LIDAR SYSTEM, VEHICLE AND OPERATING PROCEDURES

Abstract

Lidar system (1) with - a first laser (21) and a second laser (22), and - a detection unit (4) for detecting laser radiation from the first and second lasers (21, 22), wherein - the first laser (21) in a first wavelength range (L1) and the second laser (22) in a second wavelength range (L2) are set up for periodic tuning of a respective emission wavelength (E1, E2), - a first tuning time T1 of the first laser (21) differs from a second tuning time T2 of the second laser (22), - holds: m T1 = n T2 with m, n ∈ ℚ \ ℕ. N and m < n as well as n / m ∈ ℚ \ ℕ. where for all i = 1, ..., n and for all j = 1, ..., m: imjn ≤ 0.97 ∧ 1.03 ≤ imjn , as well as m T1 > FR / c and (m - 1) T1 ≤ FR / c with 0.2 ≤ F ≤ 1.5, and R is a specified maximum range of the lidar system (1) and c is the speed of light in a vacuum.

Description

[0001] A lidar system is specified. Furthermore, a vehicle with such a lidar system is specified. Furthermore, a method for operating such a lidar system is specified.

[0002] The document US 2020 / 0 386 886 A1 relates to a laser radar system that uses collocated laser beams to unambiguously determine the range of a target and the range rate at which the target is moving relative to the laser radar system. A laser radar system uses multiple laser radar sections to obtain multiple simultaneous measurements, allowing both the range and the range rate to be determined without various temporal effects introduced by systems that use individual laser sections that take consecutive measurements. Furthermore, a faster and more accurate determination of the target's range and speed can be achieved.

[0003] The publication WO 2020 / 081 188 A1 concerns a lidar system.

[0004] The publication WO 2019 / 205 163 A1 contains a lidar system with multiple emitters and multiple receivers.

[0005] A lidar system in which an angular range is scanned with at least two beams can be found in the publication US 2018 / 0 284 236 A1.

[0006] The documents EP 2 329 218 B1, US 2006 / 0 227 316 A1, US 2020 / 0 271 784 A1 and WO 2020 / 018 805 A1 relate to systems for light-based distance measurement.

[0007] A lidar system with wavelength division multiplexing is known from the publication US 2019 / 0257927 A1.

[0008] One task to be solved is to specify a lidar system with which the scanning time per pixel can be reduced.

[0009] This problem is solved by a lidar system, a vehicle, and an operating method having the features of the independent patent claims. Preferred developments are the subject of the dependent claims.

[0010] The lidar system comprises a first laser and a second laser. The lasers are preferably solid-state lasers, in particular semiconductor lasers. The lasers can be formed by separate, structurally independent lasers or by a laser system with several monolithically integrated laser emission regions on a common semiconductor substrate.

[0011] The first laser in a first wavelength range and the second laser in a second wavelength range are configured for periodic tuning of a respective emission wavelength. For example, a tuning time, also referred to as the period duration or chirp, of each laser is constant. This means that the tuning times preferably do not change over time, as intended. The wavelength ranges preferably do not overlap.

[0012] The lidar system comprises a detection unit for detecting laser radiation from the first and second lasers. In the detection unit, for example, a beat frequency of laser radiation directly from the respective laser and of laser radiation from the same laser that has been reflected by at least one object outside the lidar system is determined. This preferably occurs for each laser independently of the at least one other laser.

[0013] A first tuning time T1 of the first laser differs from a second tuning time T2 of the second laser. In other words, the lasers have different period lengths, in particular, no period lengths that are an integer multiple of each other.

[0014] Thus, the lidar system comprises a first laser and a second laser, as well as a detection unit for detecting laser radiation from the first and second lasers. The first laser is configured for periodic tuning of a respective emission wavelength in a first wavelength range, and the second laser is configured for periodic tuning of a respective emission wavelength in a second wavelength range. A first tuning time T1 of the first laser differs from a second tuning time T2 of the second laser, and / or a wavelength change per unit time of the first laser differs from a wavelength change per unit time of the second laser.

[0015] Because the lasers have different tuning times, the scan time per pixel can be shorter than the light travel time to the object to be detected and back to the lidar system, while still avoiding ambiguities in positioning. This allows for shorter overall scan times for an image and / or allows for a corresponding increase in the number of pixels.

[0016] According to at least one embodiment, T1 > T2 applies and the quotient T1 / T2 lies between 1.02 and 1.98, or between 1.05 and 1.95, or between 2.05 and 2.95, or between 3.05 and 3.95, inclusive. Preferably, the following applies: 1.05 ≤ T1 / T2 ≤ 1.7, or 1.05 ≤ T1 / T2 ≤ 1.6, or 1.1 ≤ T1 / T2 ≤ 1.5. In other words, the tuning times T1 and T2 differ noticeably from one another, but on the other hand, they do not differ too greatly.

[0017] According to at least one embodiment, at least one of the tuning times T1 and T2 is at least a factor of 2, or at least a factor of 3, or at least a factor of 5 smaller than a specified maximum range R of the lidar system divided by the vacuum speed of light c. In other words, T1, T2 ≤ R / 2c or T1, T2 ≤ R / 3c or T1, T2 ≤ R / 5c. Alternatively or additionally, this factor is at most 30, or at most 20, or at most 10, or at most 6.

