Frequency modulation continuous wave lidar phase noise compensation method and device

Abstract

The application discloses a frequency-modulated continuous wave laser radar phase noise compensation method. The frequency-modulated continuous wave signal is divided into two paths: a measurement path and an auxiliary path. The frequency-modulated continuous wave signal of the measurement path is divided into two paths, one of which is used as a reference light signal, and the other is used as a probe light signal to emit to a target. The received target reflected light signal is beat with the reference light signal to obtain a measurement path beat signal. The frequency-modulated continuous wave signal of the auxiliary path is beat after passing through a Mach-Zehnder interferometer with a fixed delay difference to obtain an auxiliary path beat signal. The frequency of the measurement path beat signal is searched and determined, and the phase noise in the auxiliary path beat signal is extracted. Then, the phase noise of the measurement path is estimated. Finally, the phase noise is compensated. The application also discloses a frequency-modulated continuous wave laser radar phase noise compensation device. Compared with the prior art, the application can greatly reduce the calculation amount of digital processing.

Description

Technical Field

[0001] The present invention relates to a laser radar phase noise compensation method, and in particular to a frequency modulated continuous wave (FMCW) laser radar phase noise compensation method and device, belonging to the field of laser radar technology. Background Art

[0002] Frequency-modulated continuous wave (FMCW) lidar is an emerging ranging and imaging technology widely used in autonomous driving, drones, and environmental monitoring. Its basic principle is to transmit a linearly frequency-modulated laser signal, receive the reflected signal from the target, and compare the frequency difference between the two to detect target information. Based on the Doppler effect, the target's velocity can be determined by analyzing the frequency shift in the echo signal. This process enables FMCW lidar to achieve high-precision and real-time ranging and velocity measurement. Compared to traditional pulsed lidar, FMCW lidar offers several significant advantages. First, FMCW lidar offers higher resolution because it uses a continuous waveform rather than discrete pulses. This allows for more data points to be acquired in a shorter timeframe, improving image quality and detail. Second, due to the continuous transmission and reception of signals, FMCW lidar can effectively detect distant targets even at low energy, making it more effective in complex environments. Furthermore, FMCW lidar is more sensitive to small objects and low-reflectivity surfaces, enhancing detection capabilities in a variety of application scenarios.

[0003] FMCW lidar can achieve signal modulation in two main ways: external modulation and internal modulation. External modulation is to frequency modulate the emitted laser signal through additional equipment after the laser is emitted. This process increases the size and cost of the system, and may also introduce additional noise and delay. Internal modulation refers to changing the frequency of the emitted laser beam directly inside the laser, usually by changing the driving current or temperature of the laser. In comparison, internal modulation has several significant advantages. First, since all operations are completed inside the laser, the connection between components is reduced, which helps to improve system stability and reliability. Second, internal modulation usually has a higher frequency response speed, making the signal adjustment faster, thereby improving ranging accuracy.

[0004] However, the performance of internally modulated FMCW lidar systems is limited by multiple factors, such as frequency sweep nonlinearity and phase noise. Frequency sweep nonlinearity can be calibrated using a preprocessed modulated current signal. Phase noise, however, primarily originates from the laser's spontaneous emission and changes in the external environment. Because FMCW lidar calculates target distance by comparing the frequency difference between the transmitted and received signals, this noise can cause a frequency offset between the transmitted and received signals, directly impacting the accuracy of ranging and velocity measurements. Furthermore, the impact of phase noise intensifies as the measured distance increases. When detecting moving objects, velocity information is obtained by analyzing the frequency offset of the echo signal. If significant phase noise is present, this frequency offset will be masked, leading to significant errors in velocity estimation.

