CN116930947A - High-precision FMCW system distance measurement method - Google Patents

Abstract

The invention belongs to the technical field of FMCW distance measurement, and particularly relates to a high-precision FMCW system distance measurement method. The method of the invention samples the beat frequency signal at two sampling rates, divides the beat frequency signal into two paths of processing, obtains the frequency gap through the phase difference obtained by the two paths of processing, and obtains the enhanced distance measurement result by using the frequency gap, thereby obviously improving the distance measurement precision of the FMCW system under low bandwidth.

Description

A high-precision distance measurement method for FMCW systems

Technical field

The invention belongs to the technical field of FMCW distance measurement, and specifically relates to a high-precision FMCW system distance measurement method.

Background technique

FMCW system, that is, frequency modulated continuous wave system, the oscillation frequency of its transmitted signal changes linearly with time. FMCW is widely used in various distance measurement systems. One common application is vehicle-mounted laser ranging radar. The distance measurement principle of the FMCW system is to emit a time-frequency diagram showing a triangular wave transformation waveform. The emitted wave propagates to the receiving antenna of the system after being reflected on the target surface and is received. The reflected wave and the emitted wave are mixed using a mixer. Due to the stable frequency difference between the reflected wave and the transmitted wave, the output result of the mixer is also a beat frequency signal with a stable frequency difference, whose frequency contains the distance information between the distance sensing system and the target to be measured. A simplified FMCW system block diagram is shown in Figure 1.

There are certain limitations in the distance measurement of the FMCW system. That is, the most basic FFT frequency measurement is performed on the obtained beat frequency signal. The distance measurement resolution will be limited by the bandwidth used by the FMCW system. The smaller the bandwidth, the lower the distance resolution. Difference. For a 1GHz FMCW system, the distance resolution is 15cm. At this resolution, if the distance difference between two measurements is less than 15cm, the measurement results obtained will be the same, thus introducing measurement errors.

Generally speaking, in order to obtain high-precision distance measurement results and avoid interference with other communication signals, FMCW systems will work at higher radio frequency frequencies, such as 24GHz and 80GHz, to obtain larger bandwidths, such as above 1GHz, to Improve distance measurement resolution.

If in some application scenarios, such as when the FMCW system and the communication system are co-deployed and the transceiver channel is multiplexed, the system operating frequency is at 2.4GHz or 5GHz. In order to ensure the communication quality, the hardware design only provides an operating bandwidth no higher than 100MHz. , at this time, the distance measurement accuracy of the FMCW system has great limitations. When using the FFT frequency measurement method, the distance resolution is at least 3m. If the system has high-precision distance measurement requirements in such an environment, such as a measurement accuracy of not less than 1cm, additional algorithms need to be used to improve the measurement accuracy of the FMCW system.

Contents of the invention

In view of the above-mentioned deficiencies in the prior art, the present invention proposes a high-precision FMCW system distance measurement method.

The technical solution of the present invention is:

A high-precision FMCW system distance measurement method includes the following steps:

S1. Let the FMCW system transmit the signal s(t) to the surface of the object to be measured. The waveform of the transmitted signal s(t) is:

Among them, B is the total bandwidth, T is the signal duration, f ₀ is the working frequency point at resting time, is the system phase:

S2. The FMCW system receives the signal reflected from the object to be measured. The received signal r(t) is:

Among them, t _d is the time delay of the waveform:

d is the distance between the object to be measured and the system, c is the speed of light in vacuum;

S3. Mix s(t) and r(t) to obtain the beat frequency signal b(t) as:

S4. Sampling the beat frequency signal b(t). In order to prevent spectrum aliasing of the signal during subsequent processing, the sampling rate is set to 2f _s to obtain a 2N point sequence.

to the sequence obtained They are processed in the following two ways:

The first way to deal with it is Truncate the first N points and pad the next N points with zeros to obtain b _extn [n]:

Perform a 2N-point FFT on b _extn [n] to get D _extn [k]:

The highest amplitude line of this spectrum corresponds to:

k _extn = [Bt _d ]

[] means rounding to the nearest integer, and the spectral line phase is:

δ represents the error between the spectral line and the actual frequency;

The second way to deal with it is to Perform double downsampling to obtain/>

The sampling rate of this sequence corresponds to f _s , for Perform N-point FFT to get B[k]:

The highest amplitude line of this spectrum corresponds to:

k _m =[Bt _d ]

The phase of the spectral line is:

S5. Difference Φ _fs and Φ _extn , and correct the difference according to the following formula to obtain ΔΦ:

According to the formula The solution of δ is δ, which can be brought into the following formula:

Get the distance measurement result d.

The beneficial effects of the present invention are: the method of the present invention samples the beat frequency signal at two sampling rates and divides it into two-way processing. The phase difference obtained by the two-way processing is used to obtain the frequency difference, and the frequency difference is used to obtain the enhanced distance measurement result. This can significantly improve the distance measurement accuracy of FMCW systems under low bandwidth.

Description of the drawings

Figure 1 is a simplified FMCW system block diagram.

Detailed ways

The technical solution of the present invention will be described in detail below with reference to the accompanying drawings.

