DE102018201304A1 - Method and device for estimating a maladjustment of a sensor of a vehicle - Google Patents

Abstract

The present invention relates to a method for estimating a misalignment of a sensor (100) of a vehicle (102), characterized in that in a step of determining a misalignment angle (a) between a nominal orientation (104) of the sensor (104) in the vehicle (102) and an actual orientation (106) of the sensor (100) in the vehicle (102) using an angular offset (ψ) between the nominal orientation (104) at a first measurement point (Mtl) and the nominal orientation (104) at a second measurement point (Mt2), a first vector (V1) detected at the first measuring point (Mtl) by the sensor (100) and relating to the alignment (106) at the first measuring point (Mtl) between the first measuring point (Mtl) and a detected object (P), and a at the second measuring point (Mt2) by the sensor (100) detected, on the alignment (106) at the second measuring point (Mt2) related second vector (V2) between the second measuring point (Mt2) and the object (P) be is true.

Description

Field of the invention

The invention relates to a method and an apparatus for estimating a maladjustment of a sensor of a vehicle.

State of the art

If a sensor for detecting a distance to an object and / or a direction to the object is rotated with respect to a proper orientation, it is misaligned. An actual detection range of the sensor is no longer consistent with a desired detection range. Captured objects appear to be located at positions where they are not actually located. As a result, sensor data of the sensor can only be used to a limited extent. The desired coverage area can be incompletely recorded. Safety functions and / or comfort functions that use the sensor data of the sensor may malfunction because the functions are based on the fact that the detected objects are actually located at the detected positions.

Disclosure of the invention

Against this background, a method for estimating a maladjustment of a sensor of a vehicle and a device for estimating a misalignment of a sensor of a vehicle, and finally a corresponding computer program product according to the independent claims are presented with the approach presented here. Advantageous developments and improvements of the approach presented here emerge from the description and are described in the dependent claims.

Advantages of the invention

Embodiments of the present invention can advantageously make it possible to detect a maladjustment of a sensor quickly and with little computational effort. A misalignment angle can be determined without knowing an airspeed. The approach presented here can be applied based on unprocessed raw data from the sensor. At a small misalignment of the sensor or when the determined misalignment angle is smaller than a threshold value, the determined misalignment angle can be used as a correction factor for correcting object positions imaged in sensor data. In the case of a strong maladjustment or if the determined misalignment angle is greater than a limit value, functions which use the sensor data can be deactivated because, for example, the detection range of the misadjusted sensor no longer covers a decisive part of an intended detection range.

A method for estimating a misalignment of a sensor of a vehicle is presented, which is characterized in that in a step of determining a misalignment angle between a nominal orientation of the sensor in the vehicle and an actual orientation of the sensor in the vehicle using
an angular offset between the nominal orientation in a first measuring point and the nominal orientation in a second measuring point,
a first vector detected by the sensor at the first measuring point and relating to the alignment at the first measuring point between the first measuring point and a detected object, and
a detected at the second measuring point by the sensor, based on the orientation at the second measuring point second vector between the second measuring point and the object is determined.

Ideas for embodiments of the present invention may be considered, inter alia, as being based on the thoughts and findings described below.

A sensor can detect objects using electromagnetic waves or sound waves. For example, the sensor can actively transmit a signal and receive an echo of the signal reflected on an object. Likewise, the sensor can only passively receive a signal coming from the object. The sensor may be, for example, a radar or lidar sensor. The sensor may also be an ultrasonic sensor. The sensor determines a distance to the object and / or a direction to the object relative to an actual orientation of the sensor. For example, the sensor may determine the distance over a transit time of the signal between the object and the sensor. The sensor can determine the direction, for example via a transit time difference of the signal, to at least two adjacently arranged receivers of the sensor. The direction can also be determined via at least one alignable transmitter and / or receiver.