[0005] Therefore, it is necessary to eliminate or compensate for the phase noise, so as to improve the measurement distance and accuracy of the FMCW laser radar system. Reference [1] proposed a phase noise compensation method based on digital processing. This method divides the frequency modulated continuous optical signal into two paths: the measurement path and the auxiliary path; the frequency modulated continuous optical signal of the measurement path is divided into two paths, one path is used as the reference optical signal, and the other path is used as the detection optical signal to be transmitted to the target, and the received target reflected optical signal is beat with the reference optical signal to obtain the measurement path beat signal; the optical signal of the auxiliary path is passed through an electro-optical modulator with a fixed frequency shift and then beat to obtain the auxiliary path beat signal; finally, the auxiliary path beat signal and the measurement path beat signal are digitally processed to achieve the compensation of the measurement path phase noise. This scheme improves the compensation accuracy of the phase noise, reduces the requirements for the laser line width and auxiliary hardware, and further improves the dynamic range of the distance measurement. However, this scheme does not accurately estimate the phase noise of the measurement path, but rather selects the optimal solution based on the results of different phase noise compensation by traversing the delay parameters of the phase noise within a certain range, and the document does not explain how to determine the above delay range. In addition, in order to obtain better results, the range needs to be continuously refined, which significantly increases the computational burden of data processing.

[0006] References:

[0007] [1]Xu C, et al.Adaptive carrier-phase-noise-canceled LiDAR for range-Doppler imaging beyond hundreds of laser coherence length[J].Optics Letters, 2024,49(15):4150-4153. Summary of the Invention

[0008] The technical problem to be solved by the present invention is to overcome the shortcomings of the existing technology and provide a frequency modulated continuous wave lidar phase noise compensation method, which can accurately estimate the phase noise of the measurement path and perform phase noise compensation accordingly without traversing all possibilities to determine the optimal solution, and can be implemented automatically, avoiding the need for manual intervention.

[0009] The present invention specifically adopts the following technical solutions to solve the above technical problems:

[0010] A method for compensating phase noise of a frequency modulated continuous wave laser radar is disclosed. The frequency modulated continuous wave laser radar divides a frequency modulated continuous light signal into two paths: a measurement path and an auxiliary path. The frequency modulated continuous light signal of the measurement path is divided into two paths, one path is used as a reference light signal, and the other path is used as a detection light signal to be transmitted to the target. The received target reflected light signal is beat with the reference light signal to obtain a measurement path beat signal. The frequency modulated continuous light signal of the auxiliary path passes through a fixed delay difference τ. r The auxiliary path beat signal is obtained by performing the following digital processing on the auxiliary path beat signal and the measurement path beat signal to obtain the phase noise of the measurement path, and then the obtained measurement path phase noise is compensated:

[0011] First, preliminarily search out the frequency range of the measurement path beat frequency signal, and then use the linear frequency modulation z-transform method to determine the frequency of the measurement path beat frequency signal within the preliminarily searched frequency range. At the same time, the phase noise in the auxiliary path beat signal is extracted The phase noise of the measurement path is estimated according to the following formula:

[0012]

[0013] in, round[·] means rounding to the nearest integer, γτ r is the beat frequency of the auxiliary path beat signal,

[0014] Preferably, a modified bisection method is used to preliminarily search for the frequency range of the beat frequency signal of the measurement path, specifically as follows: the entire frequency band interval is divided into three parts, and there is overlap between every two adjacent parts of the three parts; the average power of the three parts is calculated and the part with the largest average power is selected to repeat the above process; it is iterated in sequence until the width of the final selected partial frequency band meets the preset requirements.

[0015] Preferably, the following method is used to compensate the obtained phase noise of the measurement path: Construct compensation factor Then use the compensation factor Multiplied by the measurement path beat frequency signal.

[0016] Based on the same inventive concept, the following technical solutions can also be obtained:

[0017] A frequency modulated continuous wave laser radar phase noise compensation device, the frequency modulated continuous wave laser radar divides the frequency modulated continuous light signal into two paths: a measurement path and an auxiliary path; the frequency modulated continuous light signal of the measurement path is divided into two paths, one path is used as a reference light signal, and the other path is used as a detection light signal to be emitted to the target, and the received target reflected light signal is beat with the reference light signal to obtain the measurement path beat frequency signal; the frequency modulated continuous light signal of the auxiliary path passes through a fixed delay difference τ r The Mach-Zehnder interferometer is then used to perform beat frequency processing to obtain an auxiliary path beat frequency signal; the frequency modulated continuous wave laser radar phase noise compensation device obtains the phase noise of the measurement path by performing the following digital processing on the auxiliary path beat frequency signal and the measurement path beat frequency signal, and then compensates for the obtained measurement path phase noise, specifically including:

[0018] A frequency range search module is used to preliminarily search for the frequency range of the beat frequency signal in the measurement path;

[0019] Frequency estimation module, used to determine the frequency of the measurement path beat signal within the preliminarily searched frequency range using the linear frequency modulation z-transform method

[0020] Phase noise extraction module, used to extract the phase noise in the auxiliary path beat frequency signal The measurement path phase noise estimation module is used to estimate the phase noise of the measurement path according to the following formula

[0021]

[0022] in, round[·] means rounding to the nearest integer, γτ r is the beat frequency of the auxiliary path beat signal,

[0023] The phase noise compensation module is used to compensate for the phase noise of the obtained measurement path.

