Generation method and device of laser atmospheric transmission parameter time sequence based on generalized wavelet moment estimator

Abstract

The invention discloses a generalized wavelet moment estimator-based laser atmospheric transmission parameter time sequence generation method and device. The method comprises the following steps of: determining a model order, acquiring an observation vector sequence # imgabs0 # of laser atmospheric transmission parameters through a parameter measuring instrument, effectively judging vector data, and determining an enabling control operator; a normalized cross-correlation function of a small sample atmospheric parameter observation sequence is subjected to robust analysis by using a generalized wavelet moment estimator, autoregressive moving average model parameters of all atmospheric parameters are obtained through identification and calculation, and the autoregressive moving average model parameters are input into a Kalman filter to synchronously generate a laser atmospheric transmission parameter time sequence. According to the method, the generalized wavelet moment estimator is used for achieving synchronous and continuous generation of the laser atmospheric transmission parameter time sequence, the method can be suitable for small sample machine learning occasions, the generated data has the advantages of meeting time synchronism, continuity, correlation, parameter control and the like, and the method can be used as a pre-training processing method of machine learning.

Description

Technical Field

[0001] The invention belongs to the technical field of laser atmospheric transmission, and in particular relates to a method and a device for generating a time series of laser atmospheric transmission parameters based on a generalized wavelet moment estimator. Background Art

[0002] When the laser is transmitted in the atmosphere, it will be affected by the combined effects of atmospheric attenuation (aerosol scattering, molecular absorption, etc.), atmospheric turbulence (temperature pulsation), atmospheric thermal halo (nonlinear effect) and turbulence-thermal halo interaction. The laser amplitude, phase, and polarization will change with time and space, and the temporal and spatial coherence will decrease. There will be problems such as far-field laser energy attenuation, spot enlargement and deformation, light intensity fluctuation distortion, and power density distribution dispersion, which will lead to reduced laser transmission efficiency, greatly reducing its application in laser ranging, laser communication, laser radar, laser remote sensing, laser weapons, etc., and affecting the further development of laser systems in engineering applications. For the engineering application of laser systems, the key to judging their success is whether the efficient and directional transmission of laser energy can be achieved, and its transmission efficiency needs to be evaluated. Laser atmospheric transmission efficiency evaluation technology has undergone three generations of development since the 1970s. Since 2011, the third-generation laser atmospheric transmission efficiency evaluation technology has been characterized by taking the execution mission planning and actual efficiency evaluation of the laser system as the goal, combining the probabilistic parameters of atmospheric optical characteristics obtained by technologies such as big data analysis, machine learning and intelligent prediction, and taking the laser atmospheric transmission mechanism and laws established by technologies such as physical experiments and numerical simulation as the core to build a laser atmospheric transmission application model that supports decision-making. Therefore, the third-generation laser atmospheric transmission efficiency evaluation technology is also called auxiliary decision-making technology.

[0003] The third generation of laser atmospheric transmission efficiency evaluation technology uses machine learning methods to evaluate the efficiency of laser atmospheric transmission. It is not necessary to consider the specific details of the atmospheric effects in the middle of the transmission on the laser transmission. Its execution speed, prediction accuracy, versatility and other performance can be maximized. Since the internal architecture of the machine learning method is an adaptive bionic structure, it can directly obtain the multivariate regression relationship between the input (atmospheric parameters) and the output (efficiency evaluation). It requires large-capacity, high-information (independent and identically distributed) measured data (input atmospheric parameters and output efficiency evaluation parameters) to train and optimize the internal structure framework and internal parameters of the machine learning. Limited by factors such as site security, line of sight conditions, personnel safety and continuous working time of the laser, it is difficult to obtain stable, synchronous, continuous and full-time and space test training data for laser atmospheric transmission field tests. There is a problem of small sample machine learning with insufficient effective training sample capacity. Therefore, there are few existing machine learning research results, and the technology and application are still immature.

