Single-photon lidar horizontal direction aerosol extinction coefficient inversion method

Abstract

The application discloses a single-photon laser radar horizontal direction aerosol extinction coefficient inversion method and belongs to the technical field of atmospheric monitoring. The method is as follows: obtaining horizontal direction aerosol echo signal data, performing background denoising, geometric overlap factor correction, distance square correction and smoothing processing on the original signal; setting an initial reference point near the maximum detection distance, and combining differential comparison and wavelet change to position the mutation interval of the processed aerosol echo signal data; reselecting a new reference point between adjacent mutation intervals, and using a Collis-Fernald iteration algorithm to obtain a stable solution of the extinction coefficient boundary value near the new reference point; and substituting the obtained initial reference point and the extinction coefficient boundary value at the initial reference point into a Fernald backward integral formula to inversely obtain the aerosol extinction coefficient value on the distance between the initial reference point and the starting point of the detection path. The application can improve the precision and reliability of the extinction coefficient inversion.

Description

Technical Field

[0001] This invention belongs to the field of atmospheric monitoring technology and relates to a method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar. Background Technology

[0002] With the development of human society, environmental pollution has increasingly attracted people's attention. Air pollution is an important aspect of this, and aerosols are one of the main sources of air pollution. Aerosols are mixtures composed of various tiny particles and droplets that are suspended in the atmosphere, affecting the transparency and optical properties of the atmosphere. Therefore, the study of the optical properties of aerosols has important scientific significance and practical application value. Due to its unique technical advantages such as long detection range, good coherence, and high spatiotemporal resolution, lidar has become an indispensable tool in the field of aerosol research. However, this type of radar also has some drawbacks, such as the inability to achieve detection accuracy down to the single photon level, the influence of multiple reflections on ranging accuracy, and susceptibility to complex environmental conditions such as rain, fog, snow, haze, sandstorms, and strong light. Currently, there are various methods for measuring the extinction coefficient of aerosols in the horizontal direction, but due to the complexity of atmospheric conditions and the limitations of instruments, many problems and challenges still exist. For example, due to variations in aerosol particle size and concentration, the signal may be affected by background noise, geometric overlap factors, and signal abrupt change ranges, leading to errors and uncertainties.

[0003] Among the developed methods for detecting aerosol extinction coefficients using lidar, the prior art [1] (see Feng Shuai, Jiang Lihui, Xiong Xinglong, et al. Lidar visibility inversion with abrupt signal [J]. Infrared and Laser Engineering, 2017, 46(3):330001-0330001(7).) addresses the problem of inversion error caused by abrupt signals in the signal. It is based on locating and removing abrupt signals to perform extinction coefficient inversion. However, this method simplifies the atmosphere to a uniform distribution, which is too idealistic and not suitable for actual detection.

[0004] In prior art [2] (see Ma X, Wang C, Han G, et al. Regional atmospheric aerosolpollution detection based on LiDAR remote sensing[J].Remote Sensing,2019,11(20):2339.), a Collis-Fernald algorithm was proposed to invert the aerosol extinction coefficient. However, this method uses Fernald forward integration and does not take into account the error of the abrupt change interval in the inversion process.

[0005] In the prior art [3] (see Yang Bin, Mozus, Deng Chen, Ji Qiang, Bu Lingbing. A joint observation and extinction coefficient calibration inversion method based on lidar [P]. Jiangsu Province: CN114167448A, 2022-03-11.), the Collis algorithm is used to obtain the horizontal extinction coefficient. However, this method only uses the horizontal extinction coefficient as the initial value for the calibration of the vertical lidar extinction coefficient, and does not conduct further research on the horizontal extinction coefficient. Summary of the Invention

[0006] To address the shortcomings of existing technologies and solve the problem of significant errors in subsequent path extinction coefficient inversion caused by abrupt signal changes in traditional methods, the main objective of this invention is to provide a method for inverting the horizontal aerosol extinction coefficient using a single-photon lidar. This method uses horizontal aerosol echo signal data acquired by a single-photon lidar to invert the horizontal aerosol extinction coefficient, reducing errors in subsequent inversions caused by abrupt signal changes on the horizontal detection path and improving the accuracy and reliability of extinction coefficient inversion.

[0007] The objective of this invention is achieved through the following technical solution.

[0008] This invention discloses a method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar. The method acquires horizontal aerosol echo signal data and performs background denoising, geometric overlap factor correction, distance squared correction, and smoothing on the original signal. An initial reference point is set near the maximum detection range of the single-photon lidar, and a method combining differential comparison and wavelet transform is used to locate abrupt change intervals in the processed aerosol echo signal data. After determining the abrupt change intervals, new reference points are selected between adjacent abrupt change intervals, and the Collis-Fernald iterative algorithm is used to obtain the stable boundary value solution of the extinction coefficient at the new reference point. The obtained initial reference point and initial parameters are then compared. Substituting the boundary value of the extinction coefficient at the test point into the Fernald backward integral formula, the aerosol extinction coefficient value over the distance from the initial reference point to the starting point of the detection path is obtained. Taking the new reference point between adjacent abrupt change intervals as the starting point, the boundary value of the extinction coefficient at the new reference point is substituted into the Fernald backward integral formula again to obtain the extinction coefficient over the subsequent path from the new reference point to the starting point of the detection path. The above process is repeated until there are no abrupt change intervals. The aerosol extinction coefficient data obtained repeatedly are stitched together on the distance scale until the stitching and coverage are completed and converted into a two-dimensional aerosol extinction coefficient profile, thus realizing the horizontal aerosol extinction coefficient inversion of the single-photon lidar.

