DE102021111820A1 - LiDAR sensor system - Google Patents

Abstract

The present invention relates to a LiDAR sensor system (1) having a transmitter unit (3) with a laser source (2), a phase modulator (7) for modulating a phase of the light from the laser source (2) with a phase code (110) and a Transmitting optics (5) for emitting the light modulated by the phase modulator (7) into an environment (10) of the LiDAR sensor system (1), a receiving unit (4) with receiving optics (6) for receiving on an object (11 ) of the surroundings (10) and with an evaluation unit (8) for evaluating the light received by the receiving optics (6), the phase code (110) containing at least one repetition of a large number of nested individual codes (A, B), the Individual codes (A, B) have different code lengths (a, b), and wherein the evaluation unit (8) is designed to divide a received light signal (120) into the individual codes (A, B) of the phase code (110) based on at least one repeated single code (A, B ), in particular on the basis of each repeated individual code (A, B), to determine a Doppler frequency (fD), and on the basis of a correlation of the received light signal (120) with the phase code (110) corrected by the Doppler frequency (fD) to determine a distance (d) to the object (11).

Description

The present invention relates to a LiDAR sensor system. The LiDAR sensor system can be used in particular in a vehicle. The LiDAR sensor system enables reliable Doppler compensation.

State of the art

LiDAR sensor systems are known from the prior art. These mostly work according to the "direct detection" principle, in which an intensity of a reflected light is detected in order to infer an object in the environment. In addition, LiDAR systems are known that have coherent receivers. Such systems emit frequency-modulated signals in order to be able to detect a distance from the detected object on the one hand and a speed of the detected object on the other. The underlying method "frequency modulated continuous wave (FMCW)" is already known from radar technology. As an alternative to frequency modulation, phase codes can also be used, in which the phase of the emitted light is modulated. Such a method is, for example, from WO 2018/144853 A1 known.

Disclosure of Invention

The LiDAR system according to the invention enables the reliable estimation of distances and speeds of targets with phase-modulated light signals. This is achieved in that a received light signal can be processed in a simplified manner by suitably selecting a phase code. In this way, in particular, a reliable and low-effort Doppler estimation is made possible, which ensures a subsequent reliable determination of the light propagation time, also called time of flight, and thus the distance of an object from the LiDAR system.

The LiDAR sensor system has a transmitter unit and a receiver unit. The transmission unit in turn has a laser source, a phase modulator and transmission optics. The phase modulator serves to modulate a phase of the light from the laser source with a phase code. The transmission optics are used to emit the light modulated by the phase modulator into an area surrounding the LiDAR sensor system. The light emitted by the transmitter unit into the environment thus has a phase code that is introduced by the phase modulator.

The receiving unit has receiving optics and an evaluation unit. The receiving optics are used to receive light from the environment. In particular, it is thus possible to receive reflected light that was emitted by the transmission unit and reflected on an object in the vicinity. The evaluation unit is used to evaluate the light received by the receiving optics, with the received light representing in particular a light signal.

As already described above, the received light signal is time-shifted, with the distance to the object reflecting the light being reflected in the time shift, and also frequency-shifted due to the Doppler effect. An FFT analysis of the received signal would result in a Doppler frequency that is difficult or impossible to detect due to the spectrum of the phase code. It is therefore provided that a suitably selected phase code is used, which enables Doppler estimation to be carried out reliably and with little effort. For this purpose, the phase code contains at least one repetition of a multiplicity of interleaved individual codes. Two individual codes are particularly advantageously interleaved in order to form the phase code. All the individual codes used have different code lengths.

The evaluation unit is designed to split a received light signal into the individual codes of the phase code and to determine a Doppler frequency using at least one repeated individual code, in particular using each repeated individual code. If the light signal is divided into the individual codes, the repetition of the individual codes is an aid to estimating the Doppler frequency. The repetition of the individual codes enables a Doppler estimation unaffected by the phase code itself, since the data values along the code repetition times always have the same phase values.

If the Doppler frequency is known, the evaluation unit is designed to determine a distance to the object based on a correlation of the received light signal with the phase code corrected by the Doppler frequency. A two-stage evaluation of the received light signal is therefore provided, with the Doppler frequency being estimated in a first stage based on the known repetition of the individual codes in order to subsequently determine the light propagation time and thus the distance to the object. Both stages can be carried out with little effort, so that the demands on the computing power are minimized, particularly when using a control unit of a vehicle as the evaluation unit. At the same time, however, a reliable determination of the distance is achieved.

The dependent claims show preferred developments of the invention.