[0004] In response to the small sample problem of machine learning methods in laser atmospheric transmission, the present invention proposes a method and device for generating time series of laser atmospheric transmission parameters based on a generalized wavelet moment estimator. This method can better solve the problem of insufficient effective training sample capacity of small sample machine learning methods in laser atmospheric transmission efficiency evaluation technology, and provide theoretical guidance and technical support for the successful application of machine learning in laser atmospheric transmission efficiency evaluation. Summary of the invention

[0005] In order to overcome the deficiencies in the prior art, the present invention provides a method and device for generating a time series of laser atmospheric transmission parameters based on a generalized wavelet moment estimator, so as to achieve the purpose that the generated atmospheric transmission parameter time series has time synchronization, continuity, correlation, and parameter control, and provide excellent pre-training data for the development of machine learning algorithms.

[0006] The technical solution to achieve the purpose of the present invention is:

[0007] A method for generating time series of laser atmospheric transmission parameters based on generalized wavelet moment estimator, comprising:

[0008] Step 1: Collect the observation vector sequence of laser atmospheric transmission parameters through parameter measurement instruments And make effective judgment on the vector data to determine the enabling control operator;

[0009] Step 2: Based on the observation vector sequence The generalized wavelet moment estimation method is used to identify the parameters of the autoregressive moving average model and obtain the model parameters;

[0010] Step 3, based on the model parameters and the enabled control operator, obtain the Kalman filter state estimation value at the synchronization time k through the Kalman filter algorithm;

[0011] Step 4: cache the obtained data and repeat the above steps to obtain training sample data based on time series for evaluating the laser atmospheric transmission efficiency.

[0012] Furthermore, in step 1, the vector data is effectively judged, and determining the enabling control operator specifically includes: if data is collected at the current time k, the enabling control operator δ=1, otherwise δ=0.

[0013] Furthermore, the observation vector sequence Including temperature T, humidity RH, air pressure P, wind speed mph, visibility V, atmospheric coherence length R and refractive index structure constant

[0014] Furthermore, the step 2 specifically includes:

[0015] Analyze the statistical characteristics of the time evolution of various atmospheric parameters and determine the model orders p, d and q;

[0016] Construct the time series model of each atmospheric parameter - the difference autoregressive moving average model paradigm;

[0017] Using generalized wavelet moment estimation method combined with observation vector sequence Robustly analyze the normalized cross-correlation function of each atmospheric parameter, remove outliers, identify and calculate the differential autoregressive moving average model parameters θ(B) of each atmospheric parameter, and convert them into a state transfer matrix Φ(B) and a noise vector

[0018] Furthermore, determining the model orders p, d and q specifically includes: calculating robust autocorrelation / partial autocorrelation functions based on the time evolution statistical characteristics of each atmospheric parameter, and determining the model orders p, d and q in combination with AIC, BIC or HQ.

[0019] Furthermore, the differential autoregressive moving average model paradigm is:

[0020] Φ(B)δ d X t =θ(B)W t

[0021] Among them, B is a one-step time delay operator, Φ(B) is a regression operator, and δ d =(1-B) d is a d-order difference operator, θ(B) is a moving average operator, and (p, d, q) are the autoregressive order, the difference order, and the moving average order, respectively.

[0022] Furthermore, the Kalman filter state estimation value at the synchronization time k is obtained through the Kalman filter algorithm, which specifically includes:

[0023] Step 3-1, determine the state form of the difference autoregressive moving average model as:

[0024]

[0025] in, is the observation vector sequence The state vector of is a Gaussian white noise vector, C Φ is the mean square error of noise intensity;

[0026] Step 3-2, determine the observation equation of the one-step Kalman filter as:

[0027]

[0028] in, is the measurement vector at the kth moment, δ is the enabling control operator, if δ=1, execute step 4, if δ=0, execute step 3-3;

[0029] Step 3-3, solve and obtain the synchronous and continuous time series of laser atmospheric transmission parameters, and obtain the estimated value at time k.

