Creating a model of the environment of a motor vehicle and determining the relative speed between the motor vehicle and objects in the environment.

Abstract

Method for mapping the surroundings of a motor vehicle using an angle-measuring FMCW radar (R, 1) with range Doppler evaluation, wherein the FMCW radar (R, 1) has at least one transmitting antenna (Tx, 2) and at least two receiving antennas (Rx1, Rx2, 3, 4), the FMCW radar (R, 1) has a predetermined angle (θ) to the direction of movement and the far-field approximation is applicable, the radar (R, 1) moving in a straight line at constant speed emits signals for a limited time period in the direction of objects (Oi, O1, O2, O3) in the surroundings of the motor vehicle, the signals reflected by the objects (Oi, O1, O2, O3) are received separately by the at least two receiving antennas (Rx1, Rx2, 3, 4), the signals received in the limited time period constitute a group of M different measurement signals for each receiving antenna (Rx1, Rx2, 3, 4) forms,Each group of M measurement signals is transformed into the frequency domain by means of a two-dimensional Fourier transform, and a range-Doppler image of each group is generated. Each pixel of the range-Doppler image is assigned a distance relative to the radar (R, 1). The system is characterized in that, to estimate the angles between the radar (R, 1) and objects (Oi, O1, O2, O3), a conjugate multiplication of the at least two range-Doppler images is performed to obtain an RDA image. From the distance information of the pixels of two range-Doppler images and the angle information of the corresponding RDA image, an image of the environment of the radar (R) is generated. The Doppler shift of each pixel of the range-Doppler images is corrected, and the pixels of the environment image are interpolated to generate an equidistant grid.

Description

[0001] The invention relates to a method for mapping the environment of a motor vehicle according to the preamble of claim 1, a method for determining the speed between a motor vehicle and objects in the environment of the motor vehicle according to the preamble of claim 5, a device for mapping the environment of a motor vehicle according to the preamble of claim 8, and a device for determining the speed between a motor vehicle and objects in the environment of the motor vehicle according to the preamble of claim 9.

[0002] Synthetic aperture imaging radar systems, also known as SAR radar, are well-known in the fields of remote sensing and military technology. Synthetic apertures can, in principle, be generated by any moving object equipped with radar, either by moving a single antenna along a trajectory or by using multiple spatially distributed antennas. It is important to note that the radar signals must be in coherent phase and that the antennas or aperture points should be positioned close enough together to satisfy the spatial sampling theorem.

[0003] The transfer of SAR radar systems and methods known from remote sensing to the commercial sector has so far failed because, during antenna movement, the aperture positions must be known over the entire aperture length with an uncertainty significantly smaller than one wavelength of the radar signals used. This uncertainty is approximately 1–2 mm at a frequency of 24 GHz and approximately 0.5 mm at 79 GHz. In remote sensing, this requires highly sophisticated inertial platforms with correspondingly high costs and very complex or non-real-time post-processing algorithms. However, such solutions are completely unsuitable for the automotive, automation, and / or robotics sectors with their real-time and cost-related requirements.

[0004] A suitable aperture synthesis would achieve 360-degree imaging with the highest resolution. This high resolution is desirable for better / more accurate recognition of environmental details, which will be particularly important for future autonomous vehicles, or for increasing safety by enabling the early detection of small objects or people obscured by other objects.

[0005] Furthermore, in the microwave range, Doppler radar is a well-known method for estimating the velocity of an object. However, the Doppler effect only allows the determination of the relative velocity between a radar and an object in the radial direction. To determine the relative vectorial velocity between a radar and an object, a monopulse radar can be used, which, compared to a simple Doppler radar, has two or more receiving antennas. The distance between two adjacent receiving antennas is generally equal to or less than half the wavelength λ / 2 to ensure that the phase difference of the echo signals can be uniquely assigned to an angle relative to the object. This allows for an unambiguous estimation of direction and / or angle, or the relative lateral velocity, between a radar and an object.Angle estimation, however, assumes that no two objects are located at the same distance within the scene. Due to the limited signal bandwidth, the distance between the two objects must be greater than half the echo signal width. With a bandwidth of 250 MHz, the echo width is approximately 0.6 m; otherwise, the echo signal will be masked by the echo signals of other objects. Angle estimation, and therefore a relative vector velocity estimation, is not possible under such circumstances.

[0006] German patent application DE 10 2010 015 723 A1 describes a method for detecting the movement of a road vehicle, in which measured values ​​of the relative movement between a receiver located in the vehicle and objects in the vehicle's vicinity are acquired as a function of angle. A velocity vector of the vehicle relative to the objects in the vehicle's environment is determined by calculating an angular function, which is then derived by means of a least-squares adjustment of the angle-dependent measured values ​​of the relative movement. By using several successively determined positions of the vehicle as measuring points, a synthetic aperture positioning technique is achieved.