[0024] Preferably, the frequency range search module uses a modified bisection method to preliminarily search for the frequency range of the beat frequency signal of the measurement path, specifically as follows: the entire frequency band interval is divided into three parts, and there is overlap between every two adjacent parts of the three parts; the average power of the three parts is calculated and the part with the largest average power is selected to repeat the above process; iterate in sequence until the width of the final selected partial frequency band meets the preset requirements.

[0025] Preferably, the phase noise compensation module uses the following method to compensate the obtained phase noise of the measurement path: according to the estimated phase noise of the measurement path Construct compensation factor Then use the compensation factor Multiplied by the measurement path beat frequency signal.

[0026] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0027] The present invention is based on digital processing technology and can accurately estimate the phase noise of the measurement path. It then effectively eliminates the adverse effects of phase noise on distance and speed measurement results through phase noise compensation, thereby improving detection accuracy. The digital processing process of the present invention is simple, reduces the amount of calculation, can effectively improve the real-time performance of radar detection, and reduce the requirements for hardware performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a schematic diagram of the structural principle of the frequency modulated continuous wave laser radar phase noise compensation device of the present invention;

[0029] Figure 2 This is the principle diagram of FMCW lidar ranging and speed measurement;

[0030] Figure 3 The result diagram of the search for the frequency range of the beat signal of the measurement path;

[0031] Figure 4 This is the result diagram of the CZT frequency estimation of the measurement path beat signal;

[0032] Figure 5 This is the result diagram after the phase noise compensation processing of the beat frequency signal of the measurement path;

[0033] Figure 6 is the peak value of the signal power spectrum density after phase noise compensation for different N and p. DETAILED DESCRIPTION

[0034] In view of the shortcomings of the existing technology, the solution of the present invention is to improve its digital processing part, estimate the phase noise of the measurement path by digitally processing the auxiliary path beat signal and the measurement path beat signal, and then compensate for the obtained measurement path phase noise.

[0035] To facilitate public understanding, the technical solution of the present invention is described in detail below with reference to the accompanying drawings:

[0036] The basic structure of the FMCW laser radar phase noise compensation device of the present invention is as follows: Figure 1 As shown, an arbitrary waveform generator generates a modulation signal to control the tunable laser light, generating a linear frequency modulated continuous optical signal whose frequency varies with an applied voltage. The generated frequency modulated continuous optical signal is divided into two paths: a measurement path and an auxiliary path. The frequency modulated continuous optical signal of the measurement path is divided into two paths, one path is used as a reference optical signal, and the other path is used as a detection optical signal to be transmitted to the target. The received target reflected optical signal and the reference optical signal are subjected to a balanced detector for frequency beat, thereby obtaining a measurement path beat signal. The frequency modulated continuous optical signal of the auxiliary path passes through a Mach-Zehnder interferometer with a fixed time delay difference and is subjected to a balanced detector for frequency beat, thereby obtaining an auxiliary path beat signal. The auxiliary path beat signal and the measurement path beat signal are then digitally processed: first, a preliminary search is performed to determine the frequency range of the measurement path beat signal, and then a linear frequency modulated z-transform method is used to determine the frequency of the measurement path beat signal within the preliminary searched frequency range. Simultaneously, the phase noise in the auxiliary path beat signal is extracted. The phase noise of the measurement path is estimated based on the frequency of the measurement path beat signal and the phase noise in the auxiliary path beat signal. Finally, the obtained phase noise of the measurement path is compensated.

