frequency sound signal waveform and extract the transit time and amplitude attenuation. A total of 120*6=720 transit time and amplitude attenuation measurements are collected for each measurement.

[0061] Step 3: Build a tomographic inversion model of sound velocity and acoustic attenuation distribution

[0062] The tomographic inversion problem models of sound velocity and sound attenuation distribution are constructed respectively, and the objective function is as follows:

[0063]

[0064] where F is the second-order Laplacian matrix of the four-connected domain, which is used as a smoothing regularizer for tomography inversion. s and λ α represent the regularization coefficients in the inversion of slowness and attenuation coefficient, respectively. is the tomographic projection matrix, and the element L in the matrix i,j Represents the length of the section formed by the j-th path passing through the i-th grid. represents the transit time and amplitude attenuation measurement vector. N = 1296 is the number of tomographic imaging grids, and M = 120 is the number of projection paths. It represents the slowness and acoustic attenuation coefficient vector of the accumulated coal powder in the measured field.

[0065] Step 4: Tomography inversion solution

[0066] The tomographic inversion of sound velocity and sound attenuation distribution is realized by improving the joint iterative reconstruction method (SIRT Simultaneous Iterative Reconstruction Technique). The iterative process is as follows:

[0067]

[0068] Where t is the number of iterations and μ is the iteration step size.

[0069] Matrix P and matrix W represent the preprocessing matrix and geometric weight matrix respectively, which are used to improve the convergence performance and anti-noise performance. The calculation formula is as follows:

[0070]

[0071] in

[0072] Based on the tomographic solution method, the distribution of the slowness and acoustic attenuation coefficient of the accumulated coal powder at six frequencies was solved respectively.

[0073] Step 5: Calculate the complex wave number distribution

[0074] Calculate the corresponding multi-frequency complex wave number by using the multi-frequency slowness s and the acoustic attenuation coefficient α in each grid As shown in Formula 4:

[0075] k=ωs-iα (4)

[0076] Where ω = [200Hz 600Hz 1kHz 1.4kHz 1.8kHz 2.2kHz] is the angular frequency of the sound wave.

[0077] Step 6: Build a temperature decoupling model based on complex wavenumber

[0078] Based on the equivalent fluid assumption, the complex wave number k in the piled coal powder can be obtained from the complex volume compressibility coefficient and dynamic bulk density Uniquely determined, as shown in Formula 5:

[0079]

[0080] Based on the three-parameter JCAL (Johnson-Champoux-Allard-Lafarge) model, the complex volume compressibility and dynamic bulk density It can be calculated by the following formula:

[0081]

[0082] in and is the Padé approximation parameter, which can be calculated by the following formula:

[0083]

[0084] The multiple intermediate variables in the formula are defined as follows:

[0085]

[0086] where ρ 0 , η and N pr Respectively represent the air density, dynamic viscosity coefficient and Prandtl number in the pores of the accumulated coal powder. 0 ,γ, andω represent the gas static pressure, specific heat ratio, and sound wave angular frequency, respectively. φ, and σ s represent the porosity, median pore size and standard deviation of pore size of coal powder, respectively. is the torsion coefficient of porous media.

[0087] Air density ρ 0 and Prandtl number N pr They are all temperature-related physical quantities, and the calculation formula is as follows:

[0088]

[0089] The molar mass of air is M n =0.029 kg·mol -1 , gas constant R = 8.31 J·mol - 1.K -1 , dynamic viscosity coefficient η, constant pressure specific heat capacity C p , and thermal conductivity k g It is also a temperature-related physical quantity. For standard gas, its fitting calculation formula is as follows:

[0090]

[0091] Substituting formula (6-10) into formula (5), we can get the complex wave number, acoustic angular frequency ω, coal powder temperature T, and coal powder pore size median and the pore size standard deviation σ s And the relationship model of five parameters such as porosity φ.