[0007] German patent application DE 10 2010 051 207 A1 relates to a method and a device for the three-dimensional imaging of an object moving relative to a sensor arrangement using a combination of radar technology and optical imaging. A three-dimensional radar image of the object is acquired and fused with a surface model of the object generated by optical sensors in such a way that the radar image is focused onto the surface. In this way, a realistic representation of the imaged object is possible.

[0008] From the publication DE 199 12 370 A1, a method for radar signal processing in a motor vehicle is known in which the signals of each target lane are coherently integrated for so long that the targets, i.e. the other vehicles, are split into a plurality of Doppler cells, thereby making imaging monopulse bearing possible.

[0009] The publication DE 10 2009 030 075 A1 describes an imaging method with synthetic aperture for determining an angle of incidence and / or a distance of a sensor to at least one object in space, in which an echo profile is recorded at a number of aperture points.

[0010] German patent application DE 10 2010 048 896 A1 relates to a device for detecting the environment of a vehicle with a radar system, comprising a corresponding transmitter and receiver unit as well as an evaluation unit for evaluating echo profiles of the received signals and for deriving information about the occupancy of spatial areas in the vicinity of the vehicle. The transmitter unit includes a modulation unit for generating a frequency-modulated continuous wave signal, which is iteratively formed from a plurality of nested frequency steps. In the evaluation unit, for each iteration step, measured values ​​of the received signal are assigned to frequency levels of the individual frequency steps, and for each of the frequency steps, an echo profile is generated from the respective assigned measured values, with each echo profile containing the environment information for an aperture point of a synthetic aperture.

[0011] The invention is based on the objective of creating an improved method for imaging the environment around a moving radar of a motor vehicle and for determining the relative vectorial velocity between the radar and objects in the vehicle environment, as well as specifying corresponding devices.

[0012] This problem is solved by a method with the features of claim 1, by a method with the features of claim 5, by a device with the features of claim 8, and by a device with the features of claim 9. Preferred embodiments of the method are the subject of the dependent claims.

[0013] In the inventive method for imaging the surroundings of a motor vehicle using an angle-measuring FMCW radar with range-Doppler evaluation (FMCW: Frequency-Modulated Continuous-Wave), wherein the FMCW radar has at least one transmitting antenna and at least two receiving antennas, the FMCW radar has a predetermined angle to the direction of movement and the far-field approximation is applicable, – the radar, moving in a straight line at a constant speed, emits signals towards objects in the vicinity of the vehicle for a limited period of time, – the signals reflected by the objects are received separately by the at least two receiving antennas, – the signals received within the limited time period form a group of M different measurement signals for each receiving antenna, – each group of M measurement signals is transformed into the frequency domain using a two-dimensional Fourier transform and a range Doppler image of each group is generated, – a distance relative to the radar is assigned to each pixel of the range Doppler image, – To estimate the angles between radar and objects, a conjugate multiplication of at least two range Doppler images is performed to create an RDA image, and – an image of the radar's surroundings is generated from the distance information of the pixels of two range Doppler images and the angular information of the corresponding RDA image.

[0014] This allows for simple and cost-effective SAR imaging of the vehicle's surroundings, even without knowing the relative speed between the radar and objects in the vehicle's vicinity. In other words, SAR imaging is possible even when the exact relative speed between the radar and the object is unknown. This was previously impossible in the field of commercial SAR imaging.

[0015] Furthermore, the Doppler shift of each pixel in the range-Doppler images is preferably corrected. Since each pixel of a range-Doppler image is assigned a unique Doppler frequency, the distance of each pixel to the radar, which is distorted by a Doppler component, can be corrected in this way.

[0016] Preferably, a three-dimensional image of the vehicle's surroundings can be created using a two-dimensional antenna. In other words, extending an antenna arrangement from a line to a surface, thereby allowing angles to be determined in both spatial directions, results in three-dimensional imaging.

[0017] Preferably, the pixels of the surrounding image are interpolated to generate an equidistant grid. Since the transformations, i.e., the determination of the range Doppler images and the RDA image, generally do not result in positions lying on an equidistant grid, it is advantageous to perform interpolation to improve the surrounding image.

[0018] The inventive method for determining the relative vectorial velocity between a radar and objects in the vicinity of a motor vehicle using an angle-measuring FMCW radar with range Doppler evaluation, which may in particular include the method described above, wherein the FMCW radar has at least one transmitting antenna and at least two receiving antennas, the FMCW radar has a predetermined angle to the direction of movement and the far-field approximation is applicable, performs the following steps: – the radar, moving in a straight line at a constant speed, emits signals for a limited period of time in the direction of objects in the vicinity of the vehicle, – the signals reflected by the objects are received separately by at least two receiving antennas, – the signals received within the limited time period form a group of M different measurement signals for each receiving antenna, – Each group of M measurement signals is transformed into the frequency domain using a two-dimensional Fourier transform, and a range Doppler image of each group is generated. – To estimate the angles between radar and objects, a conjugate multiplication of at least two range Doppler images is performed to create an RDA image, and – For each pixel of the range Doppler images, a radar velocity is determined from the corresponding RDA image.