If the sensor is misadjusted, it is rotated by a misalignment angle relative to a desired orientation. The desired orientation may be referred to as the nominal orientation. A vector is characterized by a vector amount and a vector direction. At least one vector can be detected for each measurement. The vector amount is a distance value representing a distance between the sensor and the object. The vector direction is through an angle is represented between a reference axis of the sensor and a connecting line between the object and the sensor. An angular offset may indicate a rotation of the vehicle between two measurements.

An object angle between the first vector and the second vector may be determined using the angular offset, a first vector direction of the first vector, and a second vector direction of the second vector. The object angle can be determined without knowledge of a distance between two measurement points. In particular, the object angle can be determined without knowledge of a speed of the vehicle

berechnet werden.The object angle or an object angle change can be calculated using the formula $$\beta ={\phi }_{t2}-{\phi }_{t1}-\psi$$

be calculated.

A distance between the first measurement point and the second measurement point may be determined using the object angle, a first vector amount of the first vector, and a second vector amount of the second vector. A distance between the measuring points can thus be determined without a measurement by vehicle sensors. As a result, sources of error can be excluded.

berechnet werden.The distance can be calculated using the formula $$S=\sqrt{{\mathrm{<figure-callout\; id="2"\; label="r\; t"\; filenames="DE102018201304A1\_0015.png"\; state="\{\{state\}\}">r\; </figure-callout>}}_t{}^2+r_{t1}{}^2-2r_{t1}{\mathrm{<figure-callout\; id="2"\; label="r\; t"\; filenames="DE102018201304A1\_0015.png"\; state="\{\{state\}\}">r\; </figure-callout>}}_t\text{cos}\beta }$$

be calculated.

The misalignment angle may be determined using the first vector direction, the second vector magnitude, the object angle, and the distance. The misalignment angle can be determined using a few easy-to-determine values and sizes. The misalignment angle can be quickly determined because little computing power is required to determine. The quantities used can be obtained from the raw data of the sensor and do not require any previous conversion.

berechnet werden.The misalignment angle can be calculated using the formula $$\alpha ={\phi }_{t1}-\text{<figure-callout id="2" label="arcsin r t" filenames="DE102018201304A1\_0015.png" state="{{state}}">arcsin </figure-callout>}\frac{r_t}{}\frac{2_{}\text{sin}\beta }{S}$$

be calculated.

In the step of determining, a static object can be used. At least one object that acts statically with respect to the vehicle can be used. For example, a fixed infrastructure device may be used to determine the misalignment angle. Moving objects, like other road users, can be ignored.

The misalignment angle can be determined by using at least one further vector detected by the sensor at a further measuring point and relating to the alignment at the further measuring point between the further measuring point and the detected object. The misalignment angle can be determined by several measurements with increased safety. The misalignment angle can also be determined continuously. For example, one current measurement and one previous measurement can be used for a determination. In a new measurement, the previous current measurement is used as the new past measurement.

The method can be implemented, for example, in software or hardware or in a mixed form of software and hardware, for example in a control unit.

The approach presented here also provides a device for estimating a misalignment of a sensor of a vehicle, which is designed to perform the steps of a variant of the method presented here in corresponding devices to drive or implement.

The device can be an electrical device having at least one computing unit for processing signals or data, at least one memory unit for storing signals or data, and at least one interface and / or a communication interface for reading in or outputting data embedded in a communication protocol, his. The arithmetic unit can be, for example, a signal processor, a so-called system ASIC or a microcontroller for processing sensor signals and outputting data signals as a function of the sensor signals. The storage unit may be, for example, a flash memory, an EPROM or a magnetic storage unit. The interface can be designed as a sensor interface for reading in the sensor signals from a sensor and / or as an actuator interface for outputting the data signals and / or control signals to an actuator. The communication interface can be designed to read in or output the data wirelessly and / or by cable. The interfaces can also Software modules that are present for example on a microcontroller in addition to other software modules.

