Vibration error compensation method for coherent laser radar based on quadratic compensation

Abstract

The application discloses a kind of based on secondary compensation's coherent laser radar vibration error compensation method, comprising: using T-FMCW laser radar system to obtain original target echo signal, and the echo signal of each T-FMCW observation period is carried out data decomposition, the signal obtained in one period is decomposed into upper difference frequency signal and lower difference frequency signal;Using adaptive differential evolution algorithm to the upper difference frequency signal and lower difference frequency signal containing vibration error are compensated for quadratic term vibration error;The upper difference frequency signal and lower difference frequency signal after quadratic term compensation are compensated for first-order vibration error, obtain one-dimensional range image after vibration error compensation.The application can not increase additional laser and other hardware configurations, only using one T-FMCW period signal compensation time-varying vibration error, and it is applicable to the scene without strong scattering point.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of coherent laser radar, and particularly relates to a vibration error compensation method for a coherent laser radar based on secondary compensation. BACKGROUND

[0002] Coherent laser radar can achieve large size, high precision, and blind area-free distance measurement with low power, and has broad application prospects in the field of high-resolution observation. Triangular wave and sawtooth wave are two common waveforms in frequency-modulated continuous wave (FMCW) signals. Compared with sawtooth wave, triangular frequency-modulated continuous wave (T-FMCW) can effectively reduce the distance-speed coupling of the target, and can achieve higher target resolution. In addition, the triangular modulation waveform can reduce the requirement for the frequency jump of the transmitter, thereby balancing and stabilizing the frequency modulation linearity of the transmitted signal. Therefore, the radar system based on T-FMCW has good development prospects in various fields. Based on this, NASA has carried out a series of T-FMCW laser radar system development and testing work, and developed a T-FMCW coherent laser radar system based on all-optical fiber by using optical heterodyne receiving technology and advanced optical fiber technology. The system sets up multiple receiving channels to obtain high-precision speed and distance information, realizes autonomous and safe landing of the landing cabin, and thus assists the landing task on the moon and Mars. The work of NASA provides a reference for the research of airborne laser radar based on T-FMCW.

[0003] However, the wavelength of laser is in the order of microns, and the micron-level vibration between the radar installed on the airborne platform and the target will disturb the distance determination of the return signal, thereby affecting the subsequent imaging processing and target detection accuracy.

[0004] A Position and Orientation System (POS) system is usually installed in the radar airborne platform. The system obtains the position information of the platform through global navigation satellites, and obtains the attitude information of the platform through an inertial navigation system. By using the platform position and attitude information provided by the POS system, the motion error in the return data due to the motion of the radar airborne platform is preliminarily compensated, which is called a sensor-based motion error compensation method. In order to further reduce the influence of vibration on ranging and imaging, additional hardware devices such as lasers or speedometers can be added to compensate for the vibration error, which is called a hardware device-based vibration error compensation method. In order to reduce the hardware weight and cost, researchers have studied a series of data-driven vibration error compensation methods. Such methods analyze the data in depth, design corresponding vibration error compensation algorithms to compensate for the vibration error, which is called a data-driven vibration error compensation method.

[0005] The compensation accuracy of the sensor-based motion error compensation method depends on the accuracy of the POS system, and a high-precision POS system has a high cost and a relatively large weight, and when installed on a platform, such factors need to be considered comprehensively. In addition, the wavelength of the laser radar is short, and micrometer-level vibration can cause corresponding ranging errors. The POS system can preliminarily compensate for the overall motion error of the platform, but it is difficult to capture subtle vibration errors. The vibration error compensation method based on hardware devices mostly samples multiple lasers or additional devices such as tachometers to compensate for vibration errors. When multiple lasers are present in the system, the problem of asynchronization is caused, and the increase in hardware devices increases the complexity of the system and causes additional weight on the platform.