It is preferably provided that the individual codes each have a plurality of code symbols. The code length of each individual code corresponds to the number of code symbols. The phase code is formed by alternatingly lining up the code symbols from the respective individual codes. As soon as all code symbols of an individual code have been used, this individual code is repeated. In this way, the individual codes are interleaved and form the phase code.

The evaluation unit is advantageously designed to determine a spectrum from the same repeated code symbols of the received light signal for at least one code symbol per individual code. It is known that the individual codes are interleaved, so that each code symbol in the received light signal is repeated in the phase code after a respective repetition period of the associated individual code. If only these repeated code symbols are considered as a signal, then the Doppler shift of the received light signal can be isolated since the phase code always has the same phase value and is therefore effectively not present in the signal under consideration. The evaluation unit is therefore designed to estimate the Doppler frequency based on the determined spectra.

The evaluation unit particularly advantageously has a photodetector and evaluation logic. The evaluation logic includes, in particular, an analog/digital converter and a digital signal processor.

The individual codes preferably have the same bandwidth. In this way, the processing of the received light signal for estimating the Doppler frequency is simplified. In particular, the necessary computing power of the evaluation unit is minimized.

Each individual code is preferably formed by a biphase code or by a polyphase code. The biphase code is preferably a maximum length sequence or a Gold sequence or a Kasami sequence. The polyphase code is in particular a Frank code or a P1 code or a P2 code or a P3 code or a P4 code or a Chu-Zadoff code.

The code lengths are integers. In addition, it is preferably provided that the code lengths do not have any common prime factors and/or the code lengths are preferably prime numbers. This enables an unambiguous Doppler estimation with little effort.

A total duration of the phase code is preferably between 3 μs and 20 μs. Such an overall duration enables robust signal processing for typical applications.

Provision is preferably made for the evaluation unit to be a simple demodulator. Alternatively or additionally, the evaluation unit is a complex demodulator. Even the use of a simple demodulator enables a reliable distance estimation and at least an absolute estimation of the speed of the detected object in the area. If a complex demodulator is used, it is also possible to estimate the speed of the object with the correct sign.

The invention also relates to a vehicle. The vehicle has a LiDAR sensor system as previously described. This ensures safe and reliable detection of objects in the area surrounding the vehicle, with the costs for the vehicle being minimized due to the structure of the LiDAR sensor system described above.

character list

• 1 eine schematische Abbildung eines Fahrzeugs gemäß einem Ausführungsbeispiel der Erfindung,
• 2 eine schematische Ansicht einer ersten Alternative des LiDAR-Sensor-Systems des Fahrzeugs gemäß dem Ausführungsbeispiel der Erfindung,
• 3 eine schematische Ansicht eines Aufbaus eines verwendeten Phasencodes,
• 4 eine schematische Darstellung der Dopplerschätzung im Frequenzraum,
• 5 einen schematischen Ablaufplan der Signalverarbeitung zur Dopplerschätzung,
• 6 eine schematische Ansicht einer zweiten Alternative des LiDAR-Sensor-Systems des Fahrzeugs gemäß dem Ausführungsbeispiel der Erfindung,
• 7 eine schematische Darstellung einer Variante der Dopplerschätzung, und
• 8 eine schematische Darstellung einer Variante der Signalverarbeitung zur Dopplerschätzung.

Exemplary embodiments of the invention are described in detail below with reference to the accompanying drawings. In the drawing is:

• 1 a schematic illustration of a vehicle according to an embodiment of the invention,
• 2 a schematic view of a first alternative of the LiDAR sensor system of the vehicle according to the embodiment of the invention,
• 3 a schematic view of a structure of a phase code used,
• 4 a schematic representation of the Doppler estimation in the frequency domain,
• 5 a schematic flow chart of the signal processing for Doppler estimation,
• 6 a schematic view of a second alternative of the LiDAR sensor system of the vehicle according to the embodiment of the invention,
• 7 a schematic representation of a variant of Doppler estimation, and
• 8th a schematic representation of a variant of the signal processing for Doppler estimation.

Embodiments of the invention

1 FIG. 1 schematically shows a vehicle 9 according to an exemplary embodiment of the invention. The vehicle 9 in turn has a LiDAR sensor system 1 . Objects 11 in an environment 10 of the LiDAR sensor system 1 and thus of the vehicle 9 can be detected with the LiDAR sensor system 1 .

The LiDAR sensor system 1 has a transmitter unit 3 and a receiver unit 4 . The transmitter unit 3 is used to emit a light signal which is reflected by the object 11 so that this reflected signal can be received by the receiver unit 4 . Based on a time shift between the transmitted signal and the received signal, conclusions can be drawn about the light propagation time and thus about the distance d between LiDAR sensor system 1 and object 11 . The Doppler effect that occurs also enables the speed v at which the object 11 is moving to be estimated.