Also of advantage is a computer program product or computer program with program code which can be stored on a machine-readable carrier or storage medium such as a semiconductor memory, a hard disk memory or an optical memory and for carrying out, implementing and / or controlling the steps of the method according to one of the embodiments described above is used, especially when the program product or program is executed on a computer or a device.

It should be noted that some of the possible features and advantages of the invention are described herein with reference to different embodiments as a method and apparatus. A person skilled in the art will recognize that the features can be suitably combined, adapted or replaced in order to arrive at further embodiments of the invention.

list of figures

• 1 zeigt eine Darstellung einer Fahrsituation für eine Schätzung einer Dejustage eines Sensors gemäß einem Ausführungsbeispiel; und
• 2 zeigt eine Prinzipdarstellung einer Schätzung einer Dejustage eines Sensors gemäß einem Ausführungsbeispiel.

Embodiments of the invention will now be described with reference to the accompanying drawings, in which neither the drawings nor the description are to be construed as limiting the invention.

• 1 shows a representation of a driving situation for an estimate of a misalignment of a sensor according to an embodiment; and
• 2 shows a schematic diagram of an estimate of a misalignment of a sensor according to an embodiment.

The figures are only schematic and not to scale. Like reference numerals designate the same or equivalent features in the figures.

Correct misalignment estimation of radar sensors is an important part of environment modeling. Errors in factory calibration, incorrect learning of the misalignment or too slow detection of misalignment, caused for example by parking bumps can lead to the wrong object model and thus to potentially safety-critical restrictions or malfunction of the system.

Embodiments of the invention

To compensate for a small misalignment angle of the radar sensor, the misalignment is estimated continuously while driving. This estimate is used to correct object information, such as an object's position. If the learned maladjustment exceeds a limit, functions can be switched off, since a correct function mapping may not be guaranteed.

By the approach presented here, increased requirements for the misalignment estimation with respect to an increased learning speed and an improved accuracy can be met. It can be achieved a very fast maladjustment calculation.

This improvement is due to the fact that on the one hand by the very fast misalignment determination the expected system performance is available quickly and on the other hand is dispensed with the speed of the own vehicle and the associated error sources. If the sensor is outside of a permissible working range, for example caused by a parking bump, its maladjustment is already determined after a few meters and causes, for example, a functional degradation.

The method presented here is suitable both for the permanent misalignment estimation and as a dynamic calibration option at the end of the belt and / or in the workshop in which the distance to be traveled is severely limited. Furthermore, the method can be used in many types of sensors, such as radar, video, ultrasound and lidar.

The calculation of the misalignment is based on the measurement of the distance and the angle to a stationary object P over at least two cycles from different measuring points. In particular, these data are available in polar coordinates. In addition, the relative rotation ψ of the sensor axis between two adjacent measuring points is required.

1 shows a representation of a driving situation for an estimate of a misalignment of a sensor 100 according to an embodiment. The sensor 100 is in a vehicle 102 installed. The sensor 100 is designed to distances and directions to objects in an environment of the vehicle 102 capture. The sensor 100 For example, it may be a radar sensor or a lidar sensor.

The sensor 100 is ideally in a nominal orientation 104 in the vehicle 102 installed. The nominal orientation 104 corresponds to an intended viewing direction of the sensor 100 , The sensor 100 should be aligned here in the direction of a vehicle longitudinal axis. If the vehicle 102 Damage can be an actual alignment 106 of the sensor 100 from the nominal orientation 104 differ. The front of the vehicle 102 is through here an impact deformed and the sensor deformed 100 sits awry in the vehicle. The sensor 100 is twisted by a misalignment angle α to the nominal orientation 104 , All recognized objects are recognized offset by the misalignment angle α.