[0006] The vibration error compensation method based on data driving does not need to increase additional devices, compensates for vibration errors through algorithm design, and has high flexibility. One type of method comprehensively models the data of multiple echo cycles, such as the vibration compensation method based on Kalman filtering proposed by Tao and the vibration error compensation method based on time-varying Kalman proposed by Jia. However, when the radar platform is in dynamic ranging, the observation time of each observed spot is limited, and it is difficult to provide sufficient observation data for such algorithms. Another type of method can use one cycle of data for vibration compensation, such as the Doppler shift method. However, this method assumes that the vibration speed is constant, and the vibration compensation effect in a severe vibration environment is poor. The third type of method can compensate for time-varying vibration errors in one cycle, such as the time-varying vibration error compensation method based on instantaneous ranging model proposed by Wang and the time-varying vibration error compensation method based on segmented interference. However, such methods have poor effects in dealing with vibration error compensation problems in scenes without strong scattering points. SUMMARY

[0007] To solve the above problems in the prior art, the present application provides a coherent laser radar vibration error compensation method based on secondary compensation, which can compensate for time-varying vibration errors using only one T-FMCW cycle of signals without increasing additional hardware configurations such as lasers, and is suitable for scenes without strong scattering points. The technical problems to be solved by the present application are solved by the following technical solutions:

[0008] The present application provides a coherent laser radar vibration error compensation method based on secondary compensation, comprising:

[0009] S1: obtaining original target echo signals using a T-FMCW laser radar system, and decomposing the echo signals of each T-FMCW observation cycle, so as to decompose the signals obtained in one cycle into upper and lower difference frequency signals;

[0010] S2: compensating for the secondary vibration errors of the upper and lower difference frequency signals containing vibration errors using an adaptive differential evolution algorithm.

[0011] S3: performing first-order vibration error compensation on the up-mixing frequency signal and the down-mixing frequency signal after the secondary term compensation, to obtain a one-dimensional range image after vibration error compensation.

[0012] In an embodiment of the present application, the S1 comprises:

[0013] A triangular frequency modulation continuous wave is used as a transmitting signal, and the up-mixing frequency signal and the down-mixing frequency signal are obtained through coherent detection. In a frequency modulation period, the up-mixing frequency signal s up (t) and the down-mixing frequency signal s down (t) containing vibration errors are respectively represented as:

[0014]

[0015] Wherein, f c is the center frequency of the transmitting signal, K is the frequency modulation rate, t is the time, τ i = 2R i / c is the echo delay of the i-th target, c is the speed of light, R i is the distance between the i-th target and the receiving antenna, v0 and a are respectively the initial speed and acceleration of the vibrating lidar, and λ represents the wavelength t c1 and t c2 are respectively the center time of the up-mixing frequency signal and the down-mixing frequency signal.

[0016] A simplified form of the up-mixing frequency signal s up (t) and the down-mixing frequency signal s down (t) containing vibration errors is obtained:

[0017]

[0018] Wherein, m0, m1 and m2 are respectively the constant term coefficient, the first-order term coefficient and the second-order term coefficient in the phase of the up-mixing frequency signal; n0, n1 and n2 are respectively the constant term coefficient, the first-order term coefficient and the second-order term coefficient in the phase of the down-mixing frequency signal.

[0019] In an embodiment of the present application, the S2 comprises:

[0020] S2.1: modeling the vibration error of one T-FMCW period as a second-order vibration error model, and estimating the second-order vibration coefficients of the up-mixing frequency signal and the down-mixing frequency signal;

[0021] S2.2: establishing a second-order compensation filter according to the estimated second-order coefficients, and compensating the second-order vibration error by using the second-order compensation filter.

[0022] In an embodiment of the present application, the S2.1 comprises:

[0023] The minimum value of the spectrum entropy of the target echo signal is taken as a target function, and an adaptive differential evolution algorithm is used to obtain the optimal solution of the target function to estimate the quadratic term vibration coefficients of the upper and lower beat signals:

[0024]

[0025] wherein, and are the estimated quadratic term coefficients, EN[·] is the entropy value of a vector, k m and k n are vectors in the solution space in the upper and lower periods, argmin[·] represents the quadratic term coefficient when the vector takes the minimum value, and FFT[·] represents the Fourier transform.

[0026] In an embodiment of the present application, the S2.2 comprises:

[0027] According to the estimated quadratic term coefficients of the upper beat signal and the quadratic term coefficients of the lower beat signal corresponding quadratic term compensation filters are designed respectively:

[0028]

[0029] The quadratic terms in the phases of the upper beat signal and the lower beat signal containing vibration errors are compensated by using the quadratic term compensation filters respectively to obtain the upper beat signal and the lower beat signal after compensation of the quadratic terms:

[0030]

[0031] wherein, s up-m2 (t) and s down-n2 (t) represent the upper beat signal and the lower beat signal after compensation of the quadratic terms respectively.