[0092]

[0093] Where K is the transfer function based on the three-parameter JCAL model, which describes the relationship between the complex wave number and angular frequency, temperature, coal powder porosity, pore size median and standard deviation in coal powder.

[0094] Step 7: Parameter decoupling and temperature calculation

[0095] According to the temperature decoupling model of complex wave number in formula (11), parameter decoupling can be transformed into nonlinear objective function optimization. The objective function is as follows:

[0096]

[0097] in is the parameter vector to be solved, including the temperature, coal powder porosity, pore size median and standard deviation, is the multi-frequency complex wave number measurement vector, and N=6 is the number of acoustic measurement frequencies.

[0098] The nonlinear objective function optimization can be solved by hybrid particle swarm optimization algorithm. The solution process is divided into two steps, including standard particle swarm optimization calculation for approximating the global optimal solution, and interior point optimization calculation based on this for accurate solution.

[0099] Regarding the particle swarm algorithm approaching the global optimal solution, the following explains the particle swarm parameter setting, particle swarm initialization, iterative process and termination conditions.

[0100] (1) The basic parameters of the particle swarm algorithm are set as follows: the particle swarm size is 200, the self-learning and mutual learning coefficients c 1 and c 2 are 1.5 and 2.0 respectively, and the speed weight parameter w is 0.9. The ranges of the parameter solution sets to be solved are {T:273.15≤T≤900}, {φ:0≤φ≤1} and {σ s :0.2≤σ s ≤0.5}.

[0101] (2) Particle swarm initialization: According to the solution set range, 200 sets of parameters to be solved are randomly selected as the initial positions of the particles, and the corresponding initial velocities, initial neighborhood sizes, and initial inertia parameters of the particles are calculated, and the counters are initialized to 0.

[0102] (3) Iteration process: Iteration includes updating the positions and velocities of all particles, calculating the fitness of particles, and calculating the global optimal position. In the tth iteration, the iteration formula of the i-th particle is as follows:

[0103]

[0104] Where x and v are the particle position and velocity, r 1 and r 2 is a random number on [0,1], and are the historical optimal solution of the i-th particle and the global optimal solution of the population respectively.

[0105] (4) Termination condition: The iteration is terminated when any of the following conditions are met, including: 1) The relative change of the particle position iteration is less than 10 -10 , 2) The number of iterations is greater than 8000 times, 3) The number of objective function calculations is greater than 30000 times.

[0106] The interior point method is used for accurate solution in optimization calculation. The following is an explanation of the construction of the interior point method objective function, the iterative process and the termination conditions.

[0107] (1) Construction of interior point method objective function: The interior point method optimization calculation is used for accurate solution. Its initial value is inherited from the particle swarm optimization result of the previous step, and the barrier function method is used to implement the inequality constraint to ensure that the optimization result is within the solution set. The interior point method optimization problem is:

[0108]

[0109] Where G(x) is the logarithmic barrier function term, and the solution ranges of the parameters to be solved are {T:273.15≤T≤900}, {φ:0≤φ≤1} and {σ s :0.2≤σ s ≤0.5}, μ is the barrier function parameter, and its initial value is 10.

[0110] (2) Iteration process:

[0111] The iterative process of the interior point method is divided into two layers, the inner and outer layers. The inner layer iteration solves formula (15) by gradient descent, that is,

[0112]

[0113] Where H is the objective function at x t-1 The Hessian matrix of is the gradient of the objective function, and t is the number of inner loops.

[0114] When the inner loop triggers the termination condition, return to the outer loop and update the barrier function parameter μ

[0115]

[0116] Where p is the number of outer cycles.

[0117] (3) Termination condition: The iteration terminates when any of the following conditions is met, including: 1) The objective function changes by less than 10 -10 , 2) The number of iterations of the outer loop is greater than 200 times, 3) The number of iterations of the inner loop is greater than 1000 times.