[0019] This method makes it easy to determine the radar speed.

[0020] Preferably, the radar velocities determined for all pixels are converted into a one-dimensional velocity vector, and local maxima of the velocity vector are determined to ascertain the radar vehicle speed and the velocities of the moving objects. If there are no moving objects in the vicinity of the vehicle, the local maximum corresponds to the radar speed.

[0021] Preferably, to separate the radar speed from the speeds of moving objects, the speed of the vehicle can be used as the radar speed. In this way, the local maximum corresponding to the radar speed can be identified. Other local maxima then correspond to the speeds of moving objects.

[0022] This also makes it possible to determine the speeds of moving objects in the vicinity of the radar, i.e., the motor vehicle, since the speed of the motor vehicle is known and therefore the radar speed is identical.

[0023] The device according to the invention for imaging the surroundings of a motor vehicle comprises: – an angle-measuring FMCW radar with range Doppler evaluation, wherein the FMCW radar has at least one transmitting antenna and at least two receiving antennas, and – a control unit with – a device for controlling the radar and for evaluating the received measurement signals, – a device for determining range Doppler images – a device for determining an RDA image from two range Doppler images by means of conjugate multiplication of the two range Doppler images, and – a device for generating an image of the radar's surroundings from the distance information of the pixels of two range Doppler images and the angular information of the corresponding RDA image.

[0024] The device according to the invention for determining the relative vectorial velocity between a radar and objects in the vicinity of a motor vehicle by means of an angle-measuring FMCW radar with range Doppler evaluation, which may in particular include the device described above and is set up and designed for carrying out the method described above, comprises: an angle-measuring FMCW radar with range Doppler evaluation, wherein the FMCW radar has at least one transmitting antenna and at least two receiving antennas, and a control unit with – a device for controlling the radar and for evaluating the received measurement signals, – a device for determining range Doppler images, a device for determining an RDA image from two range Doppler images by means of conjugate multiplication of the two range Doppler images, and a device for determining the radar velocity for each pixel of the range Doppler images from the corresponding RDA image.

[0025] Preferably, the control device further comprises a device for transferring the radar velocities determined for all pixels into a one-dimensional velocity vector and for determining the radar speed and the velocities of the moving objects by forming local maxima.

[0026] The methods and devices described above are applied to radar technology. However, they can also be applied to other coherent wave-based measurement methods, such as ultrasound.

[0027] Preferred embodiments of the invention are explained with reference to the following drawings.

[0028] Fig. 1. A schematic representation of a first object scene to illustrate the mapping of the environment around a moving object.

[0029] Fig. 2. A schematic representation of a second object scene to illustrate the determination of a relative velocity between a radar and a moving object.

[0030] Fig. 3 the geometric relationship between moving radar and stationary objects,

[0031] Fig. 4. A first measurement situation in schematic representation,

[0032] Fig. 5 the range Doppler images (RD images) of the two receiving antennas,

[0033] Fig. 6 the RDA image resulting from the two RD images,

[0034] Fig. 7 the resulting environmental representation of the first measurement situation,

[0035] Fig. 8 a second measurement situation in a parking lot,

[0036] Fig. 9 the environmental representation of the second measurement situation,

[0037] Fig. 10 a third measurement situation in a parking lot,

[0038] Fig. 11 the environmental representation of the third measurement situation,

[0039] Fig. 12 the measured speed of the radar in the first measurement situation,

[0040] Fig. 13 a fourth measurement situation with a moving object,

[0041] Fig. 14 the measured velocities of the radar and the moving object, and

[0042] Fig. 15 a schematic representation of the device.

[0043] Fig. Figure 1 serves to explain the basic radar concept and the recording situation in a first object scene, whereby the method is based on a range-Doppler angle method, also known as RDA (Range-Doppler-Angle). An FMCW radar (FMCW: Frequency Modulated Continuous Wave) is used for angle estimation, which in this example comprises a transmitting antenna Tx and two receiving antennas Rx1 and Rx2. FMCW radars with angle measurement capability are known as amplitude or phase monopulse radars, MIMO radars, or in many other configurations. Furthermore, FMCW radars with range-Doppler evaluation are also known in the prior art.