[0032] In an embodiment of the present application, the S3 comprises:

[0033] S3.1: Fourier transform is performed on the upper beat signal and the lower beat signal after compensation of the quadratic terms to obtain the spectra of the upper beat signal and the lower beat signal;

[0034] S3.2: according to the spectra of the upper beat signal and the lower beat signal, the relative frequency shift Δf of the upper beat signal and the lower beat signal is obtained by using the spectrum correlation method;

[0035] S3.3: the vibration initial velocity is estimated by using the relative frequency shift:

[0036]

[0037] S3.4: Designing a linear term error compensation filter of the upper beat signal and the lower beat signal by using the estimated initial vibration velocity:

[0038]

[0039] S3.5: Compensating the linear term error in the phase of the quadratic term compensated upper beat signal and lower beat signal caused by the initial vibration velocity by using the linear term error compensation filter, respectively, to obtain the beat signal after compensating the vibration error:

[0040]

[0041] S3.6: Fourier transforming the beat signal after compensating the vibration error to obtain the one-dimensional range image after compensating the vibration error.

[0042] Compared with the prior art, the present application has the beneficial effects that:

[0043] 1. The coherent laser radar vibration error compensation method based on quadratic compensation of the present application completes vibration error compensation in the data domain, does not need to increase an additional laser, and can effectively avoid the non-synchronization problem between multiple lasers and reduce the hardware configuration.

[0044] 2. When the airborne coherent laser radar imaging system performs dynamic range measurement, the measurement time of a single observation spot is short, and the vibration error is time-varying. The method of the present application first divides the beat signal of one T-FMCW cycle into an upper beat signal and a lower beat signal, models and compensates the time-varying vibration error, so that the method only needs one T-FMCW cycle of beat signal to complete the time-varying vibration error compensation.

[0045] 3. The present application uses the minimum value of the signal spectrum entropy as the objective function, uses the ADE (Adaptive Differential Evolution) algorithm to obtain the optimal solution of the objective function, estimates the quadratic vibration coefficient. A good mutation vector can be generated by the adaptive mutation scheme, a good search space is obtained and fast convergence is achieved. Therefore, the method is suitable for time-varying vibration error compensation in a scene without strong scattering points.

[0046] The present application will be further described in detail below in combination with the drawings and embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0047] Figure 1 is a flowchart of a coherent laser radar vibration error compensation method based on quadratic compensation provided by an embodiment of the present application;

[0048] Figure 2is a processing process schematic diagram of a vibration error compensation method of a coherent laser radar based on quadratic compensation provided by an embodiment of the application;

[0049] Figure 3 is a basic structure diagram of a T-FMCW laser radar system provided by an embodiment of the application;

[0050] Figure 4 is a triangular wave frequency modulation ranging schematic diagram containing a linear vibration error provided by an embodiment of the application;

[0051] Figure 5a is an ideal beat frequency signal and a beat frequency signal containing a time-varying vibration error schematic diagram;

[0052] Figure 5b is Figure 5a a one-dimensional range image corresponding to the ideal beat frequency signal and the beat frequency signal containing a time-varying vibration error in

[0053] Figure 6a is a minimum entropy value schematic diagram of each iteration of the ADE algorithm;

[0054] Figure 6b is a one-dimensional range image after compensating for the quadratic vibration error;

[0055] Figure 6c is a one-dimensional range image after compensating for the linear vibration error;

[0056] Figure 6d is a one-dimensional range image corresponding to the Doppler frequency shift method. DETAILED DESCRIPTION

[0057] In order to further illustrate the technical means and effects taken by the present application to achieve the predetermined object, the following will be combined with the specific embodiments and the drawings to explain in detail a vibration error compensation method of a coherent laser radar based on quadratic compensation according to the present application.

[0058] The foregoing and other technical contents, features and effects of the present application can be clearly presented in the following specific embodiment description with the aid of the drawings. Through the description of the specific embodiments, the technical means and effects taken by the present application to achieve the predetermined object can be understood more deeply and specifically. However, the attached drawings are provided for reference and explanation only, and are not used to limit the technical solutions of the present application.

[0059] It should be noted that the relational terms herein such as first and second and the like are used solely to distinguish one from another entity or action without necessarily requiring or implying any actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion such that a process or method that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process or method. An element proceeded by "comprises a... " does not, without more constraints, preclude the existence of additional identical elements in the process or method in which the element under consideration is used.