[0018] According to at least one embodiment, m T1 = n T2 ⇔ T1 = n / m T2 with m, n EN and m < n and n / m ∈ ℚ \ ℕ. In other words, m and n are integers and the fraction m over n is not an integer.

[0019] It holds for all i = 1, ..., n and for all j=1,…,m:imjn≤0.990∧1.010≤imjn or imjn≤0.97∧1.03≤imjn or imjn≤0.95∧1.05≤imjn. This applies up to the specified maximum range R, which can be smaller than m T1 c. This means that m T1 > 2 R / c and (m - 1) T1 ≤ 2 R / c or m T1 > R / c and (m - 1) T1 ≤ R / c, or in general m T1 > FR / c and (m - 1) T1 ≤ FR / c with 0.2 ≤ F ≤ 2 or 0.2 ≤ F ≤ 1.5 or 0.3 ≤ F ≤ 1.2 or 0.4 ≤ F ≤ 0.9 also apply. A value of F < 2 means that T1 = k T2 with k ∈ ℕ only occurs well outside the specified maximum range R. This means that F can be considered as a safety factor, so that potential ambiguities only arise well outside the intended maximum range R.

[0020] In other words, i and j are integer indices ranging from 1 to n and from 1 to m, respectively. For any fraction of i times m and j times n, the fraction must not be too close to one, and this applies within the specified maximum range R of the lidar system. The lidar system typically operates in air. Since the speed of light in air differs only slightly from the speed of light in vacuum, the vacuum speed c is used for simplification to calculate the travel time from a distance.

[0021] For example, m = 7 and n = 9. Thus, T1 = 9 / 7 T2. Up to a distance of 7c T1 > R, there is no integer fraction of m and n. The tuning time periods of the second laser then end at 7 / 9 T1, at 14 / 9 T1, at 21 / 9 T1, at 28 / 9 T1, and so on. The tuning time periods of the first laser then end at 9 / 7 T2, 18 / 7 T2, 27 / 7 T2, and so on. Critical with regard to the unambiguous distinguishability of potential ambiguities are times for which the ratio of the end times of tuning periods within the maximum range are relatively close, for example 4m / 3n = 28 / 27 or 5m / 4n = 35 / 36. Up to at least 3 tuning periods T1, the end times of the tuning periods are close to each other up to a maximum of approximately a factor of 28 / 27 = 1.037, so that ambiguities can be reliably excluded.

[0022] Other example pairs of numbers for m and n are 7 and 11 or 5 and 8.

[0023] According to at least one embodiment, the intended maximum range R is at least 0.1 km and at most 0.5 km, or at least 2 km and at most 10 km. The first range of values ​​applies, for example, to trucks or motor vehicles, and the second range of values ​​applies, for example, to aircraft, ships, or railways.

[0024] According to at least one embodiment, the lidar system is configured to tune the emission wavelengths of the first laser and the second laser in the form of a triangular variation or in the form of a sawtooth variation. With a sawtooth variation, discontinuities and / or indistinguishable points in the temporal progression of the emission wavelengths may occur at the end of the tuning periods. With a triangular variation, at least discontinuities can be avoided.

[0025] According to at least one embodiment, the first laser and the second laser have different tuning slopes. The tuning slopes are defined as the wavelength difference per unit of time within the respective tuning periods. Alternatively, the lasers can also have the same tuning slopes.

[0026] According to at least one embodiment, the lidar system further comprises a third laser and a fourth laser, wherein the third laser is configured in a third wavelength range and the fourth laser is configured in a fourth wavelength range for periodically tuning a respective emission wavelength. The first, second, third, and fourth wavelength ranges differ from each other in pairs and do not overlap.

[0027] Accordingly, the detection unit is also configured to detect laser radiation from the third and fourth lasers in the same way as for the first and second lasers.

[0028] According to at least one embodiment, the third laser and the first laser form a first laser pair, and the fourth laser and the second laser form a second laser pair. Within the laser pairs, the associated tuning times are preferably the same, so that the lasers within each laser pair are then configured to be tuned synchronously. Such laser pairs can be used to measure the distance and speed of objects.

[0029] According to at least one embodiment, the tuning slopes are different within a laser pair. In other words, the lasers of the respective pair are controlled such that different wavelength changes per unit time are present in the respective chip, i.e., the respective wavelength change ramp.

[0030] According to at least one embodiment, the detection unit is configured to detect the wavelength ranges individually and independently of one another. For this purpose, the detection unit can comprise multiple detectors such as photodiodes and / or the detection unit can be a pixelated detector, wherein individual pixels and / or photodiodes can preferably detect the laser radiation of individual lasers in a spectrally selective manner.

[0031] According to at least one embodiment, the lasers are configured to all emit in a specific, common emission direction at a specific time. The detection unit is configured to detect the emission wavelengths from a spatial range that encompasses the emission direction and is larger than an angular range corresponding to the emission direction. This means that the spatial detection range can envelop the spatial emission range.

[0032] According to at least one embodiment, the lasers are configured to scan pixels, for example, in a horizontal and vertical grid. The detection unit is configured to detect the emission wavelengths from a pixel currently illuminated by the lasers and preferably also from at least five or at least ten and / or from at most 50 or from at most 20 immediately preceding pixels. This means that the detection range is not limited to the current emission range.

[0033] According to at least one embodiment, the lidar system is configured for 0.1 µs ≤ T1 ≤ 2 µs or for 0.2 µs ≤ T1 ≤ 2 µs or for 0.2 µs ≤ T1 ≤ 1.1 µs. The tuning time T1 and thus also the shorter tuning time T2 can thus be relatively short. This is particularly true for automotive applications.