[0030] Furthermore, the solution to obtain a synchronous and continuous laser atmosphere transmission parameter time series specifically includes:

[0031] P(k,k-1)=Φ(B)P(k-1,k-1)Φ T (B)+W(k)

[0032] K(k)=P(k,k-1)H T (k)[H(k)P(k,k-1)H T (k)+R(k)] -1

[0033]

[0034] P(k,k)=[I-δ·K(k)H(k)]P(k,k-1)

[0035] Among them, P(k, k-1) is the one-step state transfer covariance matrix at time k, K(k) is the gain matrix, P(k, k) is the estimated covariance at time k, and H(k) is the measurement moment under k constraint at time k.

[0036] Furthermore, the measurement matrix is:

[0037]

[0038] A laser atmospheric transmission parameter time series generation device based on a generalized wavelet moment estimator, comprising:

[0039] The acquisition unit collects the observation vector sequence of the laser atmospheric transmission parameters through the parameter measurement instrument And make effective judgment on the vector data to determine the enabling control operator;

[0040] Generalized wavelet moment estimation parameter estimator, input is the observation vector sequence The generalized wavelet moment estimation method is used to identify the parameters of the autoregressive moving average model and obtain the model parameters;

[0041] The Kalman filter, based on the enabling control operator, obtains the Kalman filter state estimation value at the synchronization time k through the Kalman filter algorithm.

[0042] Compared with the prior art, the present invention has the following advantages:

[0043] (1) The present invention introduces a method for generating a time series of laser atmospheric transmission parameters based on a generalized wavelet moment estimator, which can synchronously and continuously generate a time series of laser atmospheric transmission parameters, expand the capacity of pre-training samples required for machine learning, and provide technical support for the engineering application of machine learning in the evaluation of laser atmospheric transmission performance;

[0044] (2) The present invention can be applied to small sample machine learning scenarios. Its small sample time series data generation algorithm can provide a solution reference for solving the problem of insufficient capacity of small sample machine learning. The generated data meets the advantages of time synchronization, continuity, correlation, parameter control, etc., and can be used as a pre-training processing method for machine learning. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a diagram of the measurement device configuration for laser atmospheric transmission parameters.

[0046] Figure 2 It is the algorithm flow chart of the atmospheric parameter time series correlation model based on generalized wavelet moment estimation.

[0047] Figure 3 This is a flow chart of the algorithm for generating time series of laser atmospheric transmission parameters based on Kalman filter. DETAILED DESCRIPTION

[0048] In order to more fully understand the technical content of the present invention, the technical solution of the present invention is further introduced and illustrated in conjunction with specific embodiments below, but is not limited thereto.

[0049] This embodiment provides a laser atmospheric transmission parameter time series generation device based on a generalized wavelet moment estimator, comprising:

[0050] The acquisition unit collects the observation vector sequence of the laser atmospheric transmission parameters through the parameter measurement instrument And make effective judgment on the vector data to determine the enabling control operator;

[0051] Generalized wavelet moment estimation parameter estimator, input is the observation vector sequence The generalized wavelet moment estimation method is used to identify the parameters of the autoregressive moving average model and obtain the model parameters;

[0052] The Kalman filter, based on the enabling control operator, obtains the Kalman filter state estimation value at the synchronization time k through the Kalman filter algorithm.

[0053] This embodiment also provides a method for generating a time series of laser atmospheric transmission parameters based on the device, comprising the following steps:

[0054] Step 1: Figure 1 As shown, the observation vector sequence of the laser atmospheric transmission parameters collected for a small sample is Contains temperature T, humidity RH, air pressure P, wind speed mph measured by the automatic weather station, visibility V measured by the visibility meter, atmospheric coherence length R measured by the atmospheric coherence length meter, and refractive index structure constant measured by the temperature pulsator

[0055] Step 2: Figure 2 As shown in the figure, the generalized wavelet moment estimation algorithm for atmospheric parameter time series correlation model first uses the generalized wavelet moment estimation method to analyze the time evolution statistical characteristics of each small sample atmospheric parameter, and calculates rACF / rPACF (robust autocorrelation / partial autocorrelation function). The model order (p, d, q) is determined by comprehensive judgment based on traditional criteria (AIC, BIC or HQ). Then, the time series model of each atmospheric parameter is constructed separately - ARIMA (differential autoregressive moving average) model paradigm. The generalized wavelet moment estimation parameter estimator is combined with the small sample observation sequence. Robustly analyze the normalized cross-correlation function of each atmospheric parameter, remove outliers, identify and calculate the ARMA (autoregressive moving average) model parameters θ(B) of each atmospheric parameter, and convert them into a state transfer matrix Φ(B) and a noise vector Finally, it is input into the Kalman real-time filter and the filtering parameters are initialized.