[0060] The embodiment of the application first introduces the ranging principle of the coherent laser radar, and then analyzes the influence of time-varying vibration on the ranging and imaging of the coherent laser radar. Secondly, the time-varying vibration error is approximated as a second-order vibration error, and modeling analysis is performed thereon. Finally, a time-varying vibration error compensation method is introduced, which compensates the second-order vibration error and the first-order vibration error respectively, and simulation results verify the effectiveness of the method.

[0061] Specifically, please refer to Figure 1 and Figure 2 , Figure 1 is a flowchart of a coherent laser radar vibration error compensation method based on quadratic compensation provided by the embodiment of the application, Figure 2 is a processing process schematic diagram of a coherent laser radar vibration error compensation method based on quadratic compensation provided by the embodiment of the application. The vibration error compensation method comprises:

[0062] S1: acquiring original target echo signals by using a T-FMCW laser radar system, and performing data decomposition on the target echo signals of each period, decomposing the signals acquired in one T-FMCW observation period into upper and lower difference frequency signals.

[0063] Please refer to Figure 3 , Figure 3 is a basic structure diagram of a T-FMCW laser radar system provided by the embodiment of the application. The coherent detection of the T-FMCW laser radar system mainly comprises the following steps: a tunable laser source TLS generates a T-FMCW laser signal, which is divided into two beams after a coupler 1, one of which is emitted by an optical transmitting antenna, and is reflected to an optical receiving antenna when encountering a target. Compared with the transceiver setting by a single lens plus a circulator, the transceiver separation processing of the optical antenna can effectively reduce the interference of the end surface reflection on the signal reception of the system. The other beam of light separated by the coupler 1 is coherently coupled with the echo signal received by the optical receiving antenna at the coupler 2 to realize the oblique receiving, and the balanced detector D MThe laser signal is sampled to obtain a beat signal; subsequently, the sampled beat signal is subjected to Fourier transform to obtain a frequency domain distribution curve, and finally the frequency information is converted into the distance of the target. The transmitting signal and the receiving signal are both high-frequency optical frequency modulation signals, and the dechirp receiving technology obtains a beat signal after mixing the transmitting signal and the receiving signal, at this time, the signal frequency is small, which greatly reduces the sampling rate of the system.

[0064] Specifically, the influence of vibration on imaging is derived below taking frequency modulation observation as an example. The transmitting signal of the ideal frequency modulation continuous wave (FMCW) laser radar can be expressed as:

[0065] s t (t)=exp(j2πf c t+jπKt 2 ) (1)

[0066] Wherein, f c is the center frequency of the transmitting signal, K is the frequency modulation rate, and t is the time.

[0067] Suppose that the scene contains multiple targets, and the distance between the ith target and the receiving antenna is represented as R i , at this time, the target echo signal can be expressed as the sum of multiple components:

[0068]

[0069] Wherein, τ i =2R i / c is the echo delay of the ith target, and c is the speed of light.

[0070] Subsequently, the target echo signal is mixed with the transmitting signal, and after filtering and amplification, a beat signal is obtained:

[0071]

[0072] Wherein, * represents the conjugate of the signal. Fourier transform is performed on the above formula (3) to obtain the distance compression result, i.e. one-dimensional range image:

[0073]

[0074] Wherein, S(f) represents the one-dimensional range image of the target echo signal, T represents the signal period of the target echo signal, and f represents the signal frequency after Fourier transform. Formula (4) shows that the signal spectrum after Fourier transform is the superposition of multiple sinc functions, and the peak frequency of each target is f i =Kτ i . At this time, the peak distance of the ith target can be expressed as:

[0075]

[0076] It should be noted that the above derivation assumes that the optical path difference of the laser radar to the target to be measured is fixed and unchanging. For a vibrating laser radar, in the case of a short signal period, the measured distance containing vibration errors can be expressed in the form of a second-order Taylor expansion:

[0077]

[0078] where R i (t) is the time-varying distance between the i-th target and the receiving antenna in the vibrating environment, v0 and a are the initial velocity and acceleration of the vibrating laser radar, respectively. At this time, the beat frequency signal containing vibration errors can be expressed as:

[0079]

[0080] The frequency of the beat frequency signal is obtained by taking the derivative of the phase in the above beat frequency signal:

[0081]