[0009] This invention discloses a method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar, comprising the following steps:

[0010] Step 1: Set up a single-photon lidar in the aerosol testing area using a horizontal gimbal. Use the single-photon lidar to detect aerosols in the horizontal direction and acquire horizontal aerosol echo signal data. Use single-photon lidar to improve the detection range and accuracy in complex and harsh environments.

[0011] The complex and harsh environment includes rainy and foggy weather, snowy weather, hazy weather, sandstorms, and strong sunlight.

[0012] Preferably, the lidar mentioned in step one is a superconducting nano-single-photon lidar, comprising a transmitting module, a receiving module, a storage module, and a horizontal gimbal module. The transmitting module includes a pulsed laser that emits pulsed laser signals, which are then directed towards the target direction via a beam expander and a half-wave plate. The receiving module includes a receiving telescope, a collimating lens, a narrowband filter, a polarizing beam splitter, and a converging lens. The returned backscattered signal is collimated by the collimating lens to form a single-direction light signal, and after noise removal by the filter, it passes through the polarizing beam splitter and converging lens, resulting in parallel and perpendicular polarized light entering the storage module. The storage module consists of a photodetector, a signal amplifier, and a host computer. The optical signal is converted into an electrical signal by the photodetector; this electrical signal is horizontal aerosol echo signal data, which is amplified and stored in the host computer. The horizontal gimbal module consists of a gimbal tripod and a rotating support.

[0013] As a preferred option, the 1550nm wavelength used in the single-photon lidar in step one is invisible to the human eye, which reduces the harm to the human eye while ensuring the accuracy of horizontal detection.

[0014] Step 2: Perform background denoising, geometric overlap factor correction, distance squared correction, and smoothing on the horizontal aerosol echo signal data obtained in Step 1.

[0015] By selecting a flat signal at a far distance from the detection point and taking the average value of the flat signal as the background noise, the background noise is subtracted from the aerosol echo signal data to obtain a denoised signal, thereby reducing the interference of radiation from the sun, earth, atmosphere and artificial light sources on the detection accuracy of aerosol echo signals.

[0016] Because the pulsed laser beam emitted by the pulsed laser interacts with various substances in the atmosphere, the energy attenuates with increasing distance. This energy attenuation is inversely proportional to the square of the distance, causing the signal to attenuate too quickly at long distances, thus affecting subsequent data processing. Therefore, distance square calibration is used to eliminate this effect.

[0017] The formula for calibrating the squared distance is:

[0018] X(r)=P(r·r) 2 (1)

[0019] In equation (1), X(r) is the distance squared correction signal, P(r) is the power signal, and r is the detection distance.

[0020] In single-photon lidar measurements, the divergence angle of the laser beam differs from the receiving field of view. The geometrical overlap factor refers to the fact that when photons are backscattered, only a portion of them are received, while others are restricted by the field of view and cannot be received, thus introducing errors. Therefore, experimental methods are used to correct the geometrical overlap factor, eliminating errors and improving the detection accuracy of single-photon lidar.

[0021] To reduce the interference of atmospheric turbulence and stray light on aerosol echo signal data, the signal data needs to be smoothed after background denoising in step two. Preferably, the five-point cubic method is used for smoothing, and the signal data obtained after smoothing has higher reliability and accuracy.

[0022] The five-point three-dimensional method is as follows:

[0023]

[0024] (2) In the formula, x(t) is the aerosol echo signal data to be smoothed, t=1,2,…,N.

[0025] Step 3: Set an initial reference point near the maximum detection range of the single-photon lidar. Due to the distance constraint, there are almost no photons returning near the maximum detection range. The atmosphere near the maximum distance can be regarded as a clean atmosphere with very little aerosol particles and uniform distribution, so as to use the Collis method to determine the boundary value of the extinction coefficient at the initial reference point.

[0026] In step three, the Collis method specifically works as follows:

[0027] S(r)=ln[P(r)·r 2 (3)

[0028]

[0029] Substituting equation (3) into equation (4), the extinction coefficient under uniform atmospheric conditions is obtained as follows:

[0030]

[0031] (3) In the formula, S is the logarithmic distance correction signal; P is the power of the lidar signal intensity; r is the lidar detection distance; (4) In the formula, E is the lidar fixed parameter, which is a constant; C is the lidar correction constant; β is the total atmospheric backscattering coefficient, including the atmospheric molecular backscattering coefficient and the aerosol molecular backscattering coefficient; (5) In the formula, σ is the solved total atmospheric extinction coefficient, including the atmospheric molecular extinction coefficient and the aerosol molecular extinction coefficient.

[0032] Step 4: Due to the uneven distribution of aerosols along the detection path, there are abrupt changes in the aerosol echo signal data, which affects the inversion accuracy of the aerosol extinction coefficient in subsequent paths. The abrupt change intervals in the echo signal data are determined by the differential comparison method, and the abrupt change starting point r of the abrupt change interval in the echo signal data is obtained. bps1 With mutation endpoint r bpe1 For the same echo signal data, the abrupt change starting point r of the abrupt change interval is obtained by wavelet transform method. bps2 With mutation endpoint r bpe2 .