[0056] The general expression of GMWM (generalized wavelet moment estimation) is:

[0057]

[0058] Among them, θ is the true model parameter vector to be identified; is the parameter vector of the computational model under GLS; ψ(θ) is the wavelet variance vector of the identified data; is the wavelet variance vector of the model, which contains the parameter variables to be determined; Ω is a positive definite weight matrix, which is related to the data length.

[0059] The process of identifying the parameters of the time series related model of atmospheric parameter calibration law is as follows:

[0060] (2-1) Determination of model order (p, d, q)

[0061] The classic ACF / PACF (robust autocorrelation / partial autocorrelation function) is an important step in time series analysis, but it is only effective for stationary signals and has high requirements for outliers and data preprocessing. By performing rACF / rPACF comparative analysis on a small sample atmospheric parameter data set, the confidence intervals of each linear correlation term in the time series model can be explicitly expressed, outliers can be removed, and the small sample original data set can be preprocessed, and then the model order (p, d, q) of each atmospheric parameter ARIMA (p, d, q) can be determined.

[0062] (2-2) ARIMA (p, d, q) model parameter identification ARIMA (p, d, q) is a general standard form of non-seasonal time series model, expressed as:

[0063] Φ(B)δ d X t =θ(B)W t (2)

[0064] Among them, B is a one-step time delay operator, Φ(B) is a regression operator, and δ d =(1-B) d is a d-order difference operator, θ(B) is a moving average operator, and (p, d, q) are the autoregressive order, difference order, and moving average order corresponding to the regression operator, difference operator, and moving average operator, respectively.

[0065] The accuracy of the parameter identification Φ(B) and θ(B) of the above ARIMA (p, d, q) model is the key to ensure a good fit of the calibration function. Before parameter identification, the model order (p, d, q) needs to be determined in advance. A single traditional criterion (AIC, BIC or HQ) is difficult to effectively distinguish various statistical situations. GMWM provides a comprehensive comparative analysis tool of AIC / BIC / HQ and rACF / rPACF, which can find the simplest and most effective model order. The conventional solution methods in parameter identification are maximum likelihood least squares (ML) and correlation function method, but both lack robustness. GMWM uses the robust GMWM method to solve the model parameters, and the parameters have the advantages of stability and evaluable confidence intervals. This step finally obtains a synchronous, continuous, and correlated atmospheric parameter time series correlation model, which provides filter state parameters for the subsequent Kalman real-time filter to generate a synchronous, continuous, and correlated laser atmospheric transmission parameter time series.

[0066] Step 3: Figure 3 As shown, the laser atmospheric transmission parameter time series is generated based on the Kalman filter algorithm, including:

[0067] Step 3-1: The Kalman filter-based laser atmospheric transmission parameter time series generation algorithm first performs a small sample observation sequence Perform queue caching and make effective judgment on vector data, that is, if the above atmospheric parameter measurement instrument outputs real-time data at the current time k, then δ = 1, otherwise δ = 0;

[0068] Step 3-2, then the queue buffer outputs a length of N synchronized and continuous laser atmospheric transmission parameter time series vector Send it to the generalized wavelet moment estimation parameter estimator for model parameter identification, and the identification process is the same as step 2;

[0069] Step 3-3 Kalman filter is used to enable the control operator Determine whether the model parameters Φ(B) and Perform parameter estimation; specifically include:

[0070] A Kalman filter is constructed with a small sample atmospheric parameter data set as the observation equation and the time series correlation model ARIMA (p, d, q) of atmospheric parameters as the state equation.