[0034] According to at least one embodiment, the lasers are formed by semiconductor lasers. The semiconductor lasers are based, for example, on the AlInGaAs or AlInGaP material system.

[0035] According to at least one embodiment, the lasers are configured such that the wavelength ranges lie in the near-infrared spectral range. Near-infrared refers in particular to wavelengths ≥ 800 nm or ≥ 900 nm or ≥ 1000 nm and / or to wavelengths ≤ 1.6 µm or ≤ 1.3 µm or ≤ 1.1 µm.

[0036] Furthermore, a vehicle is provided that includes at least one lidar system as described in connection with one or more of the above-mentioned embodiments. Features of the vehicle are therefore also disclosed for the lidar system, and vice versa.

[0037] In at least one embodiment, the vehicle comprises one or more lidar systems. The at least one lidar system is configured to scan the vehicle's surroundings. The vehicle is, for example, a car, a truck, a motorcycle, a ship, a train, an aircraft, or even a satellite.

[0038] Furthermore, a method for operating a lidar system, as described in connection with one or more of the above-mentioned embodiments, is provided. Features of the lidar system and the vehicle are therefore also disclosed for the method, and vice versa.

[0039] In at least one embodiment, a method for operating a lidar system is provided and comprises: - Scanning pixels of a solid angle range with the lasers, and - Detecting laser radiation coming from pixel areas, wherein when determining the distance and / or speed of an object that reflects the laser radiation back to the lidar system, laser radiation from pixel areas that were previously scanned is also taken into account.

[0040] Below, a lidar system described herein, a vehicle described herein, and a method described herein are explained in more detail with reference to the drawings using exemplary embodiments. Like reference numerals indicate like elements in the individual figures. However, they are not drawn to scale; rather, individual elements may be exaggerated for clarity.

[0041] They show: Fig. 1 and Fig. 2 schematic representations of embodiments of lidar systems described here, Fig. 3 schematic representation of an embodiment of a vehicle with lidar systems described here, Fig. 4 and Fig. 5 schematic representations of modified operating procedures, Fig. 6 to 9 are schematic representations of embodiments of methods for operating lidar systems described here, Fig. 10 is a schematic representation of a scanning characteristic of embodiments of lidar systems described here, and Fig. 11 to 17 schematic representations of method steps of an embodiment of a method for operating lidar systems described here.

[0042] In Fig. Figure 1 shows an embodiment of a lidar system 1. The lidar system 1 comprises a laser system 2 with at least two lasers 21, 22. The laser system 2 emits laser radiation S1, which is composed of radiation from the two lasers 21, 22. The lasers 21, 22 are preferably semiconductor lasers. Within separate wavelength ranges L1, L2, the emission wavelengths E1, E2 of the lasers 21, 22 are periodically tuned. The lasers 21, 22 have a large coherence length.

[0043] The laser radiation S1 reaches a beam splitter 51, which directs a beam portion S2 to a detection unit 4 having multiple detection regions 41, 42 for the wavelength ranges L1, L2 of the lasers 21, 22. The detection regions 41, 42 are preferably configured for spectrally selective detection of the wavelength ranges L1, L2.

[0044] Furthermore, a remaining, predominant portion of the laser radiation S1 passes through the beam splitter 51 via an emission optics 52 and via an optional scanner 53 to an object 8 outside the lidar system 1. For example, a distance d of the object 8 to the lidar system 1 is several tens of meters or several hundred meters. A radiation component S3 reflected by the object 8 towards a receiving optics 54 returns to the detection unit 4 and is superimposed on the radiation component S2. Since the laser radiation S1 is periodically tuned, a beat frequency in particular can be measured at the detection unit 4, from which the distance of the object 8 and / or its relative speed to the lidar system 1 is determined.

[0045] The laser system 2 and the detection unit 4 are preferably located in a common housing 55.

[0046] In Fig. 2 shows that the laser system 2 comprises four lasers 21, 22, 33, 34, which can be arranged in two pairs 21, 33 and 22, 34. Furthermore, in Fig. 2 shows that the detection unit 4 has four detection regions 41, 42, 43, 44 for pairwise different wavelength ranges L1, L2, L3, L4 of the lasers 21, 22, 33, 34. The lasers 21, 22, 33, 34 can emit monomode.

[0047] Furthermore, in deviation from the Fig. 1 and Fig. 2, it is also possible for the reflected light S3 to be collected by the same mirror with the same optics. In this case, the incident light S3 is separated from the emitted light S1, for example, via an optical circulator.

[0048] The movable scanner 53 can be designed as a mechanical mirror, for example, a rotating mirror, or as a micromechanical mirror, or MEMS mirror for short. Such a MEMS mirror can be operated in resonance or non-resonance. In resonance, larger deflections can be achieved, but the speed of movement cannot be controlled. In non-resonance, however, the mirror's movement can be controlled almost arbitrarily.

[0049] The propagation of light S1, S2, S3 within the lidar system 1 can occur using free-space optics. However, fiber optics are preferably used, i.e., the light S1, S2, S3 is guided from the laser system 2 to the emission optics 53 and from the reception optics 54 to the detection unit 4 in preferably single-mode fibers. An additional branched light beam is often guided via a longer fiber, mixed with the reference light, and detected by another detector. The resulting fixed path difference can be used to measure a modulation of the lasers 21, 22, 33, 34 and / or to adjust it in a control loop with driver electronics.