[0033] As a preferred option, the differential comparison method in step four specifically involves: calculating the logarithmic distance correction signal and its first-order forward difference value from step three, comparing this value with a pre-set threshold G, and thus locating the abrupt change interval.

[0034] As a preferred option, the wavelet transform method in step four is as follows: wavelet decomposition is performed on the aerosol echo signal data processed in step three to obtain wavelet coefficients, threshold processing is performed on the wavelet coefficients, and the processed wavelet coefficients are reconstructed to obtain the reconstructed signal data. By calculating the error between the original signal data and the reconstructed signal data, the location of the abrupt change interval is completed.

[0035] In step four, the difference comparison method and the wavelet transform method each have their own advantages and disadvantages, resulting in different abrupt change intervals (r). bps1 ,r bpe1 ) and (r bps1 ,r bpe1 The results may differ, so the intersection method is used to extract the mutation intervals jointly determined by the two algorithms, thereby reducing the false positive rate and improving the accuracy and reliability of signal mutation interval determination.

[0036] Step 5: Using the abrupt change intervals determined in Step 4, determine new reference points for the aerosol echo signal data within adjacent abrupt change intervals, and use the Collis-Fernald iterative algorithm to stabilize the boundary value of the extinction coefficient at the new reference point.

[0037] Since the aerosol echo signal data between adjacent abrupt change intervals is relatively flat, the aerosol echo signal data between adjacent abrupt change intervals can be equivalent to a uniform distribution, so the extinction coefficient boundary value σ at the new reference point can be determined by the Collis method in step three.

[0038] Substituting the extinction coefficient boundary value σ at the new reference point into Fernald's integral formula, we obtain the average extinction coefficient σ within the distance of adjacent abrupt change intervals. mean Then determine σ and σ meanThe size of the extinction coefficient. If the boundary value σ of the extinction coefficient at the new reference point is greater than the average value σ of the extinction coefficient over that distance. mean , will σ mean As the boundary value of the extinction coefficient, it is substituted into Fernald's backward integral formula for iteration; if the boundary value σ of the extinction coefficient at the new reference point is less than the average value σ of the extinction coefficient within that distance... mean Then stop the iteration and set σ mean This serves as the boundary value for the extinction coefficient at the final new reference point.

[0039] Preferably, since the Fernald backward integration formula can handle longer pulse widths and is less susceptible to noise and interference compared to the Fernald forward integration formula, the Fernald backward integration formula is used in step five to invert the extinction coefficient, thereby improving the accuracy and reliability of the inversion of the aerosol extinction coefficient.

[0040] Fernald's backward integration formula is:

[0041]

[0042] (6) In the formula, r c For reference distance, S a =α a (r) / β a (r), S m =α m (r) / β m (r), where α m (r) represents the extinction coefficient of atmospheric molecules. The density of air molecules is obtained using the standard atmospheric model based on temperature, humidity, and pressure, and then calculated using the Rayleigh scattering theory. The boundary value α of the extinction coefficient of aerosol molecules is also given. a (r) is determined by the aerosol molecule scattering ratio 1+β a (r c ) / β m (r c To determine.

[0043] Step 6: Substitute the initial reference point and the boundary value of the extinction coefficient at the initial reference point obtained in Step 3 into the Fernald backward integral formula to invert the aerosol extinction coefficient value over the distance from the initial reference point to the starting point of the detection path.

[0044] Step 7: After completing the initial backward integration, determine if there are any abrupt change intervals after the initial reference point. If so, take the first new reference point between adjacent abrupt change intervals from Step 5 as the starting point. Substitute the extinction coefficient boundary value at the new reference point into the Fernald backward integration formula to invert the extinction coefficient along the subsequent path up to the starting point of the detection path. Repeat this process until there are no abrupt change intervals. By repeating this operation, the error in aerosol extinction coefficient inversion due to abrupt change intervals is reduced, thereby improving the detection accuracy and reliability of horizontal aerosol extinction coefficient inversion.

[0045] Step 8: Combine the initial aerosol extinction coefficient data obtained in Step 6 with the new aerosol extinction coefficient data obtained in Step 7, and continue this process until the composite data is complete.

[0046] Step 9: Obtain the one-dimensional aerosol extinction coefficient value in the horizontal detection direction of the single-photon lidar through Step 8. Based on the lidar scanning speed and angle, convert the one-dimensional aerosol extinction coefficient value into a two-dimensional aerosol extinction coefficient profile, thus realizing the inversion of the horizontal aerosol extinction coefficient.

[0047] Beneficial effects:

[0048] 1. The present invention discloses a method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar. In the adjacent abrupt change interval, the Collis method is used to determine the boundary value of the extinction coefficient at the new reference point. The value is compared with the average value of the extinction coefficient obtained by substituting it into the Fernald algorithm and solved by an iterative algorithm. After a finite number of iterations, a stable solution of the aerosol extinction coefficient boundary value at the new reference point can be obtained, thereby improving the accuracy of subsequent inversion results.

[0049] 2. The present invention discloses a method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar. After determining the abrupt change interval in the aerosol echo signal data using the difference comparison method and the wavelet transform method respectively, the final abrupt change interval is determined by the intersection method. This reduces the misjudgment rate of the abrupt change interval by a single algorithm and improves the accuracy and reliability of the determination of the signal abrupt change interval.

