Space target light micro-change calculation amplification algorithm

Abstract

The present application relates to a kind of space target luminosity micro change calculation amplification method, provide support for the space target characteristic parameter based on luminosity change inversion.The present application uses the mode of " first fitting, then extraction", regression analysis noisy luminosity micro change signal, obtain the regression curve of micro change signal;Subtract the regression signal of fitting from noisy luminosity micro change signal, obtain the luminosity micro change signal of mean removal, simply referred to as difference signal, the difference signal includes effective target luminosity micro change signal, also includes noise.Luminosity micro change extraction is converted into extracting effective micro change signal from difference signal;Based on the threshold value judgment rule of different modal component signal signal-to-noise ratio and permutation entropy estimated after decomposition, effective micro change signal is constructed from the decomposition of difference signal, effective micro change signal is added to fitting signal, based on the setting amplification multiple of addition result signal mean deviation value, output is luminosity micro change calculation amplification result.After luminosity micro change signal amplification, the mean of signal does not change, signal-to-noise ratio improves, highlights effective micro change signal feature, provides support for subsequent luminosity curve analysis.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of spatial target luminosity characteristic analysis, in particular to a spatial target luminosity micro-change calculation amplification algorithm. BACKGROUND

[0002] The luminosity change of a spatial target is related to the relative position (described by phase angle) of the sun-target-observation station, the state of the target, the pose, shape and surface material of the target and other characteristics. Therefore, perceiving the luminosity change of a spatial target to invert the related characteristics of a non-cooperative spatial target is one of the important supporting technologies of a spatial target monitoring system. However, the uncertainty factors such as detection conditions, target pose and target state make the luminosity change signal of a non-cooperative spatial target captured by the monitoring system mostly a micro-change signal. The micro-change signal leads to low estimation accuracy of the related characteristics of a spatial target on one hand, and may cause the perception system to be unable to work in time and effectively, reducing the timeliness of the system on the other hand. How to effectively perceive the luminosity change information of a spatial target plays an important role in accurately inverting the related characteristics of the target, and is also a difficult problem in the direction of spatial target detection.

[0003] In reality, the micro-change signal contains certain noise interference, and direct amplification of the signal will cause the noise signal to be amplified together. Therefore, a specific algorithm needs to be designed to extract the effective micro-change. At the same time, the actual luminosity curve is usually not sampled at equal intervals with respect to the phase angle as the independent variable due to the reality factors such as observation cost and observation conditions, so it is difficult to perform time-frequency analysis and other operations. SUMMARY

[0004] (I) Technical problems solved

[0005] In view of the problems existing in the above-mentioned practical application, the present application provides a spatial target luminosity micro-change calculation amplification algorithm. For an input luminosity micro-change signal containing noise, the algorithm amplifies the signal after fitting the overall trend of the curve and extracting the effective micro-change part, calculates the amplification of the luminosity micro-change curve while keeping the overall mean of the luminosity unchanged, and makes the amplified luminosity curve have a higher signal-to-noise ratio with respect to the original micro-change signal, that is, retains more effective features.

[0006] (II) Technical solutions

[0007] The present application provides a spatial target luminosity micro-change calculation amplification algorithm, which includes six main links of luminosity curve trend overall fitting, equal interval sampling, ensemble empirical mode decomposition (EEMD) of difference signal, extraction of effective micro-change, wavelet denoising processing and amplification according to the user's preset amplification multiple. The specific steps are as follows:

[0008] S1. Input the photometric signal, fit the photometric signal curve using the Fourier series of appropriate order, and obtain the overall trend signal of the photometric curve;

[0009] S2. Equidistantly sample the input photometric micro-variation signal and the fitted signal at an appropriate phase angle sampling rate, subtract the fitted signal from the sampled micro-variation signal to obtain an equidistantly sampled difference signal, and set the difference signal to 0 at the original non-existent data phase (indicating that no difference estimation is performed on the non-existent data);

[0010] S3. Decompose the difference signal based on the EEMD method to obtain the intrinsic mode function components (IMF) of the difference signal, denoted as IMFs;

[0011] S4. Among the series of EIMF components, if the signal-to-noise ratio is higher than the preset signal-to-noise ratio threshold, i.e., the noise is less, then it is directly considered as part of the effective micro-variation and added;