[0044] In an FMCW range-Doppler radar, a number N of linearly frequency-modulated signals (FMCW signals) are transmitted sequentially, and the entire set of resulting N received signals is then evaluated together, typically using a two-dimensional Fourier transform. This process allows the signals from an angle-measuring FMCW radar with range-Doppler evaluation to be processed in such a way that SAR imaging is possible even when the exact relative velocity between the radar and the object is unknown.

[0045] To further describe the radar concept, a simple phase monopulse radar with one transmitting antenna Tx and two separate receiving antennas Rx1 and Rx2 is assumed. Expansion to include additional transmitting and / or receiving antennas, e.g., to improve angle estimation, would be possible at any time. Furthermore, the following are located in the vicinity of the radar: Fig. 1 Three objects Oi, i = 1, 2, 3, acting as point spreaders, are arranged.

[0046] For the sake of a compact presentation, the following further simplifications are assumed: – The distance between the transmitting antenna Tx and the receiving antennas Rx1, Rx2 is much smaller than the distance between the radar and an object Oi (far field approximation). – The detection ranges of all antennas Tx, Rx1, Rx2 are assumed to be identical. – The two receiving antennas Rx1, Rx2 are assumed to be so close together that the phase difference of the two received signals from the two antennas Rx1, Rx2 can be uniquely assigned to an object angle. For this purpose, it is assumed that the distances or angles from an object Oi to each receiving antenna Rx1, Rx2 are nearly equal (far-field approximation). – The position of the radar receiving antenna Rx1 is defined as the origin of the world coordinate system (x, y). This leads to a simplified transformation between the angle θ oi in the antenna coordinates of an object Oi and the angle θ w,oi in the world coordinates θ w,Oi = θ Oi – φ. – It is assumed that the relative velocity between radar and object Oi remains almost constant during a time period Ts.

[0047] Starting from a starting position p Tx = (x Tx , y Tx ) T = (a Tx ·cosφ, a Tx ·sinφ) T sends the in Fig. 1. Radar transmitting antenna Tx shown during a limited time period T s constantly signals in the direction of the objects Oi, i = 1, 2, 3. The size a Txis the distance between the origin of the coordinate system and the transmitting antenna Tx. Furthermore, the quantity φ defines the angle of rotation of the radar antennas around the origin. Finally, let the vehicle speed be ν = (ν x , ν y = 0) T the radar system remains constant during the measurement process.

[0048] The radar beams emitted by the transmitting antenna Tx are detected by an object Oi with position p Oi = (x Oi , y Oi ) T scattered or reflected. In Fig. Figure 1 shows three objects O1, O2, and O3 as examples. The signals scattered by all objects Oi are then received separately by the two receiving antennas Rx1 and Rx2. The starting positions of antennas Rx1 and Rx2 are each defined by: p Rx1 = (x Rx1 = 0, y Rx1 = 0) T and p Rx2 = (x rx2 , y Rx2 ) T = (l cos(π + φ), l sin(π + φ))

[0049] Here, l corresponds to the distance between the receiving antennas Rx1, Rx2, which is generally less than or equal to half the wavelength of the radar signal. λ / 2 is.

[0050] The measurement with a time duration Ts leads to a group of M different measurement signals, which are subsequently denoted by s(t s , t) where t s The slow-time and the fast-time are both meant.

[0051] A transformation of the M measurement signals s(t s , t) with a two-dimensional Fourier transform into the frequency and Doppler frequency domain – as is known from FMCW range-Doppler methods according to the state of the art – the following range-Doppler images, abbreviated RD images, are obtained for each of the two receiving antennas: S Rx1 (ω d , ω) = FFT2{s Rx1 (t s , t)·W(t s , t)} (1) S Rx2 (ω d , ω) = FFT2{s Rx2 (ts , t)·W(t s , t)} (2)

[0052] Here, W denotes a two-dimensional window function, for example, a Hamming window. The range-Doppler image can be represented as a probability density distribution of distance d. Oi and Doppler velocity ν r,Oi a point spreader P Oi to display on radar.

[0053] Additional transmitting and / or receiving antennas will produce correspondingly more RD images.

[0054] A direction estimation between all objects Oi and the radar is obtained by a conjugate multiplication of two RD images:

[0055] From two RD images acquired with spatially separated antennas Rx1 and Rx2, a so-called RDA image is generated by conjugate multiplication. This RDA image assigns an angle relative to the radar to each pixel from the RD images. By using additional antennas or RD images, the angle measurement for each pixel can be improved.

[0056] If the distance l between the receiving antennas Rx1, Rx2 is equal to half the wavelength of the radar radiation, i.e., l = λ / 2, then equation (3) simplifies to:

[0057] A geometrically correct image of the environment around the radar can now be obtained as follows: In a first step, each pixel in the RD image can be assigned a unique distance relative to the radar. Since, in an FMCW range-Doppler analysis, the distance axis initially only describes a so-called pseudo-range (i.e., the range axis is still distorted by a Doppler component), this Doppler shift must first be corrected for each pixel. This is easily accomplished, however, because each pixel in the RD image can be assigned a unique Doppler frequency.