[0050] 3. The present invention discloses a method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar. The smoothing process of aerosol echo signal data uses the five-point cubic method to reduce the interference of atmospheric turbulence and stray light on aerosol echo signal data, thereby improving the detection accuracy and reliability of aerosol extinction coefficient inversion.

[0051] 4. The present invention discloses a method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar. The superconducting nano-single-photon lidar uses light invisible to the human eye with a wavelength of 1550nm for detection, which improves the detection range and accuracy in complex and harsh environments such as rain, fog, snow, haze, sandstorms and strong light, while reducing the harm to the human eye.

[0052] 5. The present invention discloses a method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar. The aerosol extinction coefficient values ​​obtained by the Fernald backward integral formula for each reference point are spliced ​​and overlaid on the data of the previous inversion, thereby reducing the inversion error caused by abrupt changes and improving the reliability of the inversion data. Attached Figure Description

[0053] Figure 1 A schematic flowchart of a method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar according to the present invention;

[0054] Figure 2 The graph showing the variation of the overlap factor of a single-photon lidar array with detection distance in this invention;

[0055] Figure 3 A schematic diagram of a horizontal aerosol extinction coefficient detection system for a single-photon lidar according to the present invention. Detailed Implementation

[0056] To better illustrate the purpose and advantages of the present invention, the invention will be further described below in conjunction with the accompanying drawings and examples.

[0057] Example 1:

[0058] like Figure 1 As shown in this embodiment, a method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar is disclosed. The specific implementation steps are as follows:

[0059] Step 1: Set up a single-photon lidar in the aerosol testing area using a horizontal gimbal. Use the single-photon lidar to detect aerosols in the horizontal direction and acquire horizontal aerosol echo signal data. Use single-photon lidar to improve the detection range and accuracy in complex and harsh environments.

[0060] The complex and harsh environment includes rainy and foggy weather, snowy weather, hazy weather, sandstorms, and strong sunlight.

[0061] The aforementioned lidar is a superconducting nanoscale single-photon lidar, comprising a transmitting module, a receiving module, a storage module, and a horizontal pan-tilt module. The transmitting module includes a pulsed laser that emits pulsed laser signals, which are then directed towards the target direction via a beam expander and a half-wave plate. The receiving module includes a receiving telescope, a collimating lens, a narrowband filter, a polarizing beam splitter, and a converging lens. The returned backscattered signal is collimated by the collimating lens to form a single-direction light signal, and after noise removal by the filter, it passes through the polarizing beam splitter and converging lens to enter the storage module as parallel and perpendicular polarized light. The storage module consists of a photodetector, a signal amplifier, and a host computer. The optical signal is converted into an electrical signal by the photodetector; this electrical signal is horizontal aerosol echo signal data, which is amplified and stored in the host computer. The horizontal pan-tilt module consists of a pan-tilt tripod and a rotating support.

[0062] In step one, the single-photon lidar uses a wavelength of 1550nm, which is invisible to the human eye, thus reducing harm to the human eye while ensuring horizontal detection accuracy.

[0063] Step 2: Perform background denoising, geometric overlap factor correction, distance squared correction, and smoothing on the horizontal aerosol echo signal data obtained in Step 1.

[0064] By selecting a flat signal at a far distance from the detection point and taking the average value of the flat signal as the background noise, the background noise is subtracted from the aerosol echo signal data to obtain a denoised signal, thereby reducing the interference of radiation from the sun, earth, atmosphere and artificial light sources on the detection accuracy of aerosol echo signals.

[0065] Because the pulsed laser beam emitted by the pulsed laser interacts with various substances in the atmosphere, the energy attenuates with increasing distance. This energy attenuation is inversely proportional to the square of the distance, causing the signal to attenuate too quickly at long distances, thus affecting subsequent data processing. Therefore, distance square calibration is used to eliminate this effect.

[0066] The formula for calibrating the squared distance is:

[0067] X(r)=P(r·r) 2 (7)

[0068] In equation (7), X(r) is the distance squared correction signal, P(r) is the power signal, and r is the detection distance.

[0069] In single-photon lidar measurements, the emission angle of the pulsed laser emitter and the receiving field of view are different. The geometrical overlap factor refers to the fact that when photons are backscattered, only a portion of them are received, while others are limited by the field of view and cannot be received, thus causing errors. Taking a parallel-axis transceiver system as an example, the emission angle of the pulsed laser emitter and the receiving field of view of the telescope are represented by α and β, respectively, and the geometrical overlap factor is represented by η.

[0070] As Figure 2 shown, the geometric overlap factor changes with the detection distance as follows: when the detection distance r ≤ R1, η = 0, and this distance is called the blind area of the single-photon lidar, that is, no laser signal enters the receiving field of view; when R1 < r < R2, 0 < η < 1, the echo signal of the single-photon lidar enters the transition region, and as the distance gets farther, more and more echo signals enter the receiving field of view; when r ≥ R2, η = 1, that is, the echo signal of the single-photon lidar completely enters the telescope receiving field of view. Therefore, the geometric overlap factor can be corrected by the experimental method to eliminate errors and improve the detection accuracy of the single-photon lidar.

[0071] The experimental method is selected to correct the geometric overlap factor: choose to calibrate on a night just after rain, at this time, the content of atmospheric aerosol is less within a certain range of the near-ground level, the air is relatively clean, and the atmospheric distribution is relatively uniform. Through the horizontal detection of the single-photon lidar and continuous signal acquisition for linear fitting calculation, the geometric overlap factor within the corresponding distance can be obtained.