[0037] The technical solution of the present invention is further described in detail below through a specific embodiment:

[0038] This embodiment uses an arbitrary waveform generator to generate a triangular waveform voltage with a period of T as a modulation signal to control the laser to generate a linear frequency modulated continuous optical signal whose optical frequency changes with the applied voltage. The frequency variation curves of the transmitted signal and the received signal over time are shown in Figure 2. Figure 2 As shown in (a), the solid line is the transmitted signal and the dotted line is the received signal. The initial modulation frequency of the laser is f0, the modulation bandwidth is B, and the modulation period is T. Figure 2 (b) is the time-frequency diagram of the ideal beat signal. Due to the Doppler frequency f D The existence of the modulation period T, the frequencies obtained by the beat frequency in the upper and lower frequency modulation stages are not equal, respectively f b+ and f b- Then, the radial distance and velocity of the target relative to the lidar can be obtained through these two frequencies:

[0039]

[0040] Where c is the speed of light in a vacuum environment, λ is the central wavelength of the laser, and τ c It measures the delay between the transmitted and received signals in the path.

[0041] Since the analysis of the up and down frequency modulation stages is similar, we will only consider the up frequency modulation stage to simplify the formula. Within the range of half a modulation period (0, T / 2), the FMCW reference frequency f(t) can be expressed as:

[0042] f(t)=f0+γt,γ=2B / T (2)

[0043] Therefore, taking phase noise into account, the optical signal output by the tunable laser can be expressed as:

[0044]

[0045] Where, is the phase noise.

[0046] Then, the optical signal output by the laser is divided into two paths, where a small part of the light is introduced into a fixed delay τ r The Mach-Zehnder interferometer (MZI) is used as an auxiliary path, and most of the rest is used for target measurement. The auxiliary path is delayed by τ r The FMCW optical signal can be expressed as:

[0047]

[0048] The auxiliary MZI couples the two FMCW signals with fixed delays and then sends them to the photodetector for beat frequency. After removing the DC and high-frequency terms, the beat frequency signal can be expressed as:

[0049]

[0050] Similarly, the beat frequency signal of the measurement path can also be expressed as:

[0051]

[0052] Next, the frequency range of the beat signal in the measurement path is estimated. First, the frequency range of the beat signal is searched. Based on the power spectrum density of the signal, the average power within a certain range can be obtained to avoid the error that may be caused by directly obtaining the maximum value of the spectrum. At the same time, this method realizes the automation of frequency range estimation and does not require manual spectrum analysis. The location of the beat signal can be determined by comparing the average power in different frequency ranges. s To sample the beat frequency signal, I2(t) is rewritten as a discrete signal I2(nT s ), where T s is the sampling time interval 1 / f s Perform discrete Fourier transform (DFT) on a discrete beat frequency signal with N sampling points:

[0053]

[0054] The actual frequency corresponding to each frequency component of X(k) is f k =kf s / N. The power spectral density (PSD) is obtained from the DFT output and is expressed as:

[0055]

[0056] The signal frequency range search can adopt conventional algorithms such as binary search. In order to improve the search efficiency, the present invention further proposes an improved binary search method. The steps of searching the signal frequency range by the improved binary search method are as follows: first, determine the entire frequency band interval as [f1, f2], and divide it into three parts. There is overlap between every two adjacent parts of the three parts; after the division, the width of each frequency band is the same as (f2-f1) / 2, and their starting frequencies decrease from high to low. The starting frequency difference between every two adjacent frequency bands is (f2-f1) / 4. Therefore, the three frequency band intervals are [f1, f1 / 2+f2 / 2], [3f1 / 4+f2 / 4, f1 / 4+3f2 / 4], and [f1 / 2+f2 / 2, f2]; by performing S x (f k ) to perform trapezoidal numerical integration and calculate the average power of the three frequency bands respectively; select the frequency band with the largest average power and repeat the above operation, and gradually narrow the search range of the beat frequency through iteration until the width of the final selected partial frequency band meets the preset requirements.

