A FMCW laser ranging method based on matched filtering

Abstract

The application discloses a kind of FMCW laser ranging methods based on matched filter, including FMCW laser, beam splitter, circulator, signal inversion unit, signal conjugate unit, convolver, balanced detector, local oscillator beam, echo light beam are carried out matched filter processing, convolver is carried out convolution operation to local oscillator beam and echo light beam after signal inversion unit, conjugate unit, the result is output to balanced detector;Local oscillator beam, echo light beam are carried out matched filter processing, utilize the maximum signal-to-noise ratio principle of matched filter and the time delay characteristics of echo light beam, after echo light beam signal is inverted, conjugate operation processing is as system function, and local oscillator beam signal is carried out convolution operation to obtain the peak of higher signal-to-noise ratio containing distance information.Can be inverted the distance information in a sweep frequency cycle, improve the ranging range of FMCW laser radar.The requirement of FMCW laser frequency modulation linearity is reduced, and the ranging precision and anti-interference ability are improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of laser ranging, and particularly relates to a FMCW laser ranging method based on matched filtering. BACKGROUND

[0002] In the prior art, the FMCW laser radar can realize the ranging function, because the FMCW laser radar uses the frequency-modulated continuous wave, and calculates the distance through the difference frequency of the local oscillator light beam and the echo light beam. The frequency-modulated continuous laser has very high bandwidth, which can generally reach the range of several GHz. The high bandwidth enables the FMCW laser radar to improve the accuracy of distance measurement.

[0003] The FMCW laser radar adopts a coherent receiving system. The principle is that the transmitted light beam is divided into transmitted light and local oscillator light through a beam splitter, the transmitted light interacts with the target object to generate echo light, and the echo light and the local oscillator light are mixed through an optical receiving system to generate difference frequency and sum frequency signals. Then the difference frequency and sum frequency signals enter the balanced detector, and since the balanced detector can only respond to the difference frequency signal, the difference frequency signal containing distance information is converted into distance information. However, the laser radar has the following shortcomings: 1. The FMCW laser cannot achieve complete linear frequency modulation, and the nonlinear effect of the light source frequency will cause the difference frequency signal to spread in the frequency domain, reduce the spectral resolution, and affect the ranging accuracy of the difference frequency calculation method; 2. With the change of the echo time delay, the maximum difference frequency change can only reach the modulation bandwidth, and the minimum sum frequency change is greater than the modulation bandwidth. In order to retain the difference frequency signal and filter out the high-frequency signals such as sum frequency, the response range of the balanced detector is usually maximized to the modulation bandwidth. Therefore, the difference frequency signal can only be used to invert the distance information of half a frequency sweeping period; 3. The stray light noise caused by optical components and the thermal noise caused by the balanced detector make the echo signal, which is already not strong after atmospheric transmission loss, more likely to be submerged in noise, making it difficult to extract the difference frequency signal and affecting the ranging effect.

[0004] A new ranging method suitable for the FMCW laser radar system is needed to solve the above technical problems. SUMMARY

[0005] The purpose of the present application is to provide a FMCW laser ranging method based on matched filtering, which solves the technical problems of small ranging range, poor noise resistance and poor nonlinear frequency modulation resistance in the prior art.

[0006] The technical solution of the present application to solve the technical problems is as follows:

[0007] A FMCW laser ranging method based on matched filtering, comprising the following steps:

[0008] S1: the FMCW laser emits a laser beam to a beam splitter, the beam splitter divides the laser beam into a local oscillator beam and a transmission beam, the local oscillator beam divided by the beam splitter is input to a convolver; the transmission beam divided by the beam splitter is input to a circulator; the transmission beam emitted by the circulator interacts with a target to generate a return beam, and the local oscillator beam and the return beam are subjected to matched filtering processing;

[0009] The time domain function expression of the local oscillator beam and the transmission beam is:

[0010] rect(·) represents a rectangular window function, T represents a sweep period, K represents a sweep slope, t is time, is an impact string signal, and δ(t-iT) is an impact signal;

[0011] The return beam is separated by the circulator and then processed by a signal inversion unit and a signal conjugate unit, and the processed return beam is input to the convolver; the time domain function expression of the return beam separated by the circulator is:

[0012] t0 is the time delay of the return beam signal relative to the local oscillator beam signal;

[0013] The transmission signal of length NT and the return signal of length NT are intercepted for analysis, and the transmission signal can be expressed as:

[0014] x1(t)=s(t)·R NT (t), N is any positive integer, and R NT (·) represents a rectangular window function with an amplitude of 1 and a width of NT,

[0015] The return signal can be expressed as:

[0016] x2(t)=s(t-t0)·R NT (t)