[0050] The Fig. 1 and Fig. Figure 2 shows the lidar systems 1 only schematically. Details on design variants and usable components can be found, for example, in the publications WO 2020 / 081188 A1, WO 2019 / 205163 A1, US 2018 / 0284236 A1, or US 2019 / 0257927 A1.

[0051] In Fig. Figure 3 shows a vehicle 10, for example, a car. The vehicle 10 includes several lidar systems 1 that scan the surroundings of the vehicle 10. The method described below allows high tuning rates and thus high spatial resolutions to be achieved.

[0052] The lidar systems 1 described here are thus each a frequency-modulated continuous-wave lidar, also known as an FMCW lidar, which enables accelerated scanning with multiple wavelengths. FMCW lidar stands for Frequency Modulated Continuous Wave Light Detection and Ranging. The lidar system 1 is particularly suitable for use in the automotive sector, in aviation and aerospace, in the defense sector, and in general metrology to perform distance and speed measurements. This allows the implementation of a scanning FMCW lidar system with which individual image points, also known as pixels, can be recorded in a time shorter than the light travel time to the object 8 and back.

[0053] In an FMCW lidar system described here, a laser 21, 22 emits a continuous laser beam whose light frequency is periodically modulated, with the frequency preferably increasing or decreasing linearly over a certain period of time Tc. Such an increase or decrease is also referred to as chirp; the terms tuning time and chirp duration or chirp are synonymous in this case. The emitted laser radiation S1 is thus radiated via the optics 52, reflected by an object 8 whose distance and / or speed is to be determined, and a portion S3 of the laser radiation is collected again via the receiver optics 54. This collected laser radiation S3 is mixed with the reference light S2 branched off from the emitted laser radiation S1, also referred to as a local oscillator, and the mixed light is detected on a fast photodetector, i.e., the detection unit 4.

[0054] Due to the different path of reference light S2 and light S1, S3 reflected at object 8 as well as due to the temporally preferably linearly variable frequency, corresponding to the wavelengths, the reference light S2 and the reflected light S3 have different frequencies and thus different wavelengths, see also Fig. 4. This leads to a beat, i.e., a periodic change in the intensity of the detected mixed light S2, S3. The beat frequency corresponds to the difference frequency Df, which is proportional to the path difference Dt of the laser beams S2, S3 and thus to the distance to the object 8. The beat frequency is measured by recording the signal from detection unit 4 and calculating a Fourier transform. This automatically filters out influences from background light and other sources of interference, since these are not coherent with the laser radiation S2, S3 and therefore do not contribute to the beat.

[0055] If the object 8 to be measured moves relative to the measuring device 1, the Doppler effect leads to a change in the measured difference frequency Df and thus to a falsification of the distance result in a single measurement. A correction and thus a simultaneous measurement of the relative velocity is possible by transmitting a second chirp with an opposite gradient, see also Fig. 5. Here, the relative velocity has an opposite effect, so it can be calculated from the difference Df1 - Df2 between the two differential frequencies. The distance is determined by their average value (Df1 + Df2) / 2.

[0056] This measurement principle normally requires a chirp duration, i.e., a tuning time, to be selected that is longer than the travel time Dt of the light S1, S3 to the object 8 and back. Only in this way can the interference of the returning light S3 with the reference light S2 be reliably measured within the same chirp, i.e., within the same tuning period. Since a second chirp with an opposite slope is required for the velocity measurement, this doubles the required measurement time, see Fig. 5, the frequency range with negative slope.

[0057] In the event that two objects are illuminated simultaneously and thus two difference frequencies are measured in each measurement, which can be easily evaluated using Fourier analysis, a third chirp with a different slope is required to achieve a clear result; this is not shown. The reason for this is that it is generally not possible to determine the correct assignment of the frequencies to the objects when two difference frequencies Df1, Df2 are measured. If, for example, two identical difference frequencies are measured in both chirps with opposite slopes, this could indicate two static objects. However, the same signal is obtained if both objects are at the same distance but have opposite relative velocities, i.e. one object is moving away from the measuring device and the other towards it.A third chirp with a different slope, where a constant frequency also serves this purpose, then provides data that only fits one of the two possible interpretations and thus solves the problem of the lack of clarity.

[0058] This concept therefore requires a measurement duration and tuning time greater than three times the light travel time to object 8 and back. This is especially true when the frequency change in the chirp varies in pitch, so only the time after the same chirp reappears from the object can be used to record the time dependence of the beat signal. Prior to this, the difference frequency is not constant, resulting in no usable Fourier transformation. The signal acquisition time corresponds to the integration time of the distance measurement. A sufficient integration time is required to determine the difference frequency(ies) with sufficient accuracy using Fourier analysis.

[0059] Thus, in a first time range A and in a third time range C according to Fig. 4 Measurements of the distance are possible, and in an intermediate second time range B at least when information about the chirp phase is available from a driver or a control unit, see Fig. 4. In contrast, a speed measurement according to Fig. 5 only possible in areas A and C, but not in area B.