[0057] Assume that the frequency range of the beat signal is finally determined to be [f start ,f end ]. Next, the linear frequency modulation z-transform (CZT) method is used to estimate the beat frequency within the selected frequency range. CZT maps the input signal to an arbitrarily selected complex plane region, which means that higher-density data processing can be performed on certain key frequency bands, thereby improving the resolution within this region. In addition, CZT can enhance the analysis capability of high-frequency components. By adjusting the parameters, computing resources can be concentrated on specific frequency bands, thereby improving the recognition accuracy of these components. According to the part with higher linearity in the frequency modulated optical signal, that is, the middle part of the up and down frequency modulation stages, the beat signal of the corresponding time period is selected because the frequency change in this time period is relatively gentle. According to the above method, N1 sampling points of the beat signal are selected for CZT, which can be expressed as:

[0058]

[0059] Where A is the complex starting point and W is the interval between adjacent sampling points, which can be expressed as:

[0060]

[0061] Assuming the CZT peak position is at k1, the corresponding peak frequency is:

[0062]

[0063] The phase noise in the auxiliary path beat signal can be extracted by digital processing. Since the MZI in the auxiliary path has a small delay difference, the beat signal in the auxiliary path is less affected by the phase noise. In this case, the beat frequency can be regarded as the frequency at which the spectrum peak is located after the discrete Fourier transform. The beat frequency γτ r It is known that the phase noise of the signal can be extracted by digital processing as follows:

[0064]

[0065] where unwrap[·], tan -1 [·], hilbert[·] represent the phase unwrapping function, the inverse tangent function, and the Hilbert function, respectively.

[0066] Then, the phase noise extracted by the auxiliary path is combined with the beat frequency roughly estimated by the measurement path to estimate the phase noise of the measurement path

[0067] First, the phase noise of the measurement path can be written as follows:

[0068]

[0069] The first term of formula (13) can be obtained by the delay summation calculation method for the phase noise of the auxiliary path:

[0070]

[0071] where round[·] represents rounding to the nearest integer.

[0072] The second term of formula (13) is regarded as a residual term when the delay τ r is small, and can be regarded as a linear relationship about the delay, so the residual term can be expressed as:

[0073]

[0074] where p = (τ c -Nτ r ) / τr is an unknown number and can be estimated

[0075] Assume that the estimated phase noise is The measured path beat frequency signal is compared with the constructed compensation factor By multiplying them, we can get the beat frequency signal of the measurement path after phase noise compensation, which is expressed as:

[0076]

[0077] According to formula (16), after the phase noise compensated signal undergoes Fourier transform, the beat frequency in the positive half spectrum is almost unaffected by the phase noise, and both the resolution and signal-to-noise ratio will be significantly improved.

[0078] In order to verify the technical effect of the present invention, the following experiments were performed:

[0079] First, a frequency modulated continuous wave laser radar ranging system was built. The specific structure is as follows Figure 1 As shown. A distributed feedback laser with a line width of 200KHz is used as the light source, and the corresponding coherence length is about 250m (round-trip distance). The AWG generates a triangle wave signal with a period of 100us and an amplitude of 0-600mV as the modulation signal. The measured modulation bandwidth is 20GHz, and the corresponding distance resolution is 0.75cm. A 2m single-mode optical fiber is used as the auxiliary path of the MZI two-way optical path difference. Then, the improved dichotomy method is used to search the frequency range of the measurement path beat frequency signal, and the results are shown as follows. Figure 3 As shown in the figure, the final frequency range is [1774MHz, 1784MHz]. Next, within the above frequency range, the CZT frequency estimation of the beat signal is performed, and the result is as follows: Figure 4 As shown, the estimated peak frequency is 1779.09 MHz.

[0080] On the other hand, the beat signal of the auxiliary path is digitally processed. First, the beat signal frequency is obtained by FFT to be 3.7861MHz. Then, the phase noise is extracted according to formula (12), and then the phase noise is estimated by combining the beat frequency of the measurement path. According to formulas (14) and (15), N = 470 and p = -0.1 are calculated. Finally, the beat signal of the measurement path is compensated, and the result is as follows: Figure 5 As shown in the figure, the coherence peak reappears, the signal-to-noise ratio is significantly improved, the spectrum peak frequency is 1778.83MHz, and the corresponding distance is 704.75m. The experimental results show that this scheme significantly improves the measurement distance in the FMCW lidar system, thereby realizing super-coherent length measurement. Figure 6As shown in FIG, the optimal solution obtained by the method of reference [1] is still N = 470, p = -0.1. Therefore, compared with reference [1], the technical solution of the present invention reduces the traversal process and greatly reduces the amount of calculation.