In principle, it is provided that the LiDAR sensor system 1 works according to coherent phase code approaches, so that the phase of an emitted laser light is varied. A first alternative for the implementation of such a LiDAR sensor system 1 is in 2 shown. As described above, the LiDAR sensor system 1 according to the first alternative has the transmission unit 3 and the reception unit 4 . The transmission unit 3 comprises a laser source 2, a phase modulator 7 and a transmission optics 5. The reception unit 4 has a reception optics 6 and an evaluation unit 8.

In order to emit a light signal by means of the transmitter unit 3, it is initially provided that laser light is emitted by means of the laser source 2. This light is guided by a splitter 12 both to the receiving unit 4 and to the phase modulator 7 . The phase modulator 7 is used to modulate the phase of the light from the laser source 2 in order to generate a phase code 110 in this way. The light modulated in this way is then emitted by the transmission optics 5 into the environment 10 .

The emitted light is reflected by an object 11 in the surroundings 10 and thus reaches the receiving optics 6 of the receiving unit 4. From there, the received light signal reaches a coupler 13, which couples in the light from the laser source 2 branched off via the splitter 12. Finally, the received signal can be evaluated via the evaluation unit 8 . The evaluation unit 8 has a photodetector 8a and evaluation logic with an analog/digital converter 8b and a digital signal processor 8c.

The receiving unit 4 thus receives a light signal 120 which is time-delayed and Doppler-shifted in relation to the transmitted signal and thus in relation to the phase code 110 . To determine the distance d of the object 11, the Doppler frequency is therefore first estimated in order to then determine the distance d.

k:

Digitale Laufvariable k ∈ N₀,

Ts:

Abtastintervall; Entspricht invertiert der Abtastzeit: f_s = 1/T_s,

Â:

Signalamplitude, beinhaltet Dämpfung über Signalpfad, Verstärkung im Mischprozess und durch Verstärker, etc.,

ωD:

Kreisfrequenz der Doppler-Frequenz:

ω0τ:

Phasenverschiebung resultierend aus Time of Flight bezogen auf Kreisträgerfrequenz,

n(k):

Zusammengefasstes diskretisiertes Rauschen (Schrotrauschen, thermisches Rauschen, Phasenrauschen),

ϕ(kTs - τ):

Zeitverschobener Phasencode.

The starting point is thus the received light signal 120, which was received by the receiving unit 4 from the environment. Also known is the phase code 110 used to generate the transmit sequence. The phase code 110 is, for example, a biphasic PN code sequence modulated by the phase modulator 7 onto a carrier. The transmission sequence generated in this way is transmitted and attenuated, time-delayed and Doppler-shifted as a light signal 120 and received again. The light signal 120 can be described mathematically as a digital signal c _RX as follows: $$c_{RX}\left(k\right)={\widehat{A}}_C\cdot cOs\left({\omega }_{D}k{T}_S+{\omega }_0\tau -\varphi \left(k{T}_S-\tau \right)\right)+n_C\left(k\right)$$

c:

digital running variable k ∈ N ₀ ,

Ts:

sampling interval; Corresponds to the inverse of the sampling time: f _s = 1/T _s ,

Â:

Signal amplitude, includes attenuation over the signal path, amplification in the mixing process and through amplifiers, etc.,

ωD:

Angular frequency of the Doppler frequency:

ω0τ:

Phase shift resulting from time of flight related to circular carrier frequency,

n(k):

Summarized discretized noise (shot noise, thermal noise, phase noise),

ϕ(kTs - τ):

Time-Shifted Phase Code.

a:

Länge des ersten Einzelcodes A; a ∈ N,

b:

Länge des zweiten Einzelcodes B; b ∈ N,

NSC:

Länge eines Einzelcodes allgemein, wobei SC für single code, also Einzelcode, steht; N_SC ∈ N,

TA:

Zeitdauer des ersten Einzelcodes A,

TB:

Zeitdauer des zweiten Einzelcodes B,

Tx:

Zeitdauer eines Einzelcodes allgemein,

T'A:

Zeitdauer des ersten Einzelcodes A im Phasencode 110,

T'B:

Zeitdauer des zweiten Einzelcodes B im Phasencode 110,

T'x:

Zeitdauer eines Einzelcodes allgemein im Phasencode 110,

BC:

Code-Bandbreite allgemein,

BCn:

Code-Bandbreite des n-ten Codes; n ∈ N,

NTC:

Länge des Phasencodes 120, wobei TC für total code, also Gesamtcode, steht; N_TC ∈ N,

TTC:

Zeitdauer des Phasencodes 110,

τ

Time of Flight / Signallaufzeit,

τmax

maximal mögliche Time of Flight / Signallaufzeit,

d

Entfernung zum Objekt 11,

dmax

maximal zulässige bzw. zu erwartende Entfernung zum Objekt 11,

fD

Dopplerfrequenz,

fD,max

maximal zu erwartende Dopplerfrequenz,

c0

Lichtgeschwindigkeit im Vakuum,

Tunamb

Zeitdauer, über welche der Phasencode 110 eindeutig ist.

For a better understanding of the following explanations, the following formula symbols are also clearly introduced

a:

Length of the first single code A; a ∈ N,

b:

length of the second single code B; b ∈ N,

NPC:

Length of a single code in general, where SC stands for single code; N _SC ∈ N,

TA:

Duration of the first individual code A,

TB:

Duration of the second individual code B,

Tx:

Duration of an individual code in general,

T'A:

Duration of the first individual code A in phase code 110,

T'B:

Duration of the second individual code B in phase code 110,

T'x:

Duration of a single code generally in phase code 110,

BC:

code bandwidth in general,

BCn:

code bandwidth of the nth code; n ∈ N,

NTC:

length of the phase code 120, where TC stands for total code; N _TC ∈ N,

TTC:

Duration of phase code 110,

τ

Time of Flight / signal propagation time,

τmax

maximum possible time of flight / signal propagation time,

i.e

distance to object 11,

d max

maximum permissible or expected distance to the object 11,

fD

doppler frequency,

fD,max

maximum expected Doppler frequency,

c0

speed of light in vacuum,

Tunamb

Amount of time over which phase code 110 is unique.

3 12 schematically shows the structure of the phase code 110. This is formed by mixing a first individual code A and a second individual code B. The first individual code A has a length a and the second individual code B has a length b, where the following applies: a≠b. The first individual code A and the second individual code B have the same bandwidth B _C .

in the in 3 In the example shown, the first individual code A has a length a=6, so that six first code symbols S _A1 ...S _A6 are present. The second individual code B has a length of b=8, so that eight second code symbols S _B1 . . . S _B8 are present. The first code symbols S _A1 . . . S _A6 and the second _{code symbols} S _B1 . Even if in 3 only two individual codes, ie the first individual code A and the second individual code B, are used to form the phase code 110, then several individual codes can also be mixed together in order to form the phase code 110. Another alternative is that the bandwidths B _Cn of the individual codes are different.

In general, the duration T _x of an individual code results from the length N _SC of the individual code and the bandwidth B _C of the individual code: $$T_x=\frac{N_{SC}}{B_C}$$

In the example according to 3 in order to: $$T_A=\frac{a}{B_C}{T}_B=\frac{b}{B_C}$$

In general, the repetition time of an individual code in the phase code 110 generated by mixing n individual codes also results accordingly: $$T{\text{'}}_x=N_{SC}\left(\frac{1}{{\mathrm{<figure-callout\; id="1"\; label="B\; C"\; filenames="DE102021111820A1\_0021.png,DE102021111820A1\_0023.png"\; state="\{\{state\}\}">B</figure-callout>}}_C}+\frac{1}{{\mathrm{<figure-callout\; id="2"\; label="B\; C"\; filenames="DE102021111820A1\_0021.png,DE102021111820A1\_0025.png"\; state="\{\{state\}\}">B</figure-callout>}}_C}+...+\frac{1}{B_{Cn}}\right)$$

In the example according to 3 in order to: $$T{\text{'}}_A=\frac{2a}{B_C}T{\text{'}}_B=\frac{2b}{B_C}$$

• • Biphasencodes: „Maximum Length Sequences“ (MLS), Gold- oder Kasami-Sequenzen
• • Polyphasencodes: Frank-, P1, P2-, P3-, P4- oder Chu-Zadoff-Codes.