[0060] Thus, in Fig. 4 shows the simplest operating mode for static situations: The frequency f of a laser increases linearly, is reduced to the minimum value in a very short time after reaching the maximum value at the end of the first time range A, and then increases linearly again. Likewise, the frequency f could also decrease linearly; this does not change the basic functionality. One period of this frequency response can be referred to as a chirp or tuning time. The light S3 reflected from object 8 shows the same frequency response, only shifted in time by the time the light travels to object 8 and back. This results in a constant frequency difference Df from the emitted light from the time the light returns from object 8 until it reaches its maximum frequency; see the third time range C in Fig. 4.

[0061] For this method, there must be a sufficient temporal overlap between the continued linear increase of the emitted light S1 and the returning increase in order to be able to determine the beat frequency from the measured data with the required accuracy. For this to happen, the duration of the chirp, i.e. the tuning time, as explained above, must be greater than the travel time of the light to object 8 and back. If this is not the case, in principle, and assuming a strict periodicity of the successive chirps, the difference frequency between the previous, returning chirp and the beginning rise of the next chirp in the second time range B could be determined. This frequency is the difference between the frequency deviation of the chirp and the desired difference frequency. Since the time of the chirp start, i.e. in Fig. 4, the smallest frequency for the emitted beam S1 in the system is known and can be used as a trigger for such a measurement, it is fundamentally possible to distinguish which of the two difference frequencies is being measured. This means that the entire duration of strictly periodic chirps can be used to measure the difference frequency and thus the distance, and thus also the second time range B, even if usually only the overlap within the same chirp is used.

[0062] If a relative movement of the object 8 to the measuring device cannot be excluded, this measurement is sufficient according to Fig. 4 is not sufficient, however, because the Doppler effect causes a frequency shift of the reflected light, which makes an unknown contribution to the measured difference frequency. In this case, the measurement is carried out according to Fig. 5. Here, two chirps with opposite linear gradients are used. From the two determined difference frequencies Df1 and Df2, the velocity can be calculated as Df1 - Df2 and the distance as (Df1 + Df2) / 2, as explained above.

[0063] In this mode according to Fig. 5, the measurement is only possible in time ranges A and C, where the transmitted chirp overlaps with the same returning chirp. During the temporal overlap with the previous chirp in the second time range B, the difference frequency changes over time due to the different slope, which means that after the necessary integration over a time sufficient for Fourier analysis, a clear difference frequency can no longer be measured.

[0064] Should the procedures of the Fig. 4 and Fig. 5 If the emitted laser beam is to be moved and scanned via an optical system 53 to obtain a three-dimensional image by means of several consecutive distance measurements, the scanning speed must be selected low enough to allow sufficient time for the measurement of each individual pixel. This limits the achievable spatial resolution and / or the frequency with which the image can be captured repeatedly, i.e., the frame rate.

[0065] For example, if the lidar system is to cover a range of 200 m, the time for twice the travel distance, i.e., to object 8 and back to lidar system 1, is approximately 1.33 µs. With an integration time of 0.66 µs, this results in a required chirp duration Tc of 2 µs and a resulting minimum measurement time of 6 µs per pixel. If a new image is to be acquired every 30 ms, the system can generate a maximum of 5000 pixels, not including any possible time loss due to mirror movement.

[0066] The typical requirements for 3D cameras for automotive applications are for much higher image resolutions up to the megapixel range, which cannot be achieved with such a system with only one wavelength for fundamental reasons.

[0067] Alternative ways to achieve higher resolution include combining multiple FMCW lidar systems and merging the low-resolution images they generate into a single high-resolution image. However, this is associated with high costs because multiple lidar systems are required.

[0068] In contrast, the lidar system 1 described here enables higher spatial resolution through parallel measurement with multiple wavelengths at the same pixel, preferably with a different frequency gradient for each wavelength, thus reducing the measurement time per pixel. A wavelength difference between the laser wavelength ranges is preferably greater than the wavelength variation during a chirp, i.e., within a tuning period.

[0069] The Lidar system 1 described here therefore includes in particular the following two combinable aspects: 1. A simultaneous measurement with opposite, but equal, slopes of two frequency chirps: This allows the two chirps required for distance measurement and speed measurement to be performed in parallel, directly halving the measurement time. Furthermore, this allows the slope of each chirp of the same wavelength to always remain the same. This means that each chirp can have a sawtooth pattern.

[0070] This makes it possible, in principle, to use the entire chirp period for the measurement, given a strict periodicity of the chirps generated by the laser, making the integration time virtually identical to the chirp duration. In principle, a third chirp from another laser with a different slope and a third wavelength can be performed in parallel to distinguish between multiple objects.

[0071] 2. The use of periodic, sawtooth-shaped chirps with different periodicities and a tuning time shorter than the time required for the emitted laser beam to return from the object 8 to be measured to the detection unit 4: Using only a single chirp here would result in ambiguity in the measured distance, since the same frequency difference to the local reference light S2 would result for multiple distances. Using two different chirp periods, i.e., tuning times, in parallel allows for the resolution and elimination of ambiguities.

[0072] This enables measurement using chirps that have a shorter duration than the light travel time to the object 8 and back to the detection unit 4. Another option is the resulting greater gradient of the chirps, i.e. the temporal wavelength change, which can reduce an absolute distance measurement error.