[0012] S5. If the signal-to-noise ratio is lower than the preset signal-to-noise ratio threshold, then determine whether the permutation entropy of the component signal is greater than the preset permutation entropy threshold. The EIMF component signal less than the threshold is added as an effective micro-variation to the component with a signal-to-noise ratio higher than the preset signal-to-noise ratio threshold as an effective photometric micro-variation signal. The IMF component greater than the permutation entropy threshold is further subjected to empirical mode decomposition (EMD) to obtain an empirical intrinsic mode function component, denoted as IMF;

[0013] S6. Determine whether the permutation entropy of the IMF component signal is greater than the preset permutation entropy threshold. The component less than the threshold is added to the effective micro-variation;

[0014] S7. The permutation entropy of the IMF component signal is greater than the preset permutation entropy threshold, and the IMF component is subjected to wavelet denoising processing and added to the sequence of effective micro-variation signal components after denoising;

[0015] S8. Calculate the amplification, add the fitted trend signal of step S1 to the effective micro-variation signal, and obtain the micro-variation calculation amplified photometric signal based on the preset amplification multiple.

[0016] The above technical solution has the following beneficial effects:

[0017] 1. The algorithm framework of "fitting first and then extracting" ensures that the overall trend characteristics of the photometric curve are retained, and the effective micro-variation signal is gradually extracted from the difference signal according to the signal-to-noise ratio, permutation entropy, and other criteria from coarse to fine, which can effectively remove the noise signal part and steadily improve the signal-to-noise ratio of the micro-variation amplified signal.

[0018] 2. The problem that non-equidistance sampling data cannot be analyzed in time-frequency domain is solved, so that modal decomposition can be carried out.

[0019] 3. EEMD decomposition utilizes the uniform filling characteristics of white noise in time domain and frequency domain, through adding white noise with a preset signal-to-noise ratio to the input signal and multiple averaging, the IMFs obtained by decomposition can have less modal aliasing phenomenon, that is, each IMF is decomposed into each modal component with actual physical meaning.

[0020] 4. The mean value of the original light micro-change signal remains unchanged after calculation and amplification. BRIEF DESCRIPTION OF DRAWINGS

[0021] The drawings constituting a part of the specification of the present application are used to provide further understanding of the present application.

[0022] Figure 1 is a whole algorithm flowchart of the present application;

[0023] Figure 2 is a specific flowchart of extracting the effective micro-change part in the whole algorithm flowchart;

[0024] Figure 3 is an input original micro-change signal diagram;

[0025] Figure 4 is a fitting signal diagram;

[0026] Figure 5 is a difference signal diagram;

[0027] Figure 6 is an IMF signal diagram after EEMD decomposition;

[0028] Figure 7 is a diagram of preliminarily extracting the effective micro-change signal by using signal-to-noise ratio;

[0029] Figure 8 is an effective micro-change signal result diagram;

[0030] Figure 9 is an output micro-change amplified signal diagram. DETAILED DESCRIPTION

[0031] The following detailed description is exemplary and is intended to provide further explanation of the present application, to make a clear and complete description of the technical solutions in the embodiments of the present application, and all technical and scientific terms used herein have the same meaning as that generally understood by ordinary skilled persons in the technical field to which the present application belongs.

[0032] In the present embodiment, as shown in Figure 1 , it is a whole flowchart of a space target light micro-change calculation amplification algorithm proposed by the present application, which comprises the following steps:

[0033] S1. Input the photometric signal, use the appropriate order Fourier series to fit the photometric signal curve, and obtain the overall trend signal of the photometric curve;

[0034] S2. Equidistantly sample the input photometric micro-variation signal and the fitting signal at an appropriate phase angle sampling rate, and subtract the fitting signal from the sampled micro-variation signal to obtain an equidistantly sampled difference signal;

[0035] S3. Decompose the difference signal based on the EEMD method to obtain the intrinsic mode function component (IMF) of the difference signal, denoted as IMFs;

[0036] S4. Among the series of EIMF components, if the signal-to-noise ratio is higher than the preset signal-to-noise ratio threshold, that is, the noise is less, then it is directly considered as part of the effective micro-variation and added;

[0037] S5. If the signal-to-noise ratio is lower than the preset signal-to-noise ratio threshold, then judge whether the permutation entropy of the component signal is greater than the preset permutation entropy threshold. The EIMF component signal less than the threshold is added as the effective micro-variation, and the component with a signal-to-noise ratio higher than the preset signal-to-noise ratio threshold is added as the effective photometric micro-variation signal. The IMF component greater than the permutation entropy threshold is further subjected to empirical mode decomposition (EMD) to obtain the empirical intrinsic mode function component, denoted as IMF;