[0082] where 2v0 / λ is the Doppler shift caused by vibration, φ(t) represents the phase of the beat frequency signal, and λ represents the wavelength of the beat frequency signal. In the airborne laser radar system, the radial distance from the laser radar platform to the target is usually between a few hundred meters and a few kilometers. If the detection distance is 10 2 ~ 10 3 orders of magnitude, the frequency modulation rate K is of the order of 10 11 , and f c is of the order of 10 14 . Therefore, the above equation (8) can be simplified to the approximate form to facilitate subsequent processing. At this time, the beat frequency signal containing vibration errors can be simplified as:

[0083]

[0084] The beat frequency signal containing vibration in the above equation (9) can be regarded as the superposition of multiple linear frequency modulation signals, and the initial frequency of each signal component is the coefficient Kτ i +2v0 / λ of the first-order term, and the frequency modulation rate is twice the coefficient of the second-order term, 4Kv0 / c+2a / λ. Due to the small laser wavelength, the Doppler effect can amplify the vibration displacement by several thousand times, causing a large deviation in the one-dimensional range image of the target. In addition, the second-order term introduced by vibration causes spectral broadening, and the scattering points of a certain distance unit of the target will spread into adjacent distance gates, causing range ambiguity. Therefore, the existence of vibration errors causes defocusing of the image and a large ranging error, and an effective vibration error compensation method is needed to improve the one-dimensional imaging and ranging effect in the vibrating environment.

[0085] The embodiment adopts a triangular frequency-modulated continuous wave as a transmitting signal, and obtains an upper beat frequency signal and a lower beat frequency signal through coherent detection. Within one frequency-modulated cycle, the upper beat frequency signal and the lower beat frequency signal containing vibration errors are respectively represented as:

[0086]

[0087] wherein s up (t) and s down (t) respectively represent the upper beat frequency signal and the lower beat frequency signal in one triangular frequency-modulated cycle, t c1 and t c2 are the center time of the upper beat frequency signal and the lower beat frequency signal respectively. The above formula (10) is simply written in the following form:

[0088]

[0089] wherein m0, m1 and m2 are respectively a constant term coefficient, a first-order term coefficient and a second-order term coefficient in the phase of the upper beat frequency signal; n0, n1 and n2 are respectively a constant term coefficient, a first-order term coefficient and a second-order term coefficient in the phase of the lower beat frequency signal. The time-varying vibration error compensation can be completed by sequentially compensating the second-order vibration error and the first-order vibration error.

[0090] S2: compensating the second-order vibration error of the upper beat frequency signal and the lower beat frequency signal containing vibration errors by using an adaptive differential evolution algorithm.

[0091] In this step, an optimal solution of a target function is obtained through adaptive mutation, crossover operation and multiple iterations by using an adaptive differential evolution algorithm (ADE), and then the optimal solution is used as a second-order vibration coefficient to compensate the second-order vibration error.

[0092] Specifically, the S2 includes:

[0093] S2.1: modeling the vibration error of one T-FMCW cycle as a second-order vibration error model, and estimating the second-order vibration coefficient of the upper beat frequency signal and the lower beat frequency signal.

[0094] Specifically, when there is no strong scattering point in the observed scene, different distance echo signals with similar scattering coefficients are mixed in the generated beat frequency signals. At this time, the vibration error of one T-FMCW cycle can be modeled as a second-order vibration error model, the minimum value of the spectral entropy of the signal is taken as a target function according to the idea of solving linear frequency modulation, and the optimal solution of the target function is obtained by using an ADE (Adaptive Differential Evolution) algorithm, so as to estimate the corresponding second-order vibration coefficient.

[0095] The optimal decision criterion is determined and the reasonable search range and interval are designed, so that the quadratic vibration error can be estimated from the beat signal. In the embodiment of the application, the entropy value of one-dimensional range image of the beat signal is used as the decision criterion. The beat signal containing the vibration error is subjected to Fourier transform to obtain the one-dimensional range image. Due to the existence of the quadratic vibration error, the spectrum of the range image is widened, the side lobe is lifted, and the entropy value is increased. When the quadratic vibration error is completely compensated, each component in the beat signal is compensated to a single frequency signal, and the entropy value of the corresponding range image is minimum. In the embodiment of the application, the minimum value of the spectrum entropy of the signal is used as the objective function, and the ADE algorithm is used to obtain the optimal solution of the objective function, so as to estimate the corresponding quadratic vibration coefficient. The optimization process can be expressed as:

[0096]

[0097] wherein, and are the estimated quadratic coefficients, EN[·] is the entropy value of the vector, k m and k n are the vectors in the solution space in the upper and lower periods respectively, argmin[·] represents the quadratic coefficient when the vector reaches the minimum value, FFT[·] represents the Fourier transform, and the vector entropy value is expressed as:

[0098]

[0099] wherein, s is the tested signal, and the data length is N; Sp is the signal energy, and is expressed as:

[0100]

[0101] S2.2: A quadratic compensation filter is established according to the estimated quadratic coefficient, and the quadratic vibration error is compensated by using the quadratic compensation filter.

[0102] Specifically, the estimated quadratic coefficients of the upper beat signal and the lower beat signal in formula (12) are used to design the quadratic compensation filter respectively:

[0103]

[0104] The quadratic terms in the phases of the upper beat signal and the lower beat signal in formula (11) are compensated by using the above quadratic compensation filter respectively, and the upper beat signal and the lower beat signal after compensation of the quadratic term can be expressed as:

[0105]

[0106] wherein, s up-m2 (t) and s​​down-n2 (t) respectively represent the upper and lower beat signals after quadratic term compensation.

[0107] S3: performing first term vibration error compensation on the upper and lower beat signals after quadratic term compensation to obtain a one-dimensional range image after vibration error compensation.

[0108] In this step, according to the triangular symmetry relationship of the T-FMCW signal, the vibration initial velocity is estimated by using the spectral cross-correlation method, and the first term vibration error is compensated by designing a first term compensation filter, so as to realize high-resolution imaging.

[0109] Specifically, the schematic diagram of triangular wave frequency modulation ranging when only the first term vibration error is contained is as shown in Figure 4 The solid line is the frequency of the transmitted signal, the dashed line is the frequency of the received signal containing the first term vibration error, and the dotted line is the frequency of the received signal containing the first term vibration error, f d is the Doppler frequency introduced by the first term vibration error. Since the Doppler frequency introduced by the first term vibration error in the upper and lower beat signals has a symmetric characteristic, the vibration initial velocity can be estimated by using the spectral peak symmetry relationship of the upper and lower beat signals, and then the first term error in the beat signal is calculated.

[0110] If there is a strong point target in the scene, the beat signal after compensating for the quadratic term vibration error in formula (16) is a single frequency signal, and there is no obvious spectrum spread in the one-dimensional range image, and the spectral peak is clear and distinguishable. Therefore, the first term vibration error can be estimated by using the spectral peak maximum method.

[0111] Since there is not necessarily a strong point target in the actual scene, the generated beat signal will also mix the echo signals of different distances. At this time, the beat signal will contain multiple peak frequencies, so the spectral peak maximum method cannot be directly used to estimate the target frequency. At this time, the spectral correlation method can be used to obtain the vibration initial velocity.

[0112] Specifically, step S3 of the embodiment includes:

[0113] S3.1: performing Fourier transform on the upper and lower beat signals after quadratic term compensation to obtain the spectrum of the upper and lower beat signals;

[0114] S3.2: according to the spectrum of the upper and lower beat signals, using the spectral correlation method to obtain the relative frequency shift Δf of the upper and lower beat signals;

[0115] S3.3: using the estimated relative frequency shift Δf to estimate the vibration initial velocity, and the estimated initial velocity can be expressed as:

[0116]

[0117] wherein, is the estimated initial velocity. A first-order error compensation filter is designed using the estimated initial velocity:

[0118]

[0119] The first-order error introduced by the initial velocity of vibration in the phase of the upper and lower beat signals in equation (16) is compensated using the above first-order error compensation filter. At this time, the beat signals that compensate for the vibration error can be respectively expressed as:

[0120]

[0121] The Fourier transform of the above equation (19) can obtain the distance compression result of the compensated vibration error, i.e., the one-dimensional range image after compensating for the vibration error.

[0122] Next, the effectiveness of the vibration error compensation method of the embodiment of the present application in processing distributed targets is verified. The scene is set to 8 targets with distances of: [500m, 500.3m, 500.4m, 500.7m, 500.8m, 501.1m, 501.2m, 501.5m]. There is no strong scattering point in the set scene, and the energy of the echo signal at each distance is the same.