[0073] To determine the difference frequency or frequencies in each individual signal, the method described here makes it possible to use almost the entire duration of the chirp to obtain a sufficient integration time. For this purpose, all chirps preferably have a sawtooth shape with the same gradient orientation. At the beginning of each emitted chirp, where the phase position of the received chirp is distance-dependent and therefore unknown, a measured difference frequency is obtained that corresponds exactly to the difference between a modulation deviation and the desired difference frequency. After the jump in the sawtooth occurs, the desired difference frequency is measured directly. With this knowledge, suitable algorithms, such as piecewise Fourier transformation, variation of the calculation interval during the Fourier transformation, or the like, can be used to use almost the entire chirp duration as the integration time.

[0074] In combination, this results in a measurement with four wavelengths, two each with different periodicity, each of which has the opposite sign of the slope of the other, see also Fig. 9. If the magnitude of the slopes is adjusted along with the periodicity, which can be done automatically if the overall frequency sweep is kept constant, the different slopes also allow for the resolution of ambiguity between multiple objects, thus eliminating the need for an additional chirp or laser. This method allows the measurement of the distance and speed of multiple objects at the same image point in a time that can be significantly shorter than the time it takes light to travel to object 8 and back.

[0075] For this purpose, each pixel is illuminated for at least the duration of the longer chirp. After that, the illumination unit, i.e., laser system 2, switches to the next pixel, while the lasers continue to periodically emit the sawtooth chirp. The detection unit 4, comprising the two or four detection areas 41, 42, 43, 44, i.e., one detection area per wavelength, captures multiple pixels, possibly even the entire field of view to be scanned, i.e., the entire scan area of ​​the lidar system, simultaneously. If a signal consisting of the four partial signals from the four wavelengths is received at the detection unit 4, the distance and speed of the object 8 to the lidar system and relative to the lidar system 1 can be calculated using Fourier transformation, possibly even for multiple objects 8 simultaneously.

[0076] From the calculated distance, the light travel time since the chirps were emitted from laser system 2 can now be calculated, and from this, the scanning direction in which the light was imaged and / or emitted at the time of its emission can be determined. This allows the measured distance to be subsequently assigned to the correct image point, even if the scanner is already pointed at a different image point at the time of reception.

[0077] This allows for the implementation of a scanning FMCW lidar system 1, which can record individual image points in a time shorter than the light travel time to the object 8 and back. This enables faster scanning measurements than with pulsed lidar systems, which are distance measuring systems but do not provide information on relative speed.

[0078] While other methods for increasing the frame rate in scanning FMCW systems rely on traditional parallelization and thus only achieve an increase in the number of measurement points per unit of time proportional to the number of parallel systems, the approach described here, in particular by performing four parallel measurements at the same pixel, allows the effective measurement time per pixel to be reduced to significantly less than a quarter of the individual measurement time. Therefore, the method described here is particularly efficient and thus also offers considerable cost savings potential.

[0079] The simple operating mode according to Fig. 4 also provides an ambiguity in the distance measurement for static objects 8. This occurs when the chirp is possibly shorter than the travel time of the light to the object 8 and back. This problem is Fig. 6. Due to the periodicity of the chirps, measuring the difference frequency effectively measures a phase shift between the transmitted and received signal. If the measurement sensitivity allows this shift to be greater than a full period, it is no longer possible to clearly determine the distance to the object with a simple measurement.

[0080] Distances that are an integer multiple of the distance corresponding to the orbital time of a chirp period T1 are also possible. Therefore, in conventional systems, the chirp period is chosen to be so long that this ambiguity only occurs at distances where detection is no longer possible, even with highly reflective objects.

[0081] For example, the tuning time T1, within which the emission wavelength E1 is linearly tuned once through the wavelength range L1, corresponds to a distance of 50 m to the object 8. The frequency difference Df corresponds to a distance of 30 m, but also fits time differences Dt for 80 m, 130 m, 180 m, 230 m, and so on. Ambiguities arise because the tuning time T1 is smaller than R / c, where R is a specified maximum range of the lidar system 1 and c is the speed of light.

[0082] A remedy is in Fig. 7: A second chirp is emitted in parallel at a second emission wavelength E2 in a second wavelength range L2 with a different tuning time T2. This results in a second series of distance values ​​from the difference frequency of this second measurement.

[0083] For example, the tuning time T2 corresponds to a distance to object 8 of 40 m. The frequency difference Df corresponds to a distance of 20 m, but also fits time differences Dt for 60 m, 100 m, 140 m, 180 m, 220 m and so on.

[0084] A comparison with the corresponding list of possible distances from the first measurement, or an equivalent mathematical procedure, leaves only one distance that matches the measurement of both wavelengths. In this example, this is a distance of 180 m. Ambiguity only arises at the lowest common multiple of the individual ambiguity distances. The tuning times T1 and T2 must therefore be selected so that such ambiguities only occur outside the maximum specified range R.

[0085] This method allows distance measurement with chirps that last shorter than the maximum measurable round-trip time of light to object 8 and back to the lidar system 1. This allows for larger frequency gradients over time with a constant frequency sweep. A constant frequency measurement error thus results in a smaller measurement error in the time of flight and thus a higher accuracy of the distance measurement.

[0086] A measurement with multiple wavelengths can be performed using a second or third laser and a second or third detector. The local oscillator can also be designed separately for each wavelength, but can also be used jointly. In other words, the light for the reference beam can be branched off before or after the radiation from the two or more lasers is combined. Between the receiver optics and the detectors, the light must be separated depending on its wavelength, for example by a grating, a prism, or a prefabricated component such as that used for wavelength demultiplexing in telecommunications applications. The mixing with the reference light of the local oscillator can take place before or after the wavelength-selective element.