A device 108 to appreciate the misalignment of the sensor 100 is designed to determine the misalignment angle α. These are at least two vectors V1 . V2 needed. The vectors V1 . V2 are detected at different times t1 and t2 and each represent a vector amount rt1 . r t2 and a vector direction φt1 . φt2 from the sensor 100 to an object P. The object P is in each case the same object P. The object P is here a guide post of a road on which the vehicle 102 moves. As the vehicle 102 moving between times, the vectors become V1 . V2 from different measuring points mt 1 . mt2 detected. Furthermore, an angular offset ψ of the sensor 100 between the two times T1 and T2 needed. The angular offset ψ is for example via an inertial sensor of a vehicle control unit 110 detected. In particular, the angular offset ψ by rotation rate sensors of the vehicle control unit 110 detected. The vehicle 102 drives along a curve of the road. The angular offset ψ between the measuring points mt 1 . mt2 in particular by a yaw angle of the vehicle 102 displayed.

The device 108 processes the vector amounts rt1 . r t2 and the vector directions φt1 . φt2 the vectors V1 . V2 and to calculate the angular offset ψ in a processing instruction by the misalignment angle α. The misalignment angle α is used as a correction factor for the sensor 100 provided.

2 shows a schematic diagram of an estimate of a misalignment of a sensor according to an embodiment. The schematic representation essentially corresponds to the driving situation in 1 , In contrast, the vehicle and the sensor and the environment are not shown here. In addition, other angular relationships are shown. By reducing to the pure geometric mapping, the trigonometric relationships can be improved.

t1 and t2 are the measuring times. P is the goal of the stand. mt 1 and mt2 are the measuring points at the time t1 or t2. rt1 and r t2 represent the radial distance at the time t1 or t2 from the respective measuring point mt 1 respectively mt2 to P.φ1 and Ø2 indicate the angles between the sensor axis and the actual orientation, respectively 106 and the vectors V1 respectively V2 at time t1 or t2. ψ is the rotation of the sensor axis at time t2 relative to t1. α is the misalignment angle of the actual alignment 106 from the nominal orientation 104 and S is the distance between mt 1 and mt2 ,

The second vector V2 hits from one by an angle β to the first vector V1 angular offset to the object P. From the triangle mt 1 mt2 P, the object angle β can be calculated: $$\left({\phi }_{t1}-\alpha \right)+\gamma +\beta =180\text{With}\gamma =\text{180}-\left({\phi }_{t2}-\alpha -\psi \right)$$

This results in $${\phi }_{t1}-{\mathrm{<figure-callout\; id="2"\; label="\phi \; t"\; filenames="DE102018201304A1\_0015.png"\; state="\{\{state\}\}">\phi \; </figure-callout>}}_t+\psi +\beta =0$$

Switched results $$\beta ={\phi }_{t2}-{\phi }_{t1}-\psi$$

Since the distance traveled between t1 and t2 does not necessarily represent the shortest path, the shortest distance S can be determined: $${\mathrm{<figure-callout\; id="2"\; label="S"\; filenames="DE102018201304A1\_0015.png"\; state="\{\{state\}\}">S</figure-callout>}}^2={\mathrm{<figure-callout\; id="2"\; label="r\; t"\; filenames="DE102018201304A1\_0015.png"\; state="\{\{state\}\}">r\; </figure-callout>}}_t{}^2+r_{t1}{}^2-2r_{t1}{\mathrm{<figure-callout\; id="2"\; label="r\; t"\; filenames="DE102018201304A1\_0015.png"\; state="\{\{state\}\}">r\; </figure-callout>}}_t\text{cos}\beta$$

This results in $$S=\sqrt{{\mathrm{<figure-callout\; id="2"\; label="r\; t"\; filenames="DE102018201304A1\_0015.png"\; state="\{\{state\}\}">r\; </figure-callout>}}_t{}^2+r_{t1}{}^2-2r_{t1}{\mathrm{<figure-callout\; id="2"\; label="r\; t"\; filenames="DE102018201304A1\_0015.png"\; state="\{\{state\}\}">r\; </figure-callout>}}_t\text{cos}\beta }$$