[0123] The simulation parameters are designed as follows: the sweep period is 1.2ms, the laser wavelength is 1.55μm, and the signal bandwidth is 1.4GHz. The vibration error with a speed of 0.05m / s and an acceleration of 20m / s 2 is added. Please refer to Figure 5a and Figure 5b , Figure 5a is a schematic diagram of the ideal beat signal and the beat signal containing the time-varying vibration error; Figure 5b is the one-dimensional range image corresponding to the ideal beat signal and the beat signal containing the time-varying vibration error in Figure 5a . It can be known from the simulation parameters that only 5 distances should be able to be distinguished in the ideal one-dimensional range image. Figure 5b 5 distances can be distinguished in , which is consistent with the theoretical analysis. In addition, the energy of each scattering point in the scene is uniform, and there is no strong scattering point. Due to the existence of the time-varying vibration error, Figure 5b the one-dimensional range image of the distributed target in deviates greatly from the theoretical value, and false peaks appear, making it difficult to directly distinguish the target distance.

[0124] The method of the embodiment of the present application is used to compensate the time-varying vibration error in the beat signal of the distributed target. First, the second-order vibration error is compensated. The minimum entropy value of the beat signal is iteratively calculated using the ADE algorithm, and the minimum entropy value of each iteration is as follows: Figure 6aAs shown, the quadratic vibration coefficient is estimated based on the minimum entropy value, thereby compensating for the quadratic vibration error. Figure 6b This is to compensate for the quadratic vibration error in the difference frequency signal, resulting in a one-dimensional range image. Compared to... Figure 5b , Figure 6b The distance distortion caused by the quadratic vibration error disappears, and the trend is consistent with the theoretical distance image, resulting in a significant improvement in distance resolution. However, due to the presence of the first-order vibration error, the overall distance image deviates from the theoretical value. The following section compensates for the first-order vibration error in the difference frequency signal. By obtaining the relative frequency shift of the difference frequency signal after compensating for the quadratic vibration error, the initial vibration velocity can be estimated, thus compensating for the first-order vibration error. The resulting one-dimensional distance image is as follows: Figure 6c As shown. Compared to Figure 5b ,Depend on Figure 6c The distance image in the model matches the theoretical value, and the distance offset phenomenon disappears. Therefore, the method of this embodiment can effectively compensate for the time-varying vibration error of distributed targets, thereby accurately estimating the target distance.

[0125] Since the Doppler frequency shift method can be used to compensate for vibration errors in distributed targets, this section uses the Doppler frequency shift method as a comparative method. Compensation Figure 5a The one-dimensional distance image corresponding to the vibration error of the mid-frequency signal is as follows Figure 6d As shown. Because the Doppler frequency shift method compensates for vibration errors by extracting the relative frequency shift from the upper and lower difference frequency signals, it can only compensate for the first-order vibration error and cannot compensate for the second-order vibration error. From Figure 6d As can be seen, after compensating for vibration errors, the target range image moves to near the theoretical value, but the range image distortion introduced by the quadratic vibration error is different. Figure 5b The distortion phenomenon is consistent with that observed in the data. Therefore, compared to the Doppler frequency shift method, the method of this embodiment can effectively compensate for the time-varying vibration error of distributed targets.

[0126] The embodiment of the application completes vibration error compensation in the data domain based on a secondary compensation coherent laser radar vibration error compensation method, does not need to increase an additional laser, can effectively avoid the non-synchronization problem between multiple lasers and reduce hardware configuration. When an airborne coherent laser radar imaging system performs dynamic distance measurement, the measurement time of a single observation spot is short, and the vibration error is time-varying. The method of the embodiment of the application first divides the beat frequency signal of one T-FMCW period into an upper beat frequency signal and a lower beat frequency signal, models and compensates the time-varying vibration error, and therefore, the method only needs the beat frequency signal of one T-FMCW period to complete time-varying vibration error compensation. The embodiment of the application takes the minimum value of the spectrum entropy of the signal as a target function, uses an ADE algorithm to obtain the optimal solution of the target function, estimates the quadratic term vibration coefficient. Through an adaptive mutation scheme, a good mutation vector can be generated, a good search space can be obtained and fast convergence can be achieved. Therefore, the method is suitable for time-varying vibration error compensation in a scene without strong scattering points.