[0087] A separate emitter optic for each wavelength is possible, but is not required due to size and cost constraints. The receiver optic is preferably used jointly for all wavelengths.

[0088] Another example is in Fig. 8, where a measurement is performed with two different slopes. If a speed measurement is required, two different chirp slopes are needed to distinguish between the influence of distance and the influence of relative speed on the difference frequency. The simplest way to evaluate is to use opposing chirps. If two wavelengths are available, the measurements can be performed with both slopes in parallel. This allows a sawtooth chirp to be used for each individual measurement, so that each time interval and each phase can be used to measure the difference frequency.

[0089] In the example of Fig. 9 is a combination of different periods and gradients. Fig. 9 a combination of the application examples of the Fig. 6, Fig. 7 and Fig. 8. By simultaneously using four emission wavelengths E1, E2, E3, and E4 in four wavelength ranges L1, L2, L3, and L4, distance and speed can be measured with chirps that can be shorter than the time it takes light to travel to the object and back. The emission wavelengths E1, E3 and E2, E4 each have the same tuning times T1 and T2.

[0090] For the measurement, a chirp period with the longer period T1 is sufficient. The lower limit of the possible measurement time results solely from the achievable or required measurement accuracy, which in turn is limited by the available laser power. If very high laser power is possible because the application may exceed the eye safety requirements of laser class 1, very short measurement times are possible, for example, when using lidar from aircraft. Since such applications often involve measuring very large distances, with correspondingly long light propagation times, the advantage of this method is particularly obvious.

[0091] A realistic speed advantage is also achieved at distances of just a few hundred meters, as is typical with eye-safe lasers in the automotive sector, for example. If this method is used in a scanning system, the emitted laser beam can be directed to the next measurement point after a chirp period of the longer chirp with a duration of T1. The detector optics must, of course, still be able to receive light from the previous emission direction, see Fig. 10.

[0092] The Lidar system 1 of the Fig. 10 has a scanning range W. A detection angle range WD of the detection unit 4 is smaller than the scanning range W and is thus time-tuned. Alternatively, the detection angle range WD can also be equal to the scanning range W. A comparatively narrow emission angle range WE lies within the detection angle range WD and is smaller than the detection angle range WD. The size of the emission angle range WE determines the tangential spatial resolution of the lidar system 1. The emission angle range WE is scanned across the scanning range W.

[0093] This can be achieved using different emitter and detector optics or even using a fixed detector optic. If the method is combined with a non-resonant mirror, the measurement duration can be adapted to the specific requirements. For example, a high scanning speed and a short measurement duration per pixel can achieve a high frame rate with a reduced range. Extending the measurement duration, for example, by adjusting the chirp duration or by integrating across multiple chirps, and scanning at a correspondingly slower speed, can increase the precision and range of the measurement in selected image areas.

[0094] When evaluating such a measurement, where the signal from object 8 is only received after the emitter has been further scanned, the measured distance can be used to calculate the pixel position of object 8. The corresponding method is described in the Fig. 11 to 17, which illustrate a scanning process in time steps. The pixel in whose direction the laser system 2 emits the laser radiation S1 in the respective time step is marked by a circle.

[0095] Fig. Figure 11 shows the first time period. The laser system 2 emits in the direction of pixel 1 for the duration of, for example, a chirp of the longer period T1. During this time period, the light just reaches a first object 8 located in this direction. No light from an object 8 has yet reached the detection unit 4 with the wavelength-selective detection areas, so nothing is detected yet.

[0096] Then the laser system 2 swings towards pixel 2 in the second time period, see Fig. 12. During this time period, the light from pixel 1 is currently on its way back to detection unit 4, but hasn't yet arrived there. The light from pixel 2 hasn't yet reached the other object 8 in pixel 2's field of view. Again, nothing is detected.

[0097] In period 3, see Fig. 13, is emitted toward pixel 3. In the field of view of pixel 2, the laser radiation S1 is being reflected by the object 8 located there. The light emitted in the area of ​​pixel 1 arrives at the detection unit 4 and is measured. This detection event I is symbolized by a star.

[0098] Since the detector optics receive the light from a larger angular range, detection unit 4 does not provide information about the direction from which this signal was received. However, detection unit 4 does provide information about the distance. From this, the speed of light can be used to calculate how long the just-detected laser radiation S3 was traveling. From this time and the current scan position, it is possible to calculate the direction in which the just-detected laser radiation was emitted. From this, the angular position of the detected object 8, in this case, pixel 1, is determined.

[0099] The Fig. Figures 14 to 17 show examples of the further course of the measurement. Please note, see Fig. 15 and Fig. 17, that the detection unit 4 is capable of detecting multiple signals from different distances simultaneously. Two reflected signals with different differential frequencies then arrive, resulting in a superposition of the beats. The two frequencies can be determined using Fourier analysis. By using two different chirp slopes, the speed and distance of both objects 8 can be unambiguously determined.

[0100] Thus, in the described measuring method, both detected objects 8, which are located according to Fig. 15 in the area of ​​pixels 2 and 4 according to Fig. 17 located in the area of ​​pixels 3 and 6 are considered separately. This means that for each object 8, the light travel time and, from this, the emission direction, and thus the corresponding pixel and the associated object position, are determined separately from the distance.