The type of individual code can be freely selected. These can be biphasic or polyphasic and do not have to be of the same type. Examples of applicable codes are:

• • Biphase codes: Maximum Length Sequences (MLS), Gold or Kasami sequences
• • Polyphase codes: Frank, P1, P2, P3, P4 or Chu-Zadoff codes.

ist und zum anderen eine eindeutige Dopplerschätzung synthetisiert werden kann, was nachfolgend erläutert wird. Am einfachsten lässt sich dies erreichen, wenn für die Längen a und b Werte genutzt werden, die keine gemeinsamen Primfaktoren besitzen, was beispielsweise bei Primzahlen der Fall ist. Somit lassen sich zwar beliebige Werte für die Längen a und b der Einzelcodes A, B verwenden, solange diese Längen a und b unterschiedlich sind, besonders vorteilhaft ist jedoch die Verwendung solcher Längen a und b ohne gemeinsame Primfaktoren.It is only provided that the individual codes have different lengths a and b. These lengths a and b are selected in such a way that, on the one hand, the phase code 110 is unambiguous over the period of time $${\tau }_{max}=\frac{2{\mathrm{i.e}}_{max}}{c_0}$$

and on the other hand a unique Doppler estimate can be synthesized, which will be explained below. The easiest way to achieve this is to use values for the lengths a and b that do not have common prime factors, which is the case with prime numbers, for example. Thus, although any values for the Län Use gen a and b of the individual codes A, B as long as these lengths a and b are different, but the use of such lengths a and b without common prime factors is particularly advantageous.

The uniqueness period T _unamb of the phase code 110 results for a mixed code from two individual codes of the same bandwidth Be with individual code lengths a and b as follows: (lcm = least common multiple, ie smallest common multiple): $$T_{\mathrm{and}namb}=2\cdot lcm\left(a,b\right)\cdot B_C^{-1}$$

und liegt für die beabsichtigte Anwendung typischerweise im Bereich zwischen 3 µs bis 20 µs. Vorzugsweise wird eine einheitliche Code-Bandbreite verwendet, welche gleich der Abtastfrequenz des Analog-Digital-Umsetzers 8b ist. Somit erhält man pro Sample des Analog-Digital-Umsetzers 8b ein Symbol S_A1...S_A6, S_B1-S_B8 des Phasencodes 110, was zu einer einfacheren Signalverarbeitung führt. Hintergrund hierfür ist, dass das empfangenen Lichtsignal 120 später in der Signalverarbeitung entsprechend der Code-Bandbreite B_C unterabgetastet wird, um die Einzelcodes trennen zu können. Existieren mehrere Samples pro Symbol S_A1...S_A6, S_B1...S_B8 des Phasencodes 110, muss diese Unterabtastung mehrfach mit Versatz geschehen, um alle Abtastwerte nutzen zu können. Diese mehrfache Prozessierung erhöht die Komplexität und kann durch B_C = f_s vermieden werden.In the case of a uniform bandwidth B _C , the total duration of the phase code 110 results in $$T_{TC}=\frac{N_{TC}}{B_C}$$

and typically ranges from 3 µs to 20 µs for the intended application. A uniform code bandwidth is preferably used, which is equal to the sampling frequency of the analog/digital converter 8b. _{One symbol S A1} . The background to this is that the received light signal 120 is subsampled later in the signal processing according to the code bandwidth B _C in order to be able to separate the individual codes. If there are several samples per symbol S _A1 . . . S _A6 , _{S B1} . This multiple processing increases the complexity and can be avoided by _BC = f _s .

In order to carry out a Doppler estimation, the data samples along the code repetition times T' _x are used in each case, since this ensures that the time-shifted code φ in the received light signal 120 always has the same symbol S _A1 . . . S _A6 , S _B1 . S _B8 and thus has the same phase position. This means that the received light signal 120 is sampled with T′ _x .

Effectively, therefore, one no longer sees a code, but only the Doppler frequency f _D .

Since Doppler shifts f _D in the range of +/- 100 MHz can occur in coherent lidar systems for automotive applications, the individual Doppler estimates obtained over the code repetition times will generally be ambiguous, since aliasing occurs because the Nyquist sampling theorem is not observed becomes: $${ƒ}_{Dmax}>\frac{2}{T\text{'}x}$$

$${ƒ}_D=m\cdot \frac{1}{T{\text{'}}_B}\pm {ƒ}_{Meas2}$$

$$n,m\in {\mathbb{N}}_0\mathrm{\Lambda }n{T}_B^{\text{'}-1}\le {ƒ}_{D,max}$$

in the in 3 In the example shown, in which a first individual code A and a second individual code B are mixed, two frequencies are estimated in this way, a first estimated frequency f _Meas1 and a second estimated frequency f _Meas2 . The first estimated frequency f _Meas1 and the second estimated frequency f _Meas2 are related to the true Doppler frequency f _D as follows. $${ƒ}_D=n\cdot \frac{1}{T{\text{'}}_A}\pm {ƒ}_{Mea\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="DE102021111820A1\_0021.png,DE102021111820A1\_0023.png"\; state="\{\{state\}\}">s</figure-callout>}1}$$