[0038] S6. Judge whether the permutation entropy of the IMF component signal is greater than the preset permutation entropy threshold. The component less than the threshold is added to the effective micro-variation;

[0039] S7. The permutation entropy of the IMF component signal is greater than the preset permutation entropy threshold, and the IMF component is subjected to wavelet denoising processing. After denoising, it is added to the sequence of effective micro-variation signal components;

[0040] S8. Calculate the amplification, add the fitting trend signal of step S1 to the effective micro-variation signal, and obtain the micro-variation calculation amplified photometric signal based on the preset amplification multiple.

[0041] In the above scheme, the input of step S1 is the original spatial target photometric variation signal, that is, the photometric micro-variation signal containing noise and non-equidistant sampling. First, the fitting step is performed, and the Fourier series fitting of the original signal obtains the fitting photometric trend signal curve. The order of the Fourier series is usually selected to be 4-6 orders. Then, the step of extracting the effective photometric micro-variation signal is entered, and the original photometric signal and the fitting trend photometric signal are used as inputs to extract the effective micro-variation. The specific process is shown in the attached figure. Figure 2The effective micro-variation signal is added to the equally-spaced sampling fitting signal, the mean of the added signal is calculated, the offset of the mean is amplified, and the amplified signal is finally output as the photometric micro-variation amplification result.

[0042] As shown in the flowchart of the overall algorithm, the specific flowchart of the effective micro-variation extraction part is shown in Figure 2

[0043] In the above scheme, step S2 is trend signal fitting and non-uniform sampling signal resampling. Taking a measured photometric micro-variation signal of a certain high-orbit satellite as an example, the photometric curve is shown in Figure 3

[0044] S21. The original photometric micro-variation signal is fitted by a fourth-order Fourier fitting to obtain an overall trend variation signal, as shown in Figure 4

[0045] S22. The original signal is interpolated by a cubic piecewise Hermite interpolation method to obtain an equally-spaced sampling signal, and the fitted trend signal is sampled at a sampling frequency of 20 Hz. The difference between the resampled original photometric signal and the trend signal is obtained to obtain a difference signal, and the difference signal is taken as 0 at the original non-existent data phase (indicating that the non-existent data is not estimated by difference), as shown in Figure 5

[0046] In the above scheme, step S3 is EEMD method decomposition of the difference signal, which specifically includes the following flow:

[0047] S31. The difference signal is decomposed by EEMD to obtain a series of IMFs, as shown in Figure 6

[0048] S32. The IMFs are sequentially superimposed in the order from bottom to top to obtain a preliminarily constructed effective micro-variation signal, as shown in Figure 7

[0049] ​​​​​​In the above scheme, the specific procedures of steps S4 and S5 are as follows: the IMFs are stacked from bottom to top (component number from large to small) to construct an effective micro-change signal, and if the signal-to-noise ratio of the effective micro-change signal with respect to the difference signal is greater than a preset threshold, the next step is entered, otherwise the current step is repeated to continue stacking the IMF components.

[0050] In the above scheme, the specific procedures of steps S6 and S7 are as follows: the permutation entropy of IMF1 is calculated, if the permutation entropy is less than a threshold (the threshold is 2), the previous effective micro-change is added, otherwise the EMD decomposition is performed to obtain a new IMF component, and the permutation entropy of each IMF component is judged again, if the permutation entropy is less than the threshold, the effective micro-change is added, otherwise the IMF component is processed by wavelet denoising, and the denoised IMF component is added to the effective micro-change signal component sequence. Since IMF1 is the last IMF component obtained by EEMD decomposition, the effective micro-change extraction step is ended (usually all the remaining IMFs should be subjected to the above calculation process of IMF1). The extracted effective micro-change is shown in FIG. 2, which contains less noise with respect to the difference signal. Figure 8

[0051] In the above scheme, the specific procedure of step S8 is as follows: the effective micro-change signal is subjected to cubic piecewise Hermite interpolation according to the independent variable range of the original signal, and is added to the value sampled by the fitting curve at the independent variable range of the original signal. The mean value of the added signal is taken, the added signal is subtracted from the mean value to obtain a bias, the bias is amplified by a preset multiple of 8, and then the mean value is added, to finally obtain a micro-change amplified signal, as shown in FIG. 3. Figure 9

