DE102016212716A1 - CONTROL DEVICE AND METHOD - Google Patents

Abstract

The present invention discloses a control device (1, 20) for a vehicle, having a first detection device (3, 6, 22, 25) which is designed to transmit vehicle information (4, 7, 23, 26) via the vehicle (2, 21 ) or the surroundings of the vehicle, and with a computing device (5, 8, 24, 27, 28), which is configured, the detected vehicle information (4, 7, 23, 26) based on the vehicle information (4, 7, 23, 26) of a second detection device (3, 6, 22, 25) to process. Furthermore, the present invention discloses a corresponding method.

Description

Field of the invention

The present invention relates to a control device for a vehicle and a corresponding method.

State of the art

Although the present invention will be described below mainly in connection with passenger cars, it is not limited thereto, but may be used with any type of vehicle.

In modern vehicles a variety of vehicle functions is used, which support the driver when driving the vehicle or take this task, so move the vehicle autonomously.

In order to enable an autonomous movement of the vehicle, the vehicle environment for the respective vehicle systems must be detected, so that z. B. can plan a trajectory for the vehicle.

The vehicle environment can be z. B. are detected by different sensors whose data are merged to obtain a meaningful model on the vehicle environment.

Disclosure of the invention

It is an object of the present invention to enable improved detection of the vehicle environment.

This object is achieved by a control device having the features of independent patent claim 1 and a method having the features of independent claim 8.

Accordingly, it is provided:
A control device for a vehicle, having a first detection device, which is configured to detect vehicle information about the vehicle or the surroundings of the vehicle, and having a computing device, which is configured to process the detected vehicle information based on the vehicle information of at least one second detection device.

It is also provided:
A method for processing sensor data for a vehicle, comprising the steps of: acquiring vehicle information about the vehicle or the surroundings of the vehicle with a first detection device, acquiring vehicle information about the vehicle or the surroundings of the vehicle with at least one second detection device, and processing the vehicle detected vehicle information of the first detection device based on