$${ƒ}_D=m\cdot \frac{1}{T{\text{'}}_B}\pm {ƒ}_{Mea\mathrm{<figure-callout\; id="2"\; label="s"\; filenames="DE102021111820A1\_0021.png,DE102021111820A1\_0025.png"\; state="\{\{state\}\}">s</figure-callout>}2}$$

$$n,m\in {\mathbb{N}}_0\mathrm{\Lambda }n{T}_B^{\text{'}-1}\le {ƒ}_{D,max}$$

If the information from the two individual estimates is combined, an increased uniqueness range 300 can be achieved from uniqueness ranges 100, 200 of the individual measurements, in particular as a function of the individual sampling intervals. 4 shows the facts in the frequency domain as an example. In this example, the actual Doppler frequency f _D can be calculated unambiguously from f _Meas1 and f _Meas2 using the equations described above. For this purpose, all possible alias frequencies are calculated and compared using the run variables n, m. Where the absolute value of the difference is minimal, the true Doppler frequency is f _D with the greatest probability. Thus the Doppler frequency f _D is estimated.

This procedure generally always leads to an unambiguous Doppler estimate as long as the Doppler frequency f _D is in the combined unambiguous range 300 and there is only one dominant target. In general, the combined sampling frequency f _syn : $${ƒ}_{syn}=lcm\left(T_{x1}^{\text{'}-1},T_{x2}^{\text{'}-1}\right)$$

The combined uniqueness range thus corresponds to f _syn /2 (the abbreviation lcm used again means least common multiple, ie smallest common multiple). Based on the example 3 becomes f _syn : $${ƒ}_{syn}=lcm\left(\frac{B_c}{2a},\frac{B_c}{2b}\right)=\frac{B_c}{2}lcm\left(\frac{1}{a},\frac{1}{b}\right)=\frac{{\mathrm{<figure-callout\; id="2"\; label="B\; c"\; filenames="DE102021111820A1\_0021.png,DE102021111820A1\_0025.png"\; state="\{\{state\}\}">B</figure-callout>}}_c}{2\cdot Gc\mathrm{i.e}\left(a,b\right)}$$

The abbreviation gcd used stands for greatest common divisor. It can be seen that the combined Sampling frequency here can be a maximum of B _C /2. For this, the greatest common divisor of the code lengths a and b must be 1, which is the case for numbers that have no common prime factors.

in the in 4 In the exemplary embodiment shown, two individual codes A, B are used to form the phase code 110 . Alternatively, more than two individual codes can also be used. This increases the robustness and precision of the Doppler estimation. At the same time, the uniqueness range decreases, so that for n individual codes with the same bandwidth B _C and with lengths without common prime factors, the following applies: $${ƒ}_{syn,max}=\frac{B_C}{n}$$

5 shows a flow chart of the signal processing for Doppler estimation. In the following explanations, individual codes of the same bandwidth B _C are again assumed.

In a first step S1, the received light signal 120 is separated into the individual codes used. As previously described, this is done by sampling the received light signal 120 with Bc (or B _C /2). Due to the unknown time of flight or signal propagation time τ, the separation of the individual codes in the first step S1 does not take place deterministically. This means that the received light signal 120 can be separated and it is known that the individual parts of the divided light signal 120 each contain an individual code, but which code is not known. Depending on the number n of mixed individual codes, n hypotheses are obtained. The subsequent steps S2 to S5 are carried out for each hypothesis, whereby the correct hypothesis will give the best results and can thus be identified in the end.

In a second step S2, the sampled values are rearranged into matrix structures according to the individual repetition times T _x of the individual codes. A first code matrix 400 is thus formed for the repetitions of the first individual code A and a second code matrix 500 for the repetitions of the second individual code B. Then, in a third step S3, an FFT calculation is performed over the code repetitions, ie along the columns of the respective code matrix 400, 500.

The spectra obtained in the third step S3 are summed in a fourth step S4. The summation of the spectra in the fourth step S4 is most simply non-coherent, i.e. the amounts of the spectra are added up. However, better performance can be achieved with coherent addition without absolute value formation, which, however, requires a phase correction so that the frequency to be measured is always added constructively. Here, among other things, the phase change of the code must be corrected, which is associated with increased computing effort.

Finally, in a fifth step S5, a peak search is carried out in the resulting spectra, as a result of which the first estimated frequency f _Meas1 and the second estimated frequency f _Meas2 are determined. From this, the Doppler frequency f _D can be calculated, as explained above.