[0072] When considering the geometric overlap factor, equation (4) can be regarded as:

[0073]

[0074] The lidar correction constant in equation (4) is equivalent to η(r) in equation (8), which is the correction result of the geometric overlap factor.

[0075] By transforming equation (8), we can get:

[0076] ln(P(r)·r 2 ) = ln(Eη(r)(β(r)) - 2σ(r)·r (9)

[0077] When the detection distance r = R2, the geometric overlap factor η = 1. At this time, the lidar echo signal is P0(r), and equation (9) can be changed to:

[0078]

[0079] Substituting A and B into equation (9) and transforming, we can get the geometric overlap factor as:

[0080]

[0081] To reduce the interference of atmospheric turbulence and stray light on aerosol echo signal data, after background denoising in step two, the signal data needs to be smoothed. The smoothing method uses a five-point cubic approach, where five adjacent data points are taken at a given point, and a cubic curve is fitted. The corresponding data values ​​on the cubic curve are then used as the smoothed result. The smoothed signal data has higher accuracy and reliability.

[0082] The five-point three-dimensional method is as follows:

[0083]

[0084] (12) In the formula, x(t) is the aerosol echo signal data to be smoothed, t=1,2,…,N.

[0085] Step 3: Set an initial reference point near the maximum detection range of the single-photon lidar. Due to the distance constraint, there are almost no photons returning near the maximum detection range. The atmosphere near the maximum distance can be regarded as a clean atmosphere with very little aerosol particles and uniform distribution, so as to use the Collis method to determine the boundary value of the extinction coefficient at the initial reference point.

[0086] Near the maximum detection range of a single-photon lidar, according to P(r)r 2 / β m (r) is used to determine the reference point as far as possible.

[0087] In step three, the Collis method specifically works as follows:

[0088] S(r)=ln[P(r)·r 2 (13)

[0089]

[0090] Substituting equation (13) into equation (14), the extinction coefficient under uniform atmospheric conditions is obtained as follows:

[0091]

[0092] (13) In the formula, S is the logarithmic distance correction signal; P is the power of the lidar signal intensity; r is the lidar detection distance; (14) In the formula, E is the lidar fixed parameter, which is a constant; C is the lidar correction constant; β is the total atmospheric backscattering coefficient, including the atmospheric molecular backscattering coefficient and the aerosol molecular backscattering coefficient; (15) In the formula, σ is the solved total atmospheric extinction coefficient, including the atmospheric molecular extinction coefficient and the aerosol molecular extinction coefficient.

[0093] Step 4: Due to the uneven distribution of aerosols along the detection path, there are abrupt changes in the aerosol echo signal data, which affects the inversion accuracy of the aerosol extinction coefficient in subsequent paths. The abrupt change intervals in the echo signal data are determined by the differential comparison method, and the abrupt change starting point r of the abrupt change interval in the echo signal data is obtained. bps1 With mutation endpoint r bpe1 For the same echo signal data, the abrupt change starting point r of the abrupt change interval is obtained by wavelet transform method. bps2 With mutation endpoint r bpe2 .

[0094] The differential comparison method in step four is as follows: calculate the logarithmic distance correction signal and its first-order forward difference value in step three, compare the value with the preset threshold G, and complete the location of the abrupt change interval.

[0095] The specific steps of the difference comparison method are as follows:

[0096] Let r0 be the distance at which the transmitting and receiving fields of view completely overlap, and let r be the starting point of the far-field region. n These are determined by the system's geometric overlap factor and the echo signal signal-to-noise ratio, respectively. The interval (r0, r...) n (r0) represents the effective region of the lidar echo signal data. To determine whether there are abrupt changes within the effective region, the logarithmic distance correction signal and its first-order forward difference value for each sampling point after r0 are calculated.

[0097] The logarithmic distance correction signal is:

[0098] S(r)=ln[P(r)·r 2 (16)

[0099] The first-order forward difference is:

[0100] ΔS i =S(r) i+1 )-S(r i (17)

[0101] The first-order forward difference value is compared with a pre-set threshold G, where G is r i The absolute value of k times the average of the first-order forward differences of the first five points, where G is:

[0102]

[0103] Furthermore, if △S i If ≥G, directly determine the current point r. i This is the starting point of the upward mutation. If 0 < ΔS i <G, to avoid misjudgment or omission, continue to calculate r. iThe first-order forward difference of the subsequent three points and the mean of S(r) at these three points are considered if there are two or more points where the first-order forward difference is greater than zero or the mean of S(r) is greater than the mean of S(r) at the current point. i If the value of r is given, then r i This is also considered the starting point of an upward mutation. Otherwise, the current point is considered not the starting point of an upward mutation, but rather a signal fluctuation caused by noise. If it has been determined that the point is not an upward mutation, then the process continues to determine whether it is the starting point of a downward mutation, using a method similar to that for upward mutations. This process continues until the mutation starting point is found and marked as r. bps If there is no signal change in the current echo signal, the visibility can be directly retrieved using the effective area of ​​the echo signal.