Now the misalignment angle α can be calculated: $$\frac{S}{\text{sin}\beta }=\frac{r_{t2}}{sin\left({\phi }_{t1}-\alpha \right)}$$

This results in $$sin\left({\phi }_{t1}-\alpha \right)=\frac{{\mathrm{<figure-callout\; id="2"\; label="r\; t"\; filenames="DE102018201304A1\_0015.png"\; state="\{\{state\}\}">r\; </figure-callout>}}_t\text{sin}\beta }{S}$$

And in turn $$\alpha ={\phi }_{t1}-\text{<figure-callout id="2" label="arcsin r t" filenames="DE102018201304A1\_0015.png" state="{{state}}">arcsin </figure-callout>}\frac{r_t}{}\frac{2_{}\text{sin}\beta }{S}$$

Finally, it should be noted that terms such as "comprising," "comprising," etc., do not exclude other elements or steps, and terms such as "a" or "an" do not exclude a multitude. Reference signs in the claims are not to be considered as limiting.

Claims (12)

A method of estimating a misalignment of a sensor (100) of a vehicle (102), characterized in that in a step of determining a misalignment angle (a) between a nominal orientation (104) of the sensor (104) in the vehicle (102) and an actual orientation (106) of the sensor (100) in the vehicle (102) using an angular offset (ψ) between the nominal orientation (104) at a first measurement point (Mtl) and the nominal orientation (104) at a second measurement point (Mt2), one at the first measuring point (Mtl) by the sensor (100) detected, on the alignment (106) at the first measuring point (Mtl) related first vector (V1) between the first measuring point (Mtl) and a detected object (P), and one the second measuring point (Mt2) detected by the sensor (100) and the second measuring point (Mt2) relative to the orientation (106) at the second measuring point (Mt2) is determined between the second measuring point (Mt2) and the object (P). Method according to Claim 1 in which, in the step of determining, an object angle (β) between the first vector (V1) and the second vector (V2) using the angular offset (ψ), a first vector direction (φt1) of the first vector (V1) and a second vector direction (φt2) of the second vector (V2) is determined. Method according to Claim 2 in which, in the step of determining, the object angle (β) is determined using the formula $$\beta ={\phi }_{t2}-{\phi }_{t1}-\psi$$

is calculated. Method according to one of Claims 2 to 3 in which, in the step of determining, further, a distance (S) between the first measuring point (Mtl) and the second measuring point (Mt2) using the object angle (β), a first vector amount (rt1) of the first vector (V1) and a second Vector amount (rt2) of the second vector (V2) is determined. Method according to Claim 4 in which, in the determining step, the distance (S) is determined using the formula $$S=\sqrt{r_{t2}{}^2+r_{t1}{}^2-2r_{t1}{r}_{t2}\text{cos}\beta }$$

is calculated. Method according to one of Claims 4 to 5 wherein, in the step of determining, the misalignment angle (a) is determined using the first vector direction (φt1), the second vector amount (rt2), the object angle (β), and the distance (S). Method according to Claim 4 in which, in the step of determining the misalignment angle (a) using the formula $$\alpha ={\phi }_{t1}-\text{arcsin}\frac{r_{t2}\text{sin}\beta }{S}$$

is calculated. Method according to one of the preceding claims, in which a static object (P) is used in the step of determining. Method according to one of the preceding claims, wherein in the step of determining the misalignment angle (a) using at least one at a further measuring point by the sensor (100) detected, on the alignment (106) at the further measuring point related further vector between the other Measuring point and the detected object (P) is determined. Device (108) for estimating a misalignment of a sensor (100) of a vehicle (102), the device being designed to implement, implement and / or control the method according to one of the preceding claims in corresponding devices. Computer program product adapted to perform the method according to any one of Claims 1 to 9 execute, implement and / or control. Machine readable storage medium carrying the computer program product according to Claim 11 is stored.

Images