[0052] The above describes the specific embodiments of the present application in combination with the accompanying drawings, but is not a limitation on the protection scope of the present application. All other embodiments obtained by those of ordinary skill in the art without creative labor on the basis of the technical solutions of the present application also belong to the protection scope of the present application.​​

Claims

1. A method for calculating the magnification of small photometric variations of a space object, characterized in that: The input is a non-equidistant sampling signal with noise of a space target luminosity value with slight changes in phase angle or time, and the output is an amplified signal of the valuable slight changes in the luminosity signal. The specific steps are as follows: S1. Input the luminosity signal, use a Fourier series of a proper order to fit the luminosity signal curve, and obtain the overall trend signal of the luminosity curve; S2. Equidistantly sample the input luminosity slight change signal and the fitting signal at a proper phase angle sampling rate, subtract the fitting signal from the sampled slight change signal to obtain an equidistantly sampled difference signal, and set the difference signal at the phase angles where there is no original data to 0 (indicating that no difference estimation is performed on the non-existing data); S3. Based on the ensemble empirical mode decomposition (EEMD) difference signal, obtain the intrinsic mode function (IMF) components of the difference signal, denoted as IMFs; S4. Add the components with a signal-to-noise ratio higher than a preset signal-to-noise ratio threshold ST in the series of IMF components to obtain an effective luminosity slight change signal; calculate the permutation entropy of the components with a signal-to-noise ratio lower than ST, add the IMF components with a permutation entropy lower than a preset permutation entropy threshold PT to the effective luminosity slight change signal, and further perform empirical mode decomposition (EMD) on the IMF components with a permutation entropy higher than PT to obtain empirical intrinsic mode function components, denoted as IMF; S5. Calculate the permutation entropy of the IMF component signal in the S4 step, add the components with a permutation entropy lower than PT to the effective luminosity slight change signal calculated in the S4 step, and perform wavelet denoising on the IMF component with a permutation entropy higher than PT, and add the denoised component to the effective luminosity slight change signal; S6. Add the fitting trend signal in the S1 step to the effective luminosity slight change signal calculated in the S5 step to obtain a luminosity slight change signal to be calculated and amplified, multiply the signal by a preset amplification multiple to obtain the luminosity signal with the slight change calculated and amplified; The step S1 uses a Fourier function to fit the luminosity signal curve to obtain the overall trend signal of the luminosity curve. The step S2 subtracts the fitting signal from the original signal to obtain a difference signal, and the difference signal is equidistantly sampled at a sampling frequency fs. The original signal and the fitting curve are sampled at fs, and the original signal is interpolated at fs. The interpolation method is piecewise cubic Hermite interpolation.

2. A method for calculating the amplification of small photometric variations of a space object according to claim 1, characterized in that, The step S3 specifically includes the following procedures: S31: Use ensemble empirical mode decomposition combined with post-processing to obtain a series of intrinsic mode function components; S32: The ensemble empirical mode decomposition selects a proper integration number and a Gaussian white noise signal-to-noise ratio for integration. The signal-to-noise ratio should be the estimated noise signal-to-noise ratio, which is the signal-to-noise ratio of the fitting signal relative to the original signal.

3. The method of claim 1, wherein the micro-variation in luminosity of the space object is calculated by amplifying the light intensity of the space object. The step S5 specifically includes the following procedures: S51: Stack the intrinsic mode function components from the bottom to the top to obtain a newly constructed effective slight change signal, calculate the signal-to-noise ratio of the signal and the difference signal, and stop the loop until the signal-to-noise ratio is greater than 0. S52: for the remaining intrinsic mode function component signals, if the permutation entropy is less than the threshold value, add to the effective micro change signal, otherwise, calculate the permutation entropy of each intrinsic mode function component after integrated empirical mode decomposition decomposition, if less than the threshold value, add to the effective micro change signal.

4. The method for calculating and amplifying the micro-variations in photometry of a space target as described in claim 1, characterized in that, The step S6 specifically includes the following process: S61: calculate the amplified micro change signal, the overall mean value of which should not change; S62: add the effective micro change signal and the fitting signal and calculate the mean value, calculate the difference between the effective micro change signal and the mean value to obtain the signal to be amplified, and preset the amplification multiple; S63: based on the preset amplification multiple, calculate the amplified signal to be amplified, and improve the signal-to-noise ratio of the photometric micro change signal.

Images