[0104] Furthermore, for the interval (r0, r bps-1 The least squares fit was performed on the data segment, and the mutation initiation r was recorded. bps The corresponding fitted value S(r) bps Then, the values ​​of S(r) at subsequent points are compared with S(r). bps The comparison continues until the two points are similar and satisfy the condition of a decreasing trend under an upward mutation condition or an upward trend under a downward mutation condition. The mutation endpoint r is then determined. bpe .

[0105] The wavelet transform method in step four is as follows: wavelet decomposition is performed on the aerosol echo signal data processed in step three to obtain wavelet coefficients, threshold processing is performed on the wavelet coefficients, and the processed wavelet coefficients are reconstructed to obtain the reconstructed signal data. By calculating the error between the original signal data and the reconstructed signal data, the abrupt change interval is located.

[0106] The specific steps of wavelet transform are as follows:

[0107] First, wavelet decomposition is performed on the signal to obtain a series of wavelet coefficients. Thresholding is then applied to the obtained wavelet coefficients, setting smaller wavelet coefficients to 0 and retaining larger ones. The processed wavelet coefficients are then reconstructed to obtain the reconstructed signal. By calculating the error between the original signal and the reconstructed signal, the location of abrupt change points is found. These abrupt change points typically correspond to larger error regions in the signal. Based on the magnitude and location of the error, the abrupt change points are located and analyzed to determine the specific abrupt change interval.

[0108] In step four, the difference comparison method and the wavelet transform method each have their own advantages and disadvantages, resulting in different abrupt change intervals (r). bps1 ,r bpe1 ) and (r bps2 ,r bpe2 The results may differ, so the intersection method is used to extract the mutation intervals jointly determined by the two algorithms, thereby reducing the false positive rate and improving the accuracy and reliability of signal mutation interval determination.

[0109] Step 5: Using the abrupt change intervals determined in Step 4, determine new reference points for the aerosol echo signal data within adjacent abrupt change intervals, and use the Collis-Fernald iterative algorithm to stabilize the boundary value of the extinction coefficient at the new reference point.

[0110] Since the aerosol echo signal data between adjacent abrupt change intervals is relatively flat, the aerosol echo signal data between adjacent abrupt change intervals can be equivalent to a uniform distribution, so the extinction coefficient boundary value σ at the new reference point can be determined by the Collis method in step three.

[0111] The new reference point is located between adjacent mutation regions via P(r)r 2 / β m (r) Determined.

[0112] Substituting the extinction coefficient boundary value σ at the new reference point into Fernald's integral formula, we obtain the average extinction coefficient σ within the distance of adjacent abrupt change intervals. mean Then determine σ and σ mean The size of the extinction coefficient. If the boundary value σ of the extinction coefficient at the new reference point is greater than the average value σ of the extinction coefficient over that distance. mean , will σ mean As the boundary value of the extinction coefficient, it is substituted into Fernald's backward integral formula for iteration; if the boundary value σ of the extinction coefficient at the new reference point is less than the average value σ of the extinction coefficient within that distance... mean Then stop the iteration and set σ mean This serves as the boundary value for the extinction coefficient at the final new reference point.

[0113] Because the Fernald backward integration formula can handle longer pulse widths and is less susceptible to noise and interference compared to the Fernald forward integration formula, the Fernald backward integration formula is used in step five to invert the extinction coefficient, thereby improving the accuracy and reliability of the aerosol extinction coefficient inversion.

[0114] Fernald's backward integration formula is:

[0115]

[0116] In equation (19), r c For reference distance, S a =α a (r) / β a (r), S m =α m (r) / β m (r), where α m(r) represents the extinction coefficient of atmospheric molecules. The density of air molecules is obtained using the standard atmospheric model based on temperature, humidity, and pressure, and then calculated using the Rayleigh scattering theory. The boundary value α of the extinction coefficient of aerosol molecules is also given. a (r) is determined by the aerosol molecule scattering ratio 1+β a (r c ) / β m (r c To determine.

[0117] Among them, S a It takes values ​​in the range of 20 to 100 Sr.

[0118] Among them, S a The value is 50.

[0119] Among them, S m The value is 8π / 3.

[0120] Step 6: Substitute the initial reference point and the boundary value of the extinction coefficient at the initial reference point obtained in Step 3 into the Fernald backward integral formula to invert the aerosol extinction coefficient value over the distance from the initial reference point to the starting point of the detection path.

[0121] Step 7: After completing the initial backward integration, determine if there are any abrupt change intervals after the initial reference point. If so, take the first new reference point between adjacent abrupt change intervals from Step 5 as the starting point. Substitute the extinction coefficient boundary value at the new reference point into the Fernald backward integration formula to invert the extinction coefficient along the subsequent path up to the starting point of the detection path. Repeat this process until there are no abrupt change intervals. By repeating this operation, the error in aerosol extinction coefficient inversion due to abrupt change intervals is reduced, thereby improving the detection accuracy and reliability of horizontal aerosol extinction coefficient inversion.

[0122] Step 8: Combine the initial aerosol extinction coefficient data obtained in Step 6 with the new aerosol extinction coefficient data obtained in Step 7, and continue this process until the combination is complete.

[0123] Step 9: Obtain the one-dimensional aerosol extinction coefficient value in the horizontal detection direction of the single-photon lidar through Step 8. Based on the lidar scanning speed and angle, convert the one-dimensional aerosol extinction coefficient value into a two-dimensional aerosol extinction coefficient profile, thus realizing the inversion of the horizontal aerosol extinction coefficient.