[0058] In a second step, the RDA image also provides a relative angular position to the radar for each pixel. Using the distance and angle information obtained for each pixel in steps 1 and 2, it is now possible in a third step to easily transform each pixel from an RD image into a spatial xy coordinate system. This coordinate transformation must then be performed for each pixel. This results in a geometrically correct representation of the environment, regardless of the relative velocity between the radar and the object during the measurement.

[0059] Signal components or pixels without radar echo signals – i.e., those containing only interference or noise – are statistically distributed across the xy-image area and hardly affect the resulting image. Signal components or pixels originating from radar echo signals or objects are systematically and correctly positioned within the image by this method.

[0060] A particular advantage of this method is that it also produces an accurate image for situations where two objects are moving at different speeds. This would not be possible with current SAR technology.

[0061] If the antenna arrangement is extended from a line to a surface, so that the angles in both spatial directions can be determined, three-dimensional imaging is also possible with the method according to the invention.

[0062] Since the transformation mentioned above generally does not result in positions that lie on an equidistant grid, interpolation is necessary, which can be performed as follows: The location information for all objects Oi can be obtained from the two RD images and the RDA image. An image of the area around a moving radar is created using the following steps: – Definition of a two-dimensional image B with a limited resolution to describe a two-dimensional space. The weight of each pixel is set to zero. – For each pixel (ω d , ω) a distance d can be determined Oi determine as follows: where c is the speed of light and μ is the sweep rate. – The combination of equations (4) and (5) yields the position P for each image point (ω). d , ω) as follows: – As a function of the determined position P, the four pixels of the interpolation image B that have the smallest distance to position P are determined and these pixels are weighted as follows:

[0063] Fig. Figure 2 shows a schematic representation of a second object scene to illustrate the determination of the relative velocity between a radar and a moving object. The difference to the situation of Fig. 1 in that the object O1 moves with the velocity ν O1 = (ν O1,x , ν O1,y ) T The object O2 and O3 are moving, while the other objects O2 and O3 remain stationary in space. This relates to the situation of... Fig. The statements made in section 1, in particular equations (1) to (7), are therefore also applicable to this situation.

[0064] As from the Fig. As can be seen in 2, the general relative Doppler velocity between a radar R and a point scatterer Oi with position p is given by Oi to: ν R,Oi = ν x ·sinθ w,Oi – |v Oi |·cosα Oi , (8) where v Oi and α OiThe speed and direction of travel of the object Oi are, and in the example of the Fig. 3 the position of the radar receiving antenna Rx1 is equated with the radar R.

[0065] For all stationary objects with |v Oi | = 0, in the Fig. Since these are objects O2 and O3, equation (8) simplifies to:

[0066] Fig. Figure 3 shows the geometric relationship between the radar speed and the stationary objects that are in Fig. 2 are exemplified by objects O2 and O3. The radar moves at the speed ν. x to the left, where the y-component of the radar velocity is zero. The radar velocity vector can therefore be vectorially decomposed into a velocity component ν. R,O2 with respect to the object O2 and a component ν R,O3 regarding object O3, as described in Fig. 3 is shown.

[0067] The relative vector velocity is now determined using the following steps: – Definition of a one-dimensional vector V with a given minimum and maximum speed and a limited resolution. For example, a minimum / maximum speed of ± 2 m / s and a given resolution of 5 mm / s can be specified to determine a radar speed of 0.5 m / s. The weights of all points are set to zero. – For each pixel (ω d , ω) is used to determine a radar velocity ν using equation (9). x (ω d , ω). The Doppler velocity in equation (9) corresponds to the Doppler velocity in equations (1) and (2), resulting in: – The calculated radar speed ν x is assigned to the one-dimensional vector V, with the weighting of the elements of the vector being carried out as follows:

[0068] This diagram enables the determination of the vectorial speed of a moving radar. Furthermore, it allows the separation of moving objects from stationary objects in a complex environment by determining the relative vectorial speed. If the range of validity of the radar speed is known, the radar speed can be obtained by determining the local maximum. Subsequently, the term |ν Oi |·cosα Oi can be determined in equation (8).

[0069] Fig. Figure 4 shows a schematic representation of a measurement setup used both for creating the environment image of a radar and for determining the relative vectorial velocity between the radar and objects in the vicinity of the radar.