[0017] According to the matched filtering requirement, the output signal is expressed as:

[0018] y(t)=[s(t)·R NT (t)]*[s * (-t-t0)·R NT (-t)]

[0019] Transformed into:

[0020] y(t)×[s(t)*s * (-t-t0)].[R NT (t)*R NT (-t)]×s out (t).T Δ (t), s out(t) is a time domain convolution function of s(t) and s * (t-t0);T Δ (·) is an isosceles triangle function with base length 2NT and height NT,

[0021] The time domain signal of the local light beam is converted into a frequency domain signal, i.e. after Fourier transform, the frequency domain function expression of the local light beam is:

[0022]

[0023] The time domain signal of the echo light beam signal after inversion and conjugation is converted into a frequency domain signal, i.e. after Fourier transform, the frequency domain function expression of the echo light beam after inversion and conjugation is obtained by taking the conjugate of the Fourier transform expression:

[0024]

[0025] Wherein, F is a Fourier transform operator;

[0026] S2: The local light beam and the echo light beam after inversion and conjugation are convolved by the convolver, and the result is output to the balanced detector;

[0027] S2.1: The partial expression s out (t) is converted into a frequency domain to obtain a frequency domain expression of the convolution function:

[0028]

[0029] S2.2: The frequency domain expression of the convolution function is converted into a time domain expression of the convolution function, and the time domain expression of the convolution function after conversion is:

[0030]

[0031] S2.3: The partial expression s out (t) is multiplied by the partial expression T Δ (t), and the time domain expression of the final output signal of the matched filter is:

[0032]

[0033] S2.4: The matched filter result is output to the balanced detector.

[0034] Further comprising a step S3: inverting the target distance;

[0035] S3.1: selecting an inversion window according to the target distance, and calculating the number of sampling points in the window;

[0036] L=cT / 2, Wherein, c is the speed of light, L is the maximum detection distance under the condition of the matched filtering algorithm, R is the window, T is the sweep frequency cycle, n is the signal sampling point number in the sweep frequency cycle T, M is the signal sampling point number in the window R;

[0037] S3.2: The data in the window R is taken as the target distance inversion signal, and the target distance inversion signal is normalized to obtain the distance peak value:

[0038]

[0039] The laser light source emitted by the FMCW laser adopts triangular wave modulation.

[0040] The sweep frequency bandwidth of the laser light source emitted by the FMCW laser is greater than 1GHz.

[0041] The repetition frequency of the laser light source emitted by the FMCW laser is 100KHz-500KHz.

[0042] The beneficial effects of the present application are: the local oscillator beam and the echo beam are subjected to matched filtering processing, the maximum signal-to-noise ratio principle of matched filtering and the time delay characteristics of the echo beam are utilized, the echo beam is taken as a system function after inversion and conjugate operation, and convolution operation is performed with the local oscillator signal to obtain a peak value with high signal-to-noise ratio containing distance information. Since the output signal spectrum of the matched filtering is within the modulation bandwidth range, there is no problem of response range limitation of the balanced detector, the distance information within one sweep frequency cycle can be inverted, and the ranging range of the FMCW laser radar is improved. In addition, the time domain filtering response of the matched filtering matches the transmitted waveform, so that the requirement for the frequency modulation linearity of the FMCW laser is reduced, and the ranging accuracy and anti-interference ability are improved. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 is a flowchart of the FMCW laser ranging method based on matched filtering of the present application;

[0044] Figure 2 is a time-frequency diagram of the processed transmitted beam signal, which adopts triangular wave frequency modulation, has a sweep frequency bandwidth of 2GHz, a repetition frequency of 200KHz, a sweep frequency cycle of 5x10 -6 s, and a six-sweep frequency cycle width rectangular window;

[0045] Figure 3 is a time-frequency diagram of the processed echo beam signal, which adopts triangular wave frequency modulation, has a sweep frequency bandwidth of 2GHz, a repetition frequency of 200KHz, a sweep frequency cycle of 5x10 -6 s, a time delay of 2.5x10 -6 s, and a six-sweep frequency cycle width rectangular window;

[0046] Figure 4 This is the time-domain waveform of the matched filter output signal when the target distance is 375 meters.

[0047] Figure 5 This is the peak distance map retrieved when the target distance of this invention is 375 meters;

[0048] Figure 6 This is the time-domain waveform of the matched filter output signal when the target distance is 750 meters.