[0118] Figure 1 This is a flow chart of the calibration-free coal powder temperature measurement method based on dual-parameter acoustic tomography of the present invention. The measurement process is mainly divided into three steps. Step one is the excitation and acquisition of the acoustic sensor array, which is used to measure the corresponding transit time and amplitude attenuation in 120 projection paths at 6 frequencies. Step two is the inversion of dual-parameter acoustic tomography, which inverts the distribution images of the slowness and attenuation coefficient of the accumulated coal powder at 6 frequencies. Step three is parameter decoupling and temperature solution, and the corresponding complex wavenumber distribution is calculated using the slowness and attenuation coefficient of the accumulated coal powder. The complex wavenumbers at 6 frequencies are used to realize the decoupling of temperature and porous medium structure parameters, and finally the temperature distribution image is solved.

[0119] Figure 2 It is a tomographic imaging sensor array, which includes 16 transceiver modules. Each transceiver module is composed of a speaker and a microphone. The microphone is placed in front of the speaker. Its size is much smaller than the speaker, so it can be regarded as receiving at the same position. The sensor array can provide 120 projection paths.

[0120] Figure 3 The three simulation models of coal powder stacking temperature selected for the examples of the present invention are one or more Gaussian point heat source temperature fields. (a) is the temperature distribution of coal powder pre-combustion process at room temperature, with a temperature range of 20-100°C. (b) and (c) are the temperature fields formed after coal powder self-combustion, with a temperature range of 300-900°C.

[0121] Figure 4 and Figure 5 The inversion results of acoustic attenuation coefficient and slowness tomography corresponding to the frequency of 2.2kHz are shown in the figure. As can be seen from the temperature distribution in the figure, acoustic tomography can accurately locate the local high temperature area through the distribution of slowness and attenuation coefficient. The imaging quality is high, with fewer artifacts and noise.

[0122] Figure 6The temperature distribution diagram after decoupling. The measured signal-to-noise ratio here is 50dB, and the particle swarm optimization algorithm parameters are set as follows: the particle swarm size is 80, the self-learning and mutual learning coefficients are 1.5 and 2.0, and the internal weight parameter is 0.9. The tomographic image quality of the three temperature field models is good, and the reconstructed image errors are 2.28%, 6.89% and 9.17%, respectively, which are all less than 10%, with fewer noise points and artifacts, and the measurement of the temperature in the local high temperature area is more accurate, and the imaging distortion in the high temperature area is not obvious.

[0123] Figure 7 The figure shows the measurement error of the temperature peak point in the local high temperature area under different measurement signal-to-noise ratios. It can be seen from the figure that when the measurement signal-to-noise ratio is greater than 45dB, the temperature field imaging error can be reduced to less than 10%. That is, the measurement signal-to-noise ratio requirement of this method is not less than 45dB.

Claims

1. A dual-parameter calibration-free coal powder temperature measurement method based on acoustic tomography, comprising the following steps: Step 1: Build the sensor array Arrange multiple speaker and microphone combination modules around the measurement field to transmit and receive acoustic signals, forming a ring-shaped acoustic sensor array; Step 2: Multi-frequency sound transit time and amplitude attenuation measurement H different excitation frequencies are selected, and each speaker is excited in turn to emit a multi-frequency sound signal; when each speaker is excited, the microphones at other positions simultaneously collect and receive the multi-frequency sound signal waveform, and extract the transit time and amplitude attenuation; Step 3: Construct a tomographic inversion problem model of sound velocity and sound attenuation distribution, and its objective function is as follows: where F is the second-order Laplacian matrix of the four-connected domain, which is used for the smoothing regularization of tomography inversion; λ s and λ α represent the regularization coefficients in the inversion of slowness and attenuation coefficient, respectively; is the tomographic projection matrix, the matrix elements L i,j represents the length of the intercept formed by the j-th path passing through the i-th grid; represents the transit time and amplitude attenuation measurement vector; N is the number of tomographic imaging grids, and M is the number of projection paths; It represents the slowness and acoustic attenuation coefficient vector of the accumulated coal powder in the measured field. Step 4: The tomographic inversion of the sound velocity and sound attenuation distribution is realized by improving the joint iterative reconstruction method SIRT, and the distribution of the slowness and sound attenuation coefficient of the accumulated coal powder at each frequency is solved; Step 5: Construct the corresponding multi-frequency complex wave number through the multi-frequency slowness s and the acoustic attenuation coefficient α in each grid: k=ωs-iα (4) Where ω is the angular frequency of the sound wave; Step 6: Based on the complex wave number and the acoustic angular frequency ω, the coal powder temperature T, and the median of the coal powder pore size and the pore size standard deviation σ s and porosity φ, to construct a temperature decoupling model of complex wavenumber; Step 7: Treat the temperature decoupling model of complex wavenumber as a nonlinear objective function optimization problem, establish a nonlinear objective function, and perform parameter decoupling and temperature calculation.