[0070] Schematically represented in Fig. 4 is a radar R, which consists, for example, of the components from the Fig. The setup consists of one known component, namely the transmitting antenna Tx and the two receiving antennas Rx1 and Rx2. In this measurement setup, the radar coordinate system coincides with the world coordinate system xy, and the radar R moves at a uniform velocity v in the x-direction, i.e., in Fig. 1 to the left. Consequently, the following applies to the measurement setup of the Fig. 4: φ = 0.

[0071] The radar R used for the measurements shown in the following figures travels at a speed of 0.05 m / s in the x-direction, with a 16 cm aperture being recorded for evaluation. The radar R, with its two receiving antennas Rx1 and Rx2, is based on the well-known FMCW concept with a center frequency of 24 GHz and a bandwidth of 250 MHz. Two angled reflectors WR1 and WR2 are positioned at a predetermined distance in the y-direction in front of a wall W, which is parallel to the x-axis. During a measurement, the radar R passes the two angled reflectors WR1 and WR2 at a constant lateral distance.

[0072] In Fig. Figure 5 shows the RD images determined according to equations 1 and 2, with the upper part of the Fig. 5 the RD image of the receiving antenna Rx1 and the lower part of the Fig. Figure 5 shows the RD image of the second receiving antenna Rx2. In both RD images, the signal S1 of the first angled reflector WR1, the signal S2 of the second angled reflector WR2, and the signal S3 of the wall W are clearly visible.

[0073] Fig. Figure 6 shows the representation of the angles of all objects in space, which is determined according to equation (4), i.e. the RDA image, which results from a conjugate multiplication of the two RD images of the Fig. 5 results.

[0074] The resulting mapping of the in Fig. The measurement environment of radar R shown in section 4 is Fig. 7, which shows the distance of wall W and the position of the two angle reflectors WR1 and WR2.

[0075] Fig. Figure 8 schematically illustrates the testing of the method for creating an environmental image in a parking lot. The measurement situation in a parking lot is as follows: the robot-driven radar R traverses the parking lot in the x-direction at a speed v and is intended to detect the parked vehicles F1, F2, and F3. The parking spaces are defined by markings PM, and three of the four parking spaces defined by the markings are occupied by vehicles F1, F2, and F3, which are positioned at different depths within the perpendicular parking spaces. The coordinate system of the radar R is rotated by 45° relative to the world system; in other words, φ = 45°.

[0076] In Fig. 9 is the determined environment of radar R of the Fig. 8 is shown, i.e. the parking lot area, where in Fig. Figure 9 schematically shows the antenna orientation of radar R and the direction of movement. The positions of the rear sections of vehicles F1, F2, and F3 are visible, and the different depths of penetration of the vehicles into the perpendicular parking spaces can also be seen in the surrounding area.

[0077] Fig. Figure 10 shows another situation in a parking lot, with the left part of the Fig. 10 a viewing direction perpendicular to the direction of movement of the radar R and the right part of the Fig. 10. The viewing direction is shown in the direction of movement of the radar R. A first vehicle F1 is positioned parallel to the direction of movement of the radar R, so that the radar's trajectory runs parallel to the driver's side of the first vehicle F1, and therefore the two driver-side wheels Rd1 and Rd2 are exposed. The distance between the two wheels is approximately 2.77 m. Furthermore, a second vehicle F2 is positioned behind the first vehicle F1 in a parked position in a perpendicular parking space, in the direction of travel of the radar R. The coordinate system of the radar R is as shown in Fig. 8 rotated by 45° relative to the world system, so φ = 45°.

[0078] Fig. Figure 11 shows the result of the radar measurement evaluation in the form of a map of the area surrounding radar R, with the x-axis running in the direction of radar R's movement and the y-direction perpendicular to it. It can be seen that vehicles F1 and F2 are located approximately 5 m away from the direction of radar R's movement represented by the x-axis, with the radar signals from wheels Rd1 and Rd2 clearly standing out compared to the radar signals on the driver's side of vehicle F1. Furthermore, the wheel spacing of approximately 2.80 m can be determined from the map.

[0079] Fig. Figure 12 shows the result of a determination of the radar R's driving speed in the Fig. 4. Experimental setup shown, where the angle φ between the radar coordinate system and the world coordinate system is φ = 0°. Starting from the experimental setup of the Fig. 4 as well as the one in the Fig. 5 and Fig. The 6 RD images and the RDA image shown can be determined from the angles of the RDA image. Fig. 6. Determine the robot's travel speed using equations (10) and (11), as shown in Fig. Figure 12 shows the driving speed determined by radar to be 0.05 m / s, which corresponds to the speed used in the experimental setup.