After the Doppler frequency f _D has been estimated, in the next step the distance d can now be determined by means of correlation between the corrected code template c _T and the received light signal 120, represented below by the symbol c _RX . If the Doppler estimation from the previous step is ambiguous, which can occur in special cases, especially when only two codes are mixed, the possible results can be checked for plausibility with the help of the correlation, since only the true Doppler frequency f _D leads to a successful correlation will lead.

Two code templates c _T are preferably correlated with the received light signal 120 c _RX since this leads to maximum detection performance. Correlation with only one code template c _T is also possible. When using two code templates c _T , the first code template is preferred: $$c_{TC}\left(k\right)=cOs\left({\tilde{\omega }}_{D}k{T}_S-\varphi \left(k{T}_S\right)\right)$$

ω̃_D beschreibt dabei jeweils die zuvor geschätzte Dopplerfrequenz f_D. Die Korrelation kann nun gleichermaßen für beide Templates mathematisch wie folgt beschrieben werden, wobei c_T jeweils für die Kosinus- oder die Sinus-Version des Templates steht: $$R_{SC}\left(u\right)={\displaystyle {\sum }_{k=0}^{N_{CS}}{c}_{RX}}\left(k\right)c_r\left(k-u\right)$$

N_CS ist hierbei die Länge des Phasencodes 110 und ergibt sich über: N_CS = T_CSƒS, wobei T_CS die Gesamtdauer des Phasencodes 110 ist. Die beiden Korrelationsfunktionen für die verschiedenen Template-Funktionen können dann folgendermaßen zusammengefasst werden: $$R_{SCges}\left(u\right)=\sqrt{R_{SCsin}^2\left(u\right)+R_{SCcos}^2\left(u\right)}$$

However, since the phase term ω ₀ τ is unknown, a cross-correlation is also performed with the sine version: $$c_{TC}\left(k\right)=cOs\left({\tilde{\omega }}_{D}k{T}_S-\varphi \left(k{T}_S\right)\right)$$

ω̃ _D describes the previously estimated Doppler frequency f _D . The correlation can now be described mathematically for both templates as follows, where c _T stands for the cosine or the sine version of the template: $$R_{SC}\left(\mathrm{and}\right)={\displaystyle {\sum }_{k=0}^{N_{CS}}{c}_{RX}}\left(k\right)c_{\mathrm{right}}\left(k-\mathrm{and}\right)$$

Here, N _CS is the length of the phase code 110 and results from: N _CS =T _CSƒS , where T _CS is the total duration of the phase code 110 . The two correlation functions for the different template functions can then be summarized as follows: $$R_{SCGes}\left(\mathrm{and}\right)=\sqrt{R_{SCsin}^2\left(\mathrm{and}\right)+R_{SCcOs}^2\left(\mathrm{and}\right)}$$

The signal propagation time or time of flight τ can be determined using the correlation, which in turn is a measure for the distance dist.

6 shows schematically a second variant of the LiDAR sensor system 1 of the vehicle 9 according to the exemplary embodiment of the invention. The transmission unit 3 is analogous to the first variant as previously described. There is a difference in the receiving unit 4, which is in 3 second variant shown has a complex demodulator 18, 19. In addition, an estimation of the speed v of the object 11 with the correct sign is made possible. For this purpose, a first evaluation unit 18 is provided for the I path and a second evaluation unit 19 for the Q path. The first evaluation unit 18 and the second evaluation unit 19 each have their own photodetector 8a and their own evaluation logic with an analog/digital converter 8b and a digital signal processor 8c.

In addition, the receiving unit 4 has a first additional splitter 15, which splits the reference signal from the splitter 12 of the transmitting unit 3 again in order to transmit it to the respective coupler 13a, 13b of the corresponding paths of the complex demodulator 18, 19. It is provided that a phase offset of 90° is introduced for the Q path. A second additional splitter 14 is also present, which splits the light received from the receiving optics 6 into the two paths of the complex demodulator 18, 19.

It was previously described how the Doppler estimation is based on the repetitions of the individual codes A, B. Likewise, the Doppler estimation can be performed based on the repetition of the entire phase code. This possibility offers an improvement in the precision and robustness of the aforementioned Doppler estimation based on the repetitions of the individual codes A, B. The underlying principle is in 7 shown. Basically the same steps as in 5 shown and applied as described above, but the received light signal 120 is not divided into the individual codes A, B. Rather, the individual phase codes 110 are considered and further processed. In this case, the duration of the repetitive phase code 110 is preferably set equal to the maximum light propagation time or time of flight τ _max to be expected. The associated sampling interval, which is used to separate the phase codes 120 in the received light signal 110, is then obtained as a function of the maximum light propagation time or time of flight τ _max : $${ƒ}_{s,TC}=\frac{1}{{\mathrm{\tau }}_{max}}$$