[0124] Example 2:

[0125] like Figure 3 As shown, a single-photon lidar horizontal aerosol extinction coefficient detection system device includes a transmitting module, a receiving module, a storage module, and a horizontal gimbal module.

[0126] Furthermore, the transmitting module includes a pulsed laser transmitter that emits pulsed laser signals, which are then directed towards the direction of the test via a beam expander and a half-wave plate. The receiving module includes a receiving telescope, a collimating lens, a narrowband filter, a polarizing beam splitter, and a converging lens. The returned backscattered signal is adjusted to a single-direction light signal by the collimating lens, and after the filter removes optical noise, it passes through the polarizing beam splitter and converging lens to enter the storage module as parallel and perpendicular polarized light. The storage module consists of a photodetector, a signal amplifier, and a host computer. The optical signal is converted into an electrical signal by the photodetector, amplified, and stored in the host computer for subsequent signal processing. The horizontal pan-tilt module consists of a pan-tilt tripod and a rotating support.

[0127] In this embodiment, the horizontal detection range of the single-photon lidar is greater than 5km, the time required to scan one horizontal circle is less than 9min, the single-line time resolution is 1-2s, the scanning speed is 0.01°~15° / s, the scanning angle error is ±0.1°, and the distance resolution is available in two levels: 15m and 30m. The repetition frequency of the pulsed laser is less than 20kHz, the wavelength is 1550nm, and the energy of a single pulse is less than 35μJ.

[0128] The above detailed description further illustrates the purpose, technical solution, and beneficial effects of the invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar, characterized in that: Includes the following steps, Step 1: Set up a single-photon lidar in the aerosol testing area using a horizontal gimbal. Use the single-photon lidar to detect aerosols in the horizontal direction and obtain aerosol echo signal data in the horizontal direction. Use single-photon lidar to improve the detection range and accuracy in complex and harsh environments. The complex and harsh environment includes rainy and foggy weather, snowy weather, hazy weather, sandstorms and strong sunlight. Step 2: Perform background denoising, geometric overlap factor correction, distance squared correction, and smoothing on the horizontal aerosol echo signal data obtained in Step 1; Step 3: Set an initial reference point near the maximum detection range of the single-photon lidar. Due to the distance constraint, there are almost no photons returning near the maximum detection range. The atmosphere near the maximum distance can be regarded as a clean atmosphere with very little aerosol particles and uniform distribution, so as to use the Collis method to determine the extinction coefficient boundary value at the initial reference point. Step 4: Due to the uneven distribution of aerosols along the detection path, there are abrupt changes in the aerosol echo signal data, which affects the inversion accuracy of the aerosol extinction coefficient in subsequent paths. The abrupt change intervals in the echo signal data are determined by the differential comparison method, and the abrupt change starting point r of the abrupt change interval in the echo signal data is obtained. bps1 With mutation endpoint r bpe1 For the same echo signal data, the abrupt change starting point r of the abrupt change interval is obtained by wavelet transform method. bps2 With mutation endpoint r bpe2 ; Step 5: Using the abrupt change intervals determined in Step 4, determine new reference points for the aerosol echo signal data within adjacent abrupt change intervals, and use the Collis-Fernald iterative algorithm to stabilize the boundary value of the extinction coefficient at the new reference point. Since the aerosol echo signal data between adjacent abrupt change intervals is relatively flat, the aerosol echo signal data between adjacent abrupt change intervals can be equivalent to a uniform distribution, so the extinction coefficient boundary value σ at the new reference point can be determined by the Collis method in step three. Substituting the extinction coefficient boundary value σ at the new reference point into Fernald's integral formula, we obtain the average extinction coefficient σ within the distance of adjacent abrupt change intervals. mean Then determine σ and σ mean The magnitude; if the boundary value σ of the extinction coefficient at the new reference point is greater than the average value σ of the extinction coefficient over that distance. mean , will σ mean As the boundary value of the extinction coefficient, it is substituted into Fernald's backward integral formula for iteration; if the boundary value σ of the extinction coefficient at the new reference point is less than the average value σ of the extinction coefficient within that distance... mean Then stop the iteration and set σ mean As the final new reference point, the boundary value of the extinction coefficient; Step 6: Substitute the initial reference point and the extinction coefficient boundary value at the initial reference point obtained in Step 3 into the Fernald backward integral formula to invert and obtain the aerosol extinction coefficient value over the distance from the initial reference point to the starting point of the detection path. Step 7: After completing the first backward integration, determine whether there is abrupt change interval after the initial reference point. If there is abrupt change interval, take the first new reference point between adjacent abrupt change intervals in Step 5 as the starting point, substitute the extinction coefficient boundary value at the new reference point into the Fernald backward integration formula, and invert to obtain the extinction coefficient of the subsequent path up to the starting point of the detection path from the new reference point. Then, make the judgment and repeat the above process until there is no abrupt change interval. Repeat this operation to reduce the error of abrupt change interval in the inversion of aerosol extinction coefficient, thereby improving the detection accuracy and reliability of horizontal aerosol extinction coefficient inversion. Step 8: Combine the initial aerosol extinction coefficient data obtained in Step 6 with the new aerosol extinction coefficient data obtained in Step 7, and continue this process until the combination is complete. Step 9: Obtain the one-dimensional aerosol extinction coefficient value in the horizontal detection direction of the single-photon lidar through Step 8. Based on the lidar scanning speed and angle, convert the one-dimensional aerosol extinction coefficient value into a two-dimensional aerosol extinction coefficient profile, thus realizing the inversion of the horizontal aerosol extinction coefficient.