[0049] Figure 7 This is the peak distance map retrieved when the target distance of this invention is 750 meters. Detailed Implementation

[0050] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0051] like Figure 1 As shown, this invention is an FMCW laser ranging method based on matched filtering, comprising the following steps:

[0052] S1: The FMCW laser emits a laser beam to a beam splitter, which splits the laser beam into a local oscillator beam and an emitted beam. The local oscillator beam split by the beam splitter is input into a convolutional unit, and the emitted beam split by the beam splitter is input into a circulator. The emitted beam emitted by the circulator interacts with the target to generate an echo beam. The local oscillator beam and the echo beam are subjected to matched filtering. The laser source emitted by the FMCW laser is usually modulated by a triangular wave, with a sweep bandwidth greater than 1 GHz and a repetition frequency of 100 kHz to 500 kHz.

[0053] The time-domain expressions for the local oscillator beam and the emitted beam are:

[0054] rect() represents a rectangular window function, T represents the sweep period, K represents the sweep slope, and t is time. For the impulse train signal, δ(t-iT) is the impulse signal;

[0055] After being separated by a circulator, the echo beam is processed by a signal inversion unit and a signal conjugation unit. The processed echo beam is then input into a convolutional unit. The time-domain function expression of the echo beam separated by the circulator is as follows:

[0056] t0 is the time delay of the echo beam signal relative to the local oscillator beam signal;

[0057] By analyzing the transmitted signal and the echo signal of length NT, the transmitted signal can be described as follows:

[0058] x1(t) = s(t) * R NT (t), N is any positive integer, R NT (·) represents a rectangular window function with an amplitude of 1 and a width of NT,

[0059] The echo signal can be expressed as:

[0060] x2(t) = s(t-t0) * R NT (t)

[0061] According to the requirement of matched filtering, the output signal is expressed as:

[0062] y(t) = [s(t) * R NT (t)] * [s * (-t-t0) * R NT (-t)]

[0063] Transformed into:

[0064] y(t) = [s(t) * s * (-t-t0)] * [R NT (t) * R NT (-t)] = s out (t) * T Δ (t), s out (t) is the time domain convolution function of s(t) and s * (-t-t0); T Δ (·) is an isosceles triangle function with a base length of 2NT and a height of NT,

[0065] The time domain signal of the local oscillator beam is converted into a frequency domain signal, i.e. after Fourier transform, the frequency domain function expression of the local oscillator beam is:

[0066]

[0067] The time domain signal of the echo beam signal after inversion and conjugation is converted into a frequency domain signal, i.e. after Fourier transform, the conjugate of the Fourier transform expression is taken, and the frequency domain function expression of the echo beam after inversion and conjugation is obtained:

[0068]

[0069] Where F is the Fourier transform operator;

[0070] S2: The local oscillator beam and the echo beam after inversion and conjugation are convolved by the convolver, and the result is output to the balanced detector;

[0071] S2.1: Convert the partial expression s out (t) of the matched filter output signal y to the frequency domain to obtain the frequency domain expression of the convolution function:

[0072]

[0073] S2.2: convert the frequency domain expression of the convolution function into the time domain expression of the convolution function, and the converted time domain expression of the convolution function is:

[0074]

[0075] S2.3: multiply the partial expression s out (t) with the partial expression T Δ (t), and the final output signal time domain expression of the matched filtering is:

[0076]

[0077] S2.4: output the matched filtering result to the balanced detector.

[0078] Further comprising a step S3: inverting the target distance;

[0079] S3.1: selecting an inversion window according to the target distance, and calculating the number of sampling points in the window;

[0080] L=cT / 2, wherein c is the speed of light, L is the maximum detection distance under the condition of the matched filtering algorithm, R is the window, T is the sweep frequency period, n is the number of signal sampling points in the sweep frequency period T, and M is the number of signal sampling points in the window R;

[0081] S3.2: taking the data in the window R as the target distance inversion signal, and performing normalization processing on the target distance inversion signal to obtain the distance peak value:

[0082]

[0083] Taking a triangular wave frequency modulation, a sweep bandwidth of 2GHz, a repetition frequency of 200KHz, a six-sweep period width, and a ranging window of 800m as an example:

[0084] As shown in Figure 2 , Figure 3 the time-frequency diagram of the processed transmitted light beam signal and the time-frequency diagram of the processed echo light beam signal are compared, the sweep period is 5x10 -6 s, the echo light beam has a time delay of 2.5x10 -6 seconds compared with the incident light beam, and the target distance is 375m.

[0085] As shown in Figure 4 , Figure 6As shown, after inverse transform processing, the time domain waveform diagram of the matched filter output signal can be seen that the time domain waveform of the matched filter output signal is a plurality of sine functions arranged in turn with a sweep period T as the interval in the isosceles triangle envelope with the base of 2NT and the height of NT. The amplitude of each sine function is the height of the isosceles triangle oblique side at the center frequency of the sine function multiplied by the sampling rate Fs, for example: when N = 6, the sweep period T = 5*10 -6 , the sampling rate Fs = 10 10 , the coefficient of the middle sine function is the isosceles triangle height 6*5*10 -6 , multiplied by the sampling rate Fs = 10 10 , then the multiplication result is 3*10 5 .