[0080] Fig. Figure 13 shows an experimental setup similar to that of the Fig. 4, where the radar's coordinate system is rotated by 45° relative to the xy-world coordinate system. In other words, φ = 45°. In the experiment, the radar R moves at a speed of 0.075 m in the x-direction, for example, traveling on a track. Data acquisition is performed analogously to the setup of the Fig. 4, i.e., the radar R comprises one transmitting antenna and two receiving antennas (not shown). In contrast to the experiment according to Fig. 4. During the measurement, a person P walks past the radar R.

[0081] Fig. Figure 14 shows the result of the speed determination based on the given conditions. Fig. 13. The radar's peak speed reading is visible at approximately 0.08 m / s, corresponding to the set speed. Furthermore, three peaks are visible at speeds between 0.9 and 1.3 m / s, which can be attributed to the person walking. The peaks at 1 m / s and 1.2 m / s correspond to the hand movements of the person walking, and the middle peak at 1.15 m / s corresponds to the walking speed v. P person P.

[0082] Fig. Figure 15 shows a schematic representation of a preferred embodiment of the device in which both the environment of the moving radar and the relative velocities between radar and objects, in particular moving objects, are determined.

[0083] The device shown includes a radar antenna. 1 with a transmitting antenna 2 and two receiving antennas 3 , 4 , where the radar antenna 1 It features a linear antenna arrangement. The radar antenna 1 is controlled by a control device 5 controlled by a radar control 6 for controlling the radar antenna 1 as well as for recording and grouping the measured values ​​according to the receiving antenna and the measurement duration. The recorded groups of measured values ​​are sent to a first transformation unit. 7 The signal is fed in, and the measurement groups are transformed into rank-Doppler images, so-called RD images, using a Fourier transform. The two RD images from the two receiving antennas... 3 , 4 are then transferred to a second transformation facility 8The images are transformed into RDA images containing angular information using conjugate multiplication. From these RD images, together with the angular information of the corresponding RDA image, a device is used to create a new image. 9 an image of the radar's surroundings 1 created. In another facility 10 A velocity vector is created from the information in the RD images and the RDA image, to which the radar's speed is added. 1 and the moving objects in the vicinity of the radar 1 can be extracted. Reference symbol list O1 Object O2 Object O3 Object Oi i-tes object p oi Position of the i-th object d oi Distance of the i-th object to the origin x x-axis world y y-axis world v Speed ​​of the radar ν x x-component of the velocity ν yy-component of the velocity Tx transmitting antenna radar Rx1 first receiving antenna radar Rx2 second receiving antenna p Tx Position of the transmitting antenna p Rx1 Position of the first receiving antenna p Rx2 Position of the second receiving antenna θ oi Angle of the i-th object in the radar coordinate system θ w,oi Angle of the ith object in the world coordinate system φ Angle between radar coordinate system – world coordinate system R Radar W Wall WR1 Angle Reflector 1 WR2 Angle Reflector 2 F1 Car 1 F2 Vehicle 2 F3 Vehicle 3 PM Parking Marking Rd1 Wheel 1 Rd2 Wheel 2 P Person v P Speed ​​of the person 1 radar 2 transmitting antennas 3 Receiving antenna 4 Receiving antenna 5 Control unit 6. Radar control and data acquisition 7. Transformation into Range Doppler Images (RD Images) 8 Transformation of two RD images into RDA images 9 Calculation of environment image 10 Calculating relative velocities QUOTES INCLUDED IN THE DESCRIPTION

[0084] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0085] DE 102010015723 A1

[0006] DE 102010051207 A1

[0007] DE 19912370 A1

[0008] DE 102009030075 A1

[0009] DE 102010048896 A1

[0010]