[0071] Kalman filtering is a cyclically executed algorithm. The filter consists of a state equation and an observation equation. The state transfer equation (i.e., the difference equation) consists of temperature T, humidity RH, air pressure P, wind speed mph, visibility V, refractive index structure constant The atmospheric coherence length R constitutes the state vector of the laser atmospheric transmission parameter vector sequence The model form is shown in formula (2), and the state transition form is:

[0072]

[0073] Among them, φ(B) is the matrix composed of the ARIMA (p, d, q) model parameters of the laser atmospheric transmission parameters; is a Gaussian white noise vector, C Φ is the mean square error of the noise intensity.

[0074] The observation equation of the one-step Kalman filter is:

[0075]

[0076] in, is the measurement vector at the kth moment, δ is the enabling control operator of the small sample atmospheric parameter data set, defined as The validity of the data here is determined based on whether the data is measured and output in real time.

[0077] The synchronous and continuous time series of laser atmospheric transmission parameters are calculated based on the real-time input of small sample atmospheric parameter data set (or real-time input of field test), which are as follows:

[0078] P(k,k-1)=φ(B)P(k-1,k-1)φ T (B)+W(k) (5)

[0079] K(k)=P(k,k-1)H T (k)[H(k)P(k,k-1)H T (k)+R(k)] -1 (6)

[0080]

[0081] P(k,k)=[I-δ·K(k)H(k)]P(k,k-1) (8)

[0082] Among them, P(k, k-1) is the one-step state transfer covariance matrix at time k, K(k) is the gain matrix, P(k, k) is the estimated covariance at time k, and H(k) is the measurement matrix under the constraint of the effective observation data δ at time k, which is defined as

[0083] The Kalman filter state estimate at synchronization time k can be obtained by equations (5)-(8): (including temperature T, humidity RH, air pressure P, wind speed mph, visibility V, refractive index structure constant Atmospheric coherence length R).

[0084] The method is used to generate a time series of laser atmosphere transmission parameters as training data for machine learning in laser atmosphere transmission efficiency evaluation technology, so that machine learning can obtain better learning effects and improve the evaluation effect of laser atmosphere transmission efficiency in laser systems.

[0085] In summary, the present invention introduces a method for generating time series of laser atmospheric transmission parameters based on a generalized wavelet moment estimator, which can synchronously and continuously generate time series of laser atmospheric transmission parameters, expand the pre-training sample capacity required for machine learning, and provide technical support for the engineering application of machine learning in laser atmospheric transmission efficiency evaluation; the present invention can be applied to small sample machine learning occasions, and its small sample time series data generation algorithm can provide a solution reference for solving the problem of insufficient capacity of small sample machine learning. The generated data meets the advantages of time synchronization, continuity, correlation, parameter control, etc., and can be used as a pre-training processing method for machine learning.

[0086] The embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0087] The embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0088] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0089] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0090] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0091] The above is only a preferred embodiment of the present invention. Without departing from the technical principle of the present invention, several improvements and modifications can be made, and these improvements and modifications should also be regarded as the protection scope of the present invention.

Claims

1. A method for generating time series of laser atmospheric transmission parameters based on generalized wavelet moment estimator, characterized in that: include: Step 1: Collect the observation vector sequence of laser atmospheric transmission parameters through parameter measurement instruments And make effective judgment on the vector data to determine the enabling control operator; Step 2: Based on the observation vector sequence The generalized wavelet moment estimation method is used to identify the parameters of the autoregressive moving average model and obtain the model parameters; Step 3, based on the model parameters and the enabled control operator, obtain the Kalman filter state estimation value at the synchronization time k through the Kalman filter algorithm; Step 4: cache the obtained data and repeat the above steps to obtain training sample data based on time series for evaluating the laser atmospheric transmission efficiency.

2. According to claim 1, a method for generating time series of laser atmospheric transmission parameters based on generalized wavelet moment estimator is characterized in that: In step 1, the vector data is effectively judged, and the enabling control operator is determined, which specifically includes: if data is collected at the current time k, the enabling control operator δ=1, otherwise δ=0.