[0086] As shown in Figure 5 , Figure 7 , the peak value of the inverse distance is obtained by inverting the target distance The abscissa of the peak value corresponds to the distance Figure 5 , the target distance in the is 375.0023m, and the absolute error is only 0.0023m. Because it is not limited to the frequency response range of the balanced detector, the farthest ranging range of the matched filter FMCW laser ranging can reach L = 750m, which is twice the maximum measurement range of the traditional difference frequency solving FMCW laser ranging algorithm, and the signal-to-noise ratio is high, and the anti-nonlinear frequency modulation capability is strong.

[0087] Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor belong to the scope of protection of the present application.

Claims

1. A matched-filter-based FMCW laser ranging method, characterized in that, Includes the following steps: S1: The FMCW laser emits a laser beam to the beam splitter, which splits the laser beam into a local oscillator beam and an emitted beam. The local oscillator beam split by the beam splitter is input into the convolutional unit; the emitted beam split by the beam splitter is input into the circulator; the emitted beam emitted by the circulator interacts with the target to generate an echo beam, and the local oscillator beam and the echo beam are subjected to matched filtering. The time-domain expressions for the local oscillator beam and the emitted beam are: rect(·) represents a rectangular window function, T represents the sweep period, K represents the sweep slope, and t is time. For the impulse train signal, δ(t-iT) is the impulse signal; After being separated by a circulator, the echo beam is processed by a signal inversion unit and a signal conjugation unit. The processed echo beam is then input into a convolutional unit. The time-domain function expression of the echo beam separated by the circulator is as follows: t0 is the time delay of the echo beam signal relative to the local oscillator beam signal; By analyzing the transmitted signal and the echo signal of length NT, the transmitted signal can be described as follows: x1(t)=s(t)·R NT (t), where N is any positive integer, and R is... NT (·) represents a rectangular window function with an amplitude of 1 and a width of NT. The echo signal can be described as: x2(t)=s(t-t0)·R NT (t) Based on the matched filtering requirements, the output signal is represented as follows: y(t)=[s(t)·R NT (t)]*[s * (-t-t0)·R NT (-t)] Transformed into: y(t) = [s(t) * s * (-t - t0)] · [R NT (t) * R NT (-t)] = s out (t) · T Δ (t), s out (t) is s(t) and s * The time-domain convolution function of (-t-t0); T Δ (·) represents an isosceles triangle function with base length 2NT and height NT. Converting the time-domain signal of the local oscillator beam to the frequency-domain signal, i.e., after Fourier transform, the frequency-domain function expression of the local oscillator beam is: The time-domain signal of the echo beam signal, after being inverted and conjugated, is converted into a frequency-domain signal. That is, after Fourier transform, the Fourier transform expression is taken as its conjugate, resulting in the frequency-domain function expression of the echo beam after inversion and conjugation: Where F is the Fourier transform operator; S2: The convolutional unit performs convolution operations on the local oscillator beam and the echo beam after inversion and conjugation, and outputs the result to the balanced detector. S2.1: The partial expression s of the matched filter output signal y out (t) Transform to the frequency domain to obtain the frequency domain expression of the convolution function: S2.2: Convert the frequency domain expression of the convolution function into the time domain expression of the convolution function. The converted time domain expression of the convolution function is as follows: S2.3: Subtract part of the expression s out (t) and part of the expression T Δ Multiplying (t) together, the final time-domain expression of the matched filter output signal is: S2.4: Output the matched filter result to the balanced detector.

2. The FMCW laser ranging method based on matched filtering according to claim 1, characterized in that: It also includes step S3: inverting the target distance; S3.1: Select the inversion window based on the target distance and calculate the number of sampling points within the window; L = cT / 2, Where c is the speed of light, L is the maximum detection distance under the matched filtering algorithm, R is the window, T is the sweep period, n is the number of signal sampling points within the sweep period T, and M is the number of signal sampling points within the window R. S3.2: Use the data within window R as the target range inversion signal, and normalize the target range inversion signal to obtain the peak range:

3. The FMCW laser ranging method based on matched filtering according to claim 1, characterized in that: The laser source emitted by the FMCW laser is modulated by a triangular wave.

4. The FMCW laser ranging method based on matched filtering according to claim 1, characterized in that: The FMCW laser emits a laser source with a sweep bandwidth greater than 1 GHz.

5. The FMCW laser ranging method based on matched filtering according to claim 1, characterized in that: The FMCW laser emits a laser source with a repetition frequency of 100KHz-500KHz.

Images