Claims

[1] Method for mapping the surroundings of a motor vehicle using an angle-measuring FMCW radar (R, 1 ) with range Doppler evaluation, where the FMCW radar (R, 1 ) at least one transmitting antenna (Tx, 2 ) and at least two receiving antennas (Rx1, Rx2, 3 , 4 ) shows, the FCMW radar (R, 1 ) has a predetermined angle (φ) to the direction of motion and the far-field approximation is applicable, the radar moving in a straight line at a constant speed (R, 1 ) emits signals towards objects (Oi, O1, O2, O3) in the vicinity of the motor vehicle for a limited period of time, the signals reflected by the objects (Oi, O1, O2, O3) from the at least two receiving antennas (Rx1, Rx2, 3 , 4 ) are received separately, the signals received within the limited time period constitute a group of M different measurement signals for each receiving antenna (Rx1, Rx2, 3 , 4 ) forms, Each group of M measurement signals is transformed into the frequency domain using a two-dimensional Fourier transform, and a range Doppler image of each group is generated. characterized by that Assigning a distance relative to the radar (R, 1 ) to each pixel of the range-Doppler image, to estimate the angles between radar (R, 1 ) and objects (Oi, O1, O2, O3) a conjugate multiplication of at least two range Doppler images is performed to create an RDA image, and From the distance information of the pixels of two range Doppler images and the angular information of the corresponding RDA image, an image of the environment of the radar (R) is generated. [2] Method according to claim 1, characterized bythat the Doppler shift of each pixel of the range Doppler images is corrected. [3] Method according to claim 1 or 2, characterized by that a three-dimensional image of the vehicle's surroundings can be created using a two-dimensional radar antenna (R, 1 ). [4] Method according to any of the preceding claims, characterized by that an interpolation of the pixels of the surrounding image is performed to generate an equidistant raster. [5] Method, in particular according to one of the preceding claims, for determining the relative vectorial velocity between a radar (R, 1 ) of a motor vehicle and objects (Oi, O1, O2, O3) in the vicinity of the motor vehicle using an angle-measuring FMCW radar (R, 1 ) with range Doppler evaluation, whereby the FMCW radar (R) at least one transmitting antenna (Tx, 2 ) and at least two receiving antennas (Rx1, Rx2, 3 ,4 ) shows, the FCMW radar (R, 1 ) has a predetermined angle (φ) to the direction of motion and the far-field approximation is applicable, the radar moving in a straight line at a constant speed (R, 1 ) emits signals in the direction of the objects (Oi, O1, O2, O3) in the vicinity of the motor vehicle for a limited period of time, the signals reflected by the objects (Oi, O1, O2, O3) from the at least two receiving antennas (Rx1, Rx2, 3 , 4 ) are received separately, the signals received within the limited time period constitute a group of M different measurement signals for each receiving antenna (Rx1, Rx2, 3 , 4 ) forms, Each group of M measurement signals is transformed into the frequency domain using a two-dimensional Fourier transform, and a range Doppler image of each group is generated. characterized by that to estimate the angles between radar (R, 1 ) and objects (Oi, O1, O2, O3) a conjugate multiplication of at least two range Doppler images is performed to create an RDA image, A radar velocity is determined for each pixel of the range Doppler images from the corresponding RDA image. [6] Method according to claim 5, characterized by that the radar velocities determined for all pixels are transferred into a one-dimensional velocity vector and local maxima are determined to ascertain the radar speed as well as the velocities of the moving objects. [7] Method according to claim 6, characterized by that the speed of the motor vehicle is used as the radar speed to separate the radar speed from the speeds of moving objects (Oi, O1, O2, O3). [8] Device for imaging the surroundings of a radar (R, 1) of a motor vehicle and to determine the relative vectorial velocity between the radar (R, 1 ) and objects (Oi, O1, O2, O3) in the vicinity of the motor vehicle, wherein the device for carrying out the method according to one of claims 1 to 4 is set up and designed, with an angle-measuring FMCW radar (R, 1 ) with range Doppler evaluation, where the FMCW radar (R, 1 ) at least one transmitting antenna (Tx, 2 ) and at least two receiving antennas (Rx1, Rx2, 3 , 4 ) exhibits, and a control unit ( 5 ) with an institution ( 6 ) to control the radar (R, 1 ) and for evaluating the received measurement signals, whereby the control unit ( 5 ) an institution ( 7 ) for determining range Doppler images characterized by that the control unit ( 5 ) further includes: an institution ( 8 ) to determine an RDA image from two range Doppler images by conjugate multiplication of the two range Doppler images, and an institution ( 9 ) to generate an image of the radar's surroundings (R, 1 ) from the distance information of the pixels of two range Doppler images and the angle information of the corresponding RDA image. [9] Device for determining the relative vectorial velocity between a radar (R, 1 ) of a motor vehicle and objects (Oi, O1, O2, O3) in the vicinity of the motor vehicle, in particular according to claim 8, wherein the device for carrying out the method according to one of claims 5 to 7 is set up and designed, with an angle-measuring FMCW radar (R, 1 ) with range Doppler evaluation, where the FMCW radar (R, 1 ) at least one transmitting antenna (Tx, 2) and at least two receiving antennas (Rx1, Rx2, 3 , 4 ) exhibits, and a control unit ( 5 ) with an institution ( 6 ) for controlling the radar and for evaluating the received measurement signals, wherein the control unit ( 5 ) an institution ( 7 ) for determining range Doppler images characterized by that the control unit ( 5 ) further includes: an institution ( 8 ) to determine an RDA image from two range Doppler images by conjugate multiplication of the two range Doppler images, and an institution ( 10 ) to determine the radar velocity for each pixel of the range Doppler images from the corresponding RDA image. [10] Device according to claim 9, characterized by that the facility ( 10) to determine the radar speed, the radar speeds determined for all pixels are incorporated into a one-dimensional velocity vector, and local maxima of the velocity vector are formed to determine the radar vehicle speed and the speeds of the moving objects (Oi, O1, O2, O3).

Images