[0101] Thus, the detection events I in the time step to pixel 5 of the Fig. 15, for example, distances of 25 m and 75 m, which means that the respective objects 8 are one pixel or three pixels behind in the time domain and are therefore to be assigned to pixels 2 and 4. Similarly, the detection events I in the time step to pixel 7 of the Fig. 17 for example, distances of 25 m and 100 m, which means that the objects 8 in question are one pixel or four pixels behind in the time domain and can therefore be assigned to pixels 3 and 6 List of reference symbols 1 Lidar system 2 laser system 21 first laser 22 second laser 33 third laser 34 fourth laser 4 Detection unit 41..44 detection range 51 beam splitters 52 Emission optics 53 scanners 54 Receiving optics 55 housings 8 Object 9 modified lidar system 10 vehicles A first time range B second time range C third time range c speed of light d distance Dd Frequency difference or difference frequency Dt runtime difference f Frequency as a measure of the wavelength E emission wavelength I Detection event In laser radiation reflected from the object, received Laser radiation emitted to the object L Wavelength range P pixels S laser radiation R intended maximum range t time T Tuning time W Scan range WD detection angle range WE emission angle range

Claims

[1] Lidar system (1) with - a first laser (21) and a second laser (22), and - a detection unit (4) for detecting laser radiation from the first and second lasers (21, 22), wherein - the first laser (21) in a first wavelength range (L1) and the second laser (22) in a second wavelength range (L2) are arranged for periodic tuning of a respective emission wavelength (E1, E2), - a first tuning time T1 of the first laser (21) differs from a second tuning time T2 of the second laser (22), - holds: m T1 = n T2 with m, n ∈ ℚ \ ℕ. N and m < n as well as n / m ∈ ℚ \ ℕ. where for all i = 1, ..., n and for all j = 1, ..., m holds: imjn≤0.97∧1.03≤imjn, as well as m T1 > FR / c and (m - 1) T1 ≤ FR / c with 0.2 ≤ F ≤ 1.5, and R is a specified maximum range of the lidar system (1) and c is the vacuum speed of light. [2] Lidar system (1) according to the preceding claim, wherein at least one of the tuning times T1 and T2 is at least a factor of 2 smaller than the intended maximum range R of the lidar system (1) divided by the vacuum speed of light c. [3] Lidar system (1) according to one of the preceding claims, wherein m = 7 and n = 9 or m = 7 and n = 11 or m = 5 and n = 8. [4] Lidar system (1) according to one of the preceding claims, wherein the intended maximum range R is at least 0.1 km and at most 0.5 km or at least 2 km and at most 10 km. [5] Lidar system (1) according to one of the preceding claims, which is arranged to tune the emission wavelengths (E1, E2) of the first laser (21) and the second laser (22) in the form of a triangular variation or in the form of a sawtooth variation. [6] Lidar system (1) according to claim 5, in which the first laser (21) and the second laser (22) have different tuning slopes, where the tuning slopes are defined as wavelength difference per unit time, within respective tuning periods. [7] Lidar system (1) according to one of the preceding claims, comprising a third laser (33) and a fourth laser (34), wherein the third laser (33) in a third wavelength range (L3) and the fourth laser (34) in a fourth wavelength range (L4) are arranged for periodically tuning a respective emission wavelength (E3, E4), wherein the first, second, third and fourth wavelength ranges (L1, L2, L3, L4) differ from each other in pairs and do not overlap each other. [8] Lidar system (1) according to claim 7, in which the third laser (33) and the first laser (21) form a first laser pair and the fourth laser (34) and the second laser (22) form a second laser pair, where within the laser pairs the corresponding tuning times are the same, but the tuning slopes are different, so that the lasers within each laser pair are set up to be tuned synchronously and wherein the detection unit (4) is also configured to detect laser radiation from the third and fourth lasers (21, 22). [9] Lidar system (1) according to one of the preceding claims, in which the detection unit (4) is designed to detect the wavelength ranges (L1, L2, L3, L4) individually and independently of one another, wherein the lasers (21, 22, 33, 34) are arranged to all emit in a specific, common emission direction at a specific time, and wherein the detection unit (4) is configured to detect the emission wavelengths (E1, E2, E3, E4) from a spatial area that includes and envelops the emission direction. [10] Lidar system (1) according to claim 9, in which the lasers (21, 22, 33, 34) are arranged to scan pixels, wherein the detection unit (4) is configured to detect the emission wavelengths (E1, E2, E3, E4) from a pixel currently illuminated by the lasers (21, 22, 33, 34) and from at least five immediately preceding pixels. [11] Lidar system (1) according to one of the preceding claims, which is set for 0.1 µs ≤ T1 ≤ 2 µs. [12] Lidar system (1) according to one of the preceding claims, in which the lasers (21, 22, 33, 34) are formed by semiconductor lasers and which is designed so that the wavelength ranges lie in the near-infrared spectral range. [13] Vehicle (10) comprising at least one lidar system (1) according to one of the preceding claims, wherein the at least one lidar system (1) is configured to scan an environment of the vehicle (10). [14] Method for operating a lidar system (1) according to one of claims 1 to 12, comprising: - scanning pixels of a solid angle range with the lasers (21, 22, 33, 34), and - Detecting laser radiation from the lasers (21, 22, 33, 34) coming from pixel areas, wherein, when determining the distance and / or the speed of an object (8) which reflects the laser radiation back to the lidar system (1), laser radiation from pixel areas which were previously scanned is also taken into account.

Images