8th finally shows an alternative form of signal processing. As an alternative to the third step S3 and the fourth step S4 as described above and in 5 shown, one-dimensional signal vectors can also be generated from the code matrices of the second step S2 described above. To do this, the columns of the code matrices are lined up, as in 8th shown as an example for the first code matrix 400 from which a one-dimensional signal vector 401 is formed. A code with the bandwidth BW _res =T ⁻¹ TC, which is also called residual code, remains in the signal vector 401 that is produced in this way. The FFT calculation of this signal vector delivers a spectrum with a significantly increased number of support points compared to the procedure presented before. The bandwidth of the residual code is so small that a reliable frequency estimate is still possible. It is advantageous that the detailed spectrum enables the use of filters, averaging or pattern matching methods, which can increase the detection probability and detection accuracy. After the one-to-one frequencies have been estimated from the frequency spectra, the Doppler frequency ƒ _D is again estimated by the fifth step S5 as previously described.

QUOTES INCLUDED IN DESCRIPTION

This list of documents cited by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Patent Literature Cited

• WO 2018/144853 A1 [0002]

Claims (10)

LiDAR sensor system (1) having • a transmission unit (3) with a laser source (2), a phase modulator (7) for modulating a phase of the light from the laser source (2) with a phase code (110) and transmission optics (5) for emitting the light modulated by the phase modulator (7) into an environment (10) of the LiDAR sensor system (1), • a receiving unit (4) with receiving optics (6) for receiving at an object (11) in the environment (10) reflected light and with an evaluation unit (8) for evaluating the light received by the receiving optics (6), • wherein the phase code (110) contains at least one repetition of a large number of nested individual codes (A, B), • wherein the individual codes (A, B) have different code lengths (a, b), and • wherein the evaluation unit (8) is designed, ◯ divide a received light signal (120) into the individual codes (A, B) of the phase code (110), ◯ based on at least one repeated individual code (A, B), in particular based on vo n to determine a Doppler frequency (f _D ) for each repeated individual code (A, B), and ◯ a distance (d) to a correlation of the received light signal (120) with the phase code (110) corrected by the Doppler frequency (f _D ). to determine the object (11). LiDAR sensor system (1) according to claim 1 , characterized in that the individual codes (A, B) each have a plurality of code symbols (S _A1 ... S _A6 , S _B1 ... S _B8 ), wherein the code length (a, b) corresponds to the number of code symbols, and wherein the phase code (110) is formed by alternating the code symbols (S _A1 ...S _A6 , S _B1 ...S _B8 ) from the respective individual codes (A, B) in a row. LiDAR sensor system (1) according to one of the preceding claims, characterized in that the evaluation unit (8) is designed for at least one code symbol (S _A1 ...S _A6 , S _B1 ...S _B8 ) per individual code ( A, B) to determine a spectrum from the same repeated code symbols (S _A1 ...S _A6 , S _B1 ...S _B8 ) of the received light signal (120), with each code symbol (S _A1 ...S _A6 , S _B1 ... S _B8 ) in the received light signal (120) after a respective repetition period (T' _A , T' _B ) of the associated individual code (A, B) in the phase code (110), and wherein the evaluation unit (8) is formed , to estimate the Doppler frequency (f _D ) from the spectra. LiDAR sensor system (1) according to one of the preceding claims, characterized in that the individual codes (A, B) have the same bandwidth. LiDAR sensor system (1) according to one of the preceding claims, characterized in that each individual code (A, B) by a biphase code, in particular by a maximum length sequence or by a Gold sequence or by a Kasami sequence, or by a polyphase code, in particular a Frank code or by a P1 code or by a P2 code or by a P3 code or by a P4 code or by a Chu-Zadoff code. LiDAR sensor system (1) according to one of the preceding claims, characterized in that the code lengths (a, b) are integers and have no common prime factors and/or are prime numbers. LiDAR sensor system (1) according to one of the preceding claims, characterized in that a total duration (T _TC ) of the phase code (110) is between 3 µs and 20 µs. LiDAR sensor system (1) according to one of the preceding claims, characterized in that the evaluation unit (8) has a photodetector (8A) and evaluation logic, in particular an analog/digital converter (8B) and a digital signal processor (8C), having. LiDAR sensor system (1) according to one of the preceding claims, characterized in that the evaluation unit (8) is a simple demodulator or has a complex demodulator (18, 19). Vehicle (9) having a LiDAR sensor system (1) according to one of the preceding claims.

Images