Claims

1. A method for compensating phase noise of a frequency modulated continuous wave laser radar. The frequency modulated continuous wave laser radar divides a frequency modulated continuous wave optical signal into two paths: a measurement path and an auxiliary path. The frequency modulated continuous wave optical signal of the measurement path is divided into two paths, one path is used as a reference optical signal, and the other path is used as a detection optical signal to be transmitted to the target. The received target reflected optical signal is beat with the reference optical signal to obtain a measurement path beat signal. The frequency modulated continuous wave optical signal of the auxiliary path passes through a fixed delay difference τ. r The auxiliary path beat frequency signal is obtained by performing a beat frequency measurement after the Mach-Zehnder interferometer; the characteristic of the beat frequency measurement is that: The phase noise of the measurement path is obtained by performing the following digital processing on the auxiliary path beat signal and the measurement path beat signal, and then the obtained phase noise of the measurement path is compensated: First, preliminarily search out the frequency range of the measurement path beat frequency signal, and then use the linear frequency modulation z-transform method to determine the frequency of the measurement path beat frequency signal within the preliminarily searched frequency range. At the same time, the phase noise in the auxiliary path beat signal is extracted The phase noise of the measurement path is estimated according to the following formula: in, round[·] means rounding to the nearest integer, γτ r is the beat frequency of the auxiliary path beat signal, 2. The method for compensating phase noise of a frequency modulated continuous wave laser radar according to claim 1, wherein: The frequency range of the beat signal in the measurement path is initially searched using a modified binary search method. Specifically, the entire frequency band is divided into three parts, with overlap between adjacent parts. The average power of the three parts is calculated and the part with the largest average power is selected to repeat the above process. The process is iterated until the width of the selected frequency band meets the preset requirements.

3. The method for compensating phase noise of a frequency modulated continuous wave laser radar according to claim 1, wherein: The phase noise of the obtained measurement path is compensated using the following method: Based on the estimated phase noise of the measurement path Construct compensation factor Then use the compensation factor Multiplied by the measurement path beat frequency signal.

4. A frequency modulated continuous wave laser radar phase noise compensation device, wherein the frequency modulated continuous wave laser radar divides the frequency modulated continuous wave optical signal into two paths: a measurement path and an auxiliary path; the frequency modulated continuous wave optical signal of the measurement path is divided into two paths, one path is used as a reference optical signal, and the other path is used as a detection optical signal to be transmitted to the target, and the received target reflected optical signal is beat with the reference optical signal to obtain the measurement path beat frequency signal; the frequency modulated continuous wave optical signal of the auxiliary path passes through a fixed delay difference τ r The auxiliary path beat frequency signal is obtained by performing a beat frequency measurement after the Mach-Zehnder interferometer; the characteristic of the beat frequency measurement is that: The frequency modulated continuous wave laser radar phase noise compensation device obtains the phase noise of the measurement path by performing the following digital processing on the auxiliary path beat frequency signal and the measurement path beat frequency signal, and then compensates the obtained phase noise of the measurement path, specifically including: A frequency range search module is used to preliminarily search for the frequency range of the beat frequency signal in the measurement path; Frequency estimation module, used to determine the frequency of the measurement path beat signal within the preliminarily searched frequency range using the linear frequency modulation z-transform method Phase noise extraction module, used to extract the phase noise in the auxiliary path beat frequency signal The measurement path phase noise estimation module is used to estimate the phase noise of the measurement path according to the following formula in, round[·] means rounding to the nearest integer, γτ r is the beat frequency of the auxiliary path beat signal, The phase noise compensation module is used to compensate for the phase noise of the obtained measurement path.

5. The FMCW laser radar phase noise compensation device according to claim 4, characterized in that: The frequency range search module uses a modified bisection method to preliminarily search for the frequency range of the measurement path beat frequency signal. Specifically, the entire frequency band interval is divided into three parts, and there is overlap between each two adjacent parts of these three parts; the average power of the three parts is calculated and the part with the largest average power is selected and repeated. The above process is iterated in sequence until the width of the final selected partial frequency bandwidth meets the preset requirements.

6. The FMCW laser radar phase noise compensation device according to claim 4, characterized in that: The phase noise compensation module uses the following method to compensate the phase noise of the obtained measurement path: Construct compensation factor Then use the compensation factor Multiplied by the measurement path beat frequency signal.

Images