2. The method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar as described in claim 1, characterized in that: The lidar described in step one is a superconducting nano-single-photon lidar, comprising a transmitting module, a receiving module, a storage module, and a horizontal gimbal module. The transmitting module includes a pulsed laser that emits pulsed laser signals, which are then directed towards the target direction via a beam expander and a half-wave plate. The receiving module includes a receiving telescope, a collimating lens, a narrowband filter, a polarizing beam splitter, and a converging lens. The returned backscattered signal is adjusted to a single-direction light signal by the collimating lens, and after the filter removes optical noise, it passes through the polarizing beam splitter and the converging lens to enter the storage module as parallel and perpendicular polarized light. The storage module consists of a photodetector, a signal amplifier, and a host computer. The optical signal is converted into an electrical signal by the photodetector, which is horizontal aerosol echo signal data, amplified, and stored in the host computer. The horizontal gimbal module consists of a gimbal tripod and a rotating support.

3. The method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar as described in claim 2, characterized in that: In step one, the single-photon lidar uses a wavelength of 1550nm, which is invisible to the human eye, thus reducing harm to the human eye while ensuring horizontal detection accuracy.

4. The method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar as described in claim 1, 2, or 3, characterized in that: In step two, by selecting a flat signal at the far end of the detection distance, the average value of the flat signal is taken as the background noise. The background noise is subtracted from the aerosol echo signal data to obtain the denoised signal, thereby reducing the interference of radiation from the sun, earth, atmosphere and artificial light sources on the detection accuracy of aerosol echo signals. Because the pulsed laser beam emitted by the pulsed laser interacts with various substances in the atmosphere, the energy attenuates with increasing distance. This energy attenuation is inversely proportional to the square of the distance, causing the signal to attenuate too quickly at long distances, thus affecting subsequent data processing. Therefore, distance square calibration is used to eliminate this effect. The formula for calibrating the squared distance is: X(r)=P(r)·r 2 (1) In equation (1), X(r) is the distance squared correction signal, P(r) is the power signal, and r is the detection distance; When a single-photon lidar is used for measurement, the divergence angle of the laser beam and the receiving field of view are different. The geometric overlap factor refers to the fact that when photons are backscattered, only a portion of the photons will be received, while other photons are limited by the field of view angle and cannot be received, thus producing errors. Therefore, the geometric overlap factor is corrected by experimental methods to eliminate errors and improve the detection accuracy of single-photon lidar.

5. The method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar as described in claim 4, characterized in that: The smoothing method uses the five-point cubic method, and the signal data obtained after smoothing has higher reliability and accuracy. The five-point three-dimensional method is as follows: (2) In the formula, x(t) is the aerosol echo signal data to be smoothed, t=1,2,…,N.

6. The method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar as described in claim 5, characterized in that: In step three, the Collis method specifically works as follows: S(r)=ln[P(r)·r 2 ](3) Substituting equation (3) into equation (4), the extinction coefficient under uniform atmospheric conditions is obtained as follows: (3) In the formula, S is the logarithmic distance correction signal; P is the power of the lidar signal intensity; r is the lidar detection distance; (4) In the formula, E is the lidar fixed parameter, which is a constant; C is the lidar correction constant; β is the total atmospheric backscattering coefficient, including the atmospheric molecular backscattering coefficient and the aerosol molecular backscattering coefficient; (5) In the formula, σ is the solved total atmospheric extinction coefficient, including the atmospheric molecular extinction coefficient and the aerosol molecular extinction coefficient.

7. The method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar as described in claim 6, characterized in that: In step four, The differential comparison method in step four specifically involves: calculating the logarithmic distance correction signal and its first-order forward difference value from step three, comparing this value with a pre-set threshold G, and thus locating the abrupt change interval. The wavelet transform method in step four is as follows: wavelet decomposition is performed on the aerosol echo signal data processed in step three to obtain wavelet coefficients, threshold processing is performed on the wavelet coefficients, and the processed wavelet coefficients are reconstructed to obtain the reconstructed signal data. By calculating the error between the original signal data and the reconstructed signal data, the location of the abrupt change interval is completed. By employing the intersection of the differential comparison method and the wavelet transform method, the mutation interval jointly determined by the two algorithms is extracted, thereby reducing the false judgment rate and improving the accuracy and reliability of signal mutation interval determination.

8. The method for inverting the horizontal aerosol extinction coefficient of a single-photon lidar as described in claim 7, characterized in that: Step five uses the Fernald backward integral formula to invert the extinction coefficient, improving the accuracy and reliability of the inversion of the aerosol extinction coefficient: Fernald's backward integration formula is: (6) In the formula, r c For reference distance, S a =α a (r) / β a (r), S m =α m (r) / β m (r), where α m (r) represents the extinction coefficient of atmospheric molecules. The density of air molecules is obtained using the standard atmospheric model based on temperature, humidity, and pressure, and then calculated using the Rayleigh scattering theory. The boundary value α of the extinction coefficient of aerosol molecules is also given. a (r) is determined by the aerosol molecule scattering ratio 1+β a (r c ) / β m (r c To determine.

Images