As shown in Figure 1, it is the signal transmission process of the FMCW system. The waveform characteristic of the FMCW system is that its frequency changes linearly with time. Consider a system with duration T, total bandwidth B, and working frequency point f ₀ at rest. Its frequency can be expressed as:

The system phase is expressed as the integral of frequency, then the system phase is expressed as:

Use complex signal form to represent the FMCW system waveform:

The signal is transmitted to the surface of the object to be measured through the transmitting antenna, is reflected on the surface of the object, and is transmitted back to the receiving antenna of the device. At this time, the waveform has experienced two path propagation delays, and the final time delay of the waveform is:

where d is the distance between the object to be measured and the system, and c is the speed of light in vacuum. After a time delay of t _d , the signal received by the system should be expressed as:

By mixing s(t) and r(t), the beat frequency signal can be obtained:

It can be seen from the above formula that the frequency of the beat signal is Among them, B and T are constant terms that have been determined before generating the FMCW waveform. Therefore, by frequency resolution of the mixed signal, the propagation time can be estimated and the distance measurement result can be further obtained.

Generally, frequency resolution will be performed in the digital domain. The first step of this process is to sample the beat signal b(t). Let the sampling last for T. The number of points obtained is N, and the corresponding sampling rate is f _s :

Perform an N-point FFT on this sequence:

When the spectrum reaches the maximum amplitude, the corresponding spectral line k _m satisfies:

k _m =[Bt _d ]#(2-9)

In the formula, [] means rounding to the nearest integer. Using FFT to resolve frequency is to determine the frequency of the signal by finding the spectral line with the largest amplitude. After obtaining the spectrum line, the frequency corresponding to the spectrum line is considered to be the target frequency, and the distance measurement result is:

However, it can be seen from this derivation that there is always an error δ between the spectral line and the actual frequency:

δ＝ _Btd -k#(2-11)

Since the spectral line number is a discrete integer, the range of δ is The range of distance measurement error due to δ is/> That is, the distance resolution under this scheme is/> For a 1GHz bandwidth FMCW system, the distance resolution is 15cm, which is not suitable for higher-precision measurement tasks. In view of this, the present invention proposes the following method to improve the measurement accuracy.

First, the beat frequency signal b(t) is sampled at twice the sampling rate to obtain a 2N point sequence:

The obtained sequence is processed in two ways. The first way is to Truncate the first N points and pad the next N points with zeros:

Perform 2N-point FFT on this signal:

At this time, the highest amplitude line of the spectrum corresponds to:

k _extn =[Bt _d ]#(2-15)

The phase of this spectral line is:

Another path for the beat frequency signal is obtained by downsampling the signal twice. Then its expression form is the same as (2-7), and the corresponding spectrum is (2-8). Note that under this spectrum, the spectral line number with the highest amplitude is also [Bt _d ]. Get the phase of this spectral line:

Comparing (2-16) and (2-17), the difference between the two is:

From (2-18), it can be concluded that the phase difference between the FFT spectral line phases of the downsampling and zero-padding sequences only contains the difference δ between the real frequency and the spectral line. Therefore, the actually obtained signal can be processed to obtain the measurement value of δ and enhance the distance measurement results. The results are as follows:

Since in actual measurement, With Φ _extn is obtained by using the arctan function on the complex values of the FFT spectrum, the result is limited to [-π, π], this effect is called phase ambiguity. The range of ΔΦ itself is/> There is no need to consider the phase blur problem, but/> Both Φ extn and Φ _extn will be affected by phase ambiguity. Considering the interference of noise, certain corrections are made to ΔΦ:

The overall process of the present invention is as follows:

1) Use 2f _s to sample the beat signal b(t) to obtain

2) By pairing Perform downsampling and truncation and zero padding to obtain/> with b _extn [n].

3) Perform FFT on the two to obtain the corresponding spectrum and the highest amplitude spectral line k _m .

4) Obtain the phase of the highest amplitude spectral line between the two with Φ _extn .

5) will Make the difference with Φ _extn , and correct the difference according to Equation (2-20) to obtain ΔΦ.

6) Obtain δ from the back solution of equation (2-18).

7) Obtain the enhanced estimate of distance from equation (2-19).

Claims (1)

1. A high-precision FMCW system distance measurement method, which is characterized by including the following steps: S1. Let the FMCW system transmit the signal s(t) to the surface of the object to be measured. The waveform of the transmitted signal s(t) is: Among them, B is the total bandwidth, T is the signal duration, f ₀ is the working frequency point at resting time, is the system phase: S2. The FMCW system receives the signal reflected from the object to be measured. The received signal r(t) is: Among them, t _d is the time delay of the waveform: d is the distance between the object to be measured and the system, c is the speed of light in vacuum; S3. Mix s(t) and r(t) to obtain the beat frequency signal b(t) as: S4. Sampling the beat frequency signal b(t). In order to prevent spectrum aliasing of the signal during subsequent processing, the sampling rate is set to 2f _s to obtain a 2N point sequence. to the sequence obtained They are processed in the following two ways: The first way to deal with it is Truncate the first N points and pad the next N points with zeros to obtain b _extn [n]: Perform a 2N-point FFT on b _extn [n] to get D _extn [k]: Among them, the spectrum line with the highest amplitude corresponds to: k _extn = [Bt _d ] [] means rounding to the nearest integer, and the spectral line phase is: δ represents the error between the spectral line and the actual frequency; The second way to deal with it is to Perform double downsampling to obtain/> The sampling rate of this sequence corresponds to f _s , for Perform N-point FFT to get B[k]: The highest amplitude line of this spectrum corresponds to: k _m =[Bt _d ] The phase of the spectral line is: S5, will Make the difference with Φ _extn , and correct the difference according to the following formula to get ΔΦ: According to the formula Solve for δ and put it into the following formula: Get the distance measurement result d.