3. The method for generating time series of laser atmospheric transmission parameters based on generalized wavelet moment estimator according to claim 1, characterized in that: The observation vector sequence Including temperature T, humidity RH, air pressure P, wind speed mph, visibility V, atmospheric coherence length R and refractive index structure constant 4. The method for generating time series of laser atmospheric transmission parameters based on generalized wavelet moment estimator according to claim 1 is characterized in that: The step 2 specifically includes: Analyze the statistical characteristics of the time evolution of various atmospheric parameters and determine the model orders p, d and q; Construct the time series model of each atmospheric parameter - the difference autoregressive moving average model paradigm; Using generalized wavelet moment estimation method combined with observation vector sequence Robustly analyze the normalized cross-correlation function of each atmospheric parameter, remove outliers, identify and calculate the differential autoregressive moving average model parameters θ(B) of each atmospheric parameter, and convert them into a state transfer matrix Φ(B) and a noise vector 5. The method for generating time series of laser atmospheric transmission parameters based on generalized wavelet moment estimator according to claim 4 is characterized in that: Determining the model orders p, d and q specifically includes: calculating the robust autocorrelation / partial autocorrelation function based on the time evolution statistical characteristics of each atmospheric parameter, and determining the model orders p, d and q in combination with AIC, BIC or HQ.

6. The method for generating time series of laser atmospheric transmission parameters based on generalized wavelet moment estimator according to claim 4, characterized in that: The difference autoregressive moving average model paradigm is: Φ(B)δ d X t =θ(B)W t Among them, B is a one-step time delay operator, Φ(B) is a regression operator, and δ d =(1-B) d is a d-order difference operator, θ(B) is a moving average operator, and (p, d, q) are the autoregressive order, the difference order, and the moving average order, respectively.

7. The method for generating time series of laser atmospheric transmission parameters based on generalized wavelet moment estimator according to claim 2, characterized in that: The Kalman filter state estimation value at the synchronization time k is obtained through the Kalman filter algorithm, which specifically includes: Step 3-1, determine the state form of the difference autoregressive moving average model as: in, is the observation vector sequence The state vector of is a Gaussian white noise vector, C φ is the mean square error of noise intensity; Step 3-2, determine the observation equation of the one-step Kalman filter as: in, is the measurement vector at the kth moment, δ is the enabling control operator, if δ=1, execute step 4, if δ=0, execute step 3-3; Step 3-3, solve and obtain the synchronous and continuous time series of laser atmospheric transmission parameters, and obtain the estimated value at time k.

8. The method for generating time series of laser atmospheric transmission parameters based on generalized wavelet moment estimator according to claim 7 is characterized in that: The synchronous and continuous time series of laser atmospheric transmission parameters obtained by solving the above-mentioned calculation specifically includes: P(k,k-1)=Φ(B)P(k-1,k-1)ΦT(B)+W(k) K(k)=P(k,k-1)H T (k)[H(k)P(k,k-1)H T (k)+R(k)] -1 P(k,k)=[I-δ·K(k)H(k)]P(k,k-1) Among them, P(k, k-1) is the one-step state transfer covariance matrix at time k, K(k) is the gain matrix, P(k, k) is the estimated covariance at time k, and H(k) is the measurement moment under the δ constraint at time k.

9. The method for generating time series of laser atmospheric transmission parameters based on generalized wavelet moment estimator according to claim 8, characterized in that: The measurement matrix is:

10. A device for generating time series of laser atmospheric transmission parameters based on a generalized wavelet moment estimator for implementing any of the methods described in claims 1-9, characterized in that: include: The acquisition unit collects the observation vector sequence of the laser atmospheric transmission parameters through the parameter measurement instrument And make effective judgment on the vector data to determine the enabling control operator; Generalized wavelet moment estimation parameter estimator, input is the observation vector sequence The generalized wavelet moment estimation method is used to identify the parameters of the autoregressive moving average model and obtain the model parameters; The Kalman filter, based on the enabling control operator, obtains the Kalman filter state estimation value at the synchronization time k through the Kalman filter algorithm.

Images