2. The dual-parameter calibration-free coal powder temperature measurement method according to claim 1 is characterized in that: In step 1, the frequency range of the acoustic waves emitted and received by the sensor should cover 100 Hz to 6 kHz, and the radius of the circular array should not be less than 1 m.

3. The dual-parameter calibration-free coal powder temperature measurement method according to claim 1 is characterized in that: Step 4: The iterative process of solving the tomographic inversion of the sound velocity and the sound attenuation distribution by improving the joint iterative reconstruction method SIRT is as follows: Where t is the number of iterations, μ represents the iteration step size; Matrix P and matrix W represent the preprocessing matrix and geometric weight matrix respectively, which are used to improve the convergence performance and anti-noise performance. The calculation formula is as follows: in The distribution of the slowness and acoustic attenuation coefficient of accumulated coal powder at each frequency is solved respectively.

4. The dual-parameter calibration-free coal powder temperature measurement method according to claim 3 is characterized in that: The method for step six is ​​as follows: Based on the equivalent fluid assumption, the complex wave number k in the piled coal powder is given by the complex volume compressibility coefficient and dynamic bulk density The only thing that is certain is: Based on the three-parameter JCAL model, the complex volume compressibility and dynamic bulk density Calculated by the following formula: in and is the Padé approximation parameter, calculated by the following formula: The multiple intermediate variables in the formula are defined as follows: where ρ0, η and N pr Respectively represent the air density, dynamic viscosity coefficient and Prandtl number in the pores of the accumulated coal powder; P0, γ and ω represent the gas static pressure, specific heat ratio, and angular frequency of the sound wave; φ、 and σ s represent porosity, median pore size of pulverized coal, and standard deviation of pore size, respectively; is the torsion coefficient of porous media; According to formula (5), based on the complex wave number and the acoustic angular frequency ω, the coal powder temperature T, the median value of the coal powder pore size S, and the pore size standard deviation σ s And the porosity φ, construct the temperature decoupling model of the complex wave number: Where K is the transfer function based on the three-parameter JCAL model, which describes the relationship between the complex wave number and angular frequency, temperature, coal powder porosity, and the median and standard deviation of pore size in coal powder.

5. The dual-parameter calibration-free coal powder temperature measurement method according to claim 4 is characterized in that: In step 7, the temperature decoupling model of complex wavenumber is regarded as a nonlinear objective function optimization problem, and the nonlinear objective function is established: in is the parameter vector to be solved, which is related to temperature, coal powder porosity, pore size median and standard deviation.

6. The dual-parameter calibration-free coal powder temperature measurement method according to claim 1, characterized in that: In step seven, the nonlinear objective function optimization is solved by a hybrid particle swarm optimization algorithm; the solution process is divided into two steps, including the standard particle swarm optimization calculation for approximating the global optimal solution, and the interior point method optimization based on this for accurate solution.

Images