[0127] The above is a further detailed description of the application in combination with specific preferred embodiments, and the specific implementation of the application cannot be limited to these descriptions. For ordinary skilled persons in the art to which the application belongs, several simple deductions or substitutions can be made without departing from the concept of the application, and all of them should be regarded as falling within the protection scope of the application.

Claims

1. A method for vibration error compensation of a coherent laser radar based on quadratic compensation, characterized in that, The application relates to a vibration error compensation method for a T-FMCW laser radar system. S1: acquiring original target echo signals by using a T-FMCW laser radar system, and performing data decomposition on echo signals of each T-FMCW observation period, so that signals acquired in one period are decomposed into upper difference frequency signals and lower difference frequency signals; S2: performing quadratic term vibration error compensation on the upper difference frequency signals and the lower difference frequency signals containing vibration errors by using an adaptive differential evolution algorithm; S3: performing linear term vibration error compensation on the upper difference frequency signals and the lower difference frequency signals after quadratic term compensation, so as to obtain one-dimensional range images after vibration error compensation. The S2 comprises: S2.1: modeling vibration errors in one T-FMCW period as a second-order vibration error model, and estimating quadratic term vibration coefficients of the upper difference frequency signals and the lower difference frequency signals; S2.2: establishing a quadratic term compensation filter according to the estimated quadratic term coefficients, and compensating quadratic term vibration errors by using the quadratic term compensation filter; The S2.1 comprises: taking a spectrum entropy minimum value of the target echo signals as a target function, obtaining an optimal solution of the target function by using an adaptive differential evolution algorithm, and estimating quadratic term vibration coefficients of the upper difference frequency signals and the lower difference frequency signals: wherein and are estimated quadratic coefficients, is the entropy value of the vector, and are vectors in the solution space in the upper and lower cycle, respectively, denotes the quadratic coefficient when the vector takes the minimum value, denotes the Fourier transform.

2. The method according to claim 1, wherein, The S1 comprises: The triangular frequency-modulated continuous wave is used as a transmitting signal, and the upper and lower difference frequency signals are obtained through coherent detection. In a frequency-modulated period, the upper difference frequency signal containing vibration error and the lower difference frequency signal are respectively represented as: and ​ wherein is the center frequency of the transmitted signal, is the frequency modulation, t is the time, is the echo delay of the i target, is the speed of light, is the distance between the i target and the receiving antenna, and are the initial velocity and the acceleration of the vibrating lidar, respectively, denotes the wavelength, and are the center instants of the upper and lower beat signals, respectively; Obtaining a beat signal containing vibration errors and a beat signal in a simplified form: wherein, , , are constant term coefficient, first order term coefficient and second order term coefficient in the phase of the upper beat signal, respectively; , , are constant term coefficient, first order term coefficient and second order term coefficient in the phase of the lower beat signal, respectively.

3. The method according to claim 2, wherein, The S2.2 comprises: According to the estimated quadratic coefficient of the upper difference frequency signal and the quadratic coefficient of the lower difference frequency signal corresponding quadratic compensating filters are designed respectively: compensating quadratic terms in phases of the upper difference frequency signals and the lower difference frequency signals containing vibration errors by using the quadratic term compensation filter, so as to obtain the upper difference frequency signals and the lower difference frequency signals after compensation of the quadratic terms: wherein and respectively represent the upper and lower beat signals after compensation of the quadratic term.

4. The method according to claim 3, wherein, The S3 comprises: S3.1: performing Fourier transform on the upper difference frequency signals and the lower difference frequency signals after quadratic term compensation, so as to obtain spectra of the upper difference frequency signals and the lower difference frequency signals; S3.2: According to the spectrum of the upper difference frequency signal and the lower difference frequency signal, the relative frequency shift of the upper difference frequency signal and the lower difference frequency signal is obtained by using a spectrum correlation method ; S3.3: estimating vibration initial velocity by using the relative frequency shift: S3.4: designing a linear term error compensation filter of the upper difference frequency signals and the lower difference frequency signals by using the estimated vibration initial velocity: ； S3.5: compensating linear term errors in phases of the upper difference frequency signals and the lower difference frequency signals after quadratic term compensation by using the linear term error compensation filter, so as to obtain the difference frequency signals after vibration error compensation: ； S3.6: performing Fourier transform on the difference frequency signals after vibration error compensation, so as to obtain one-dimensional range images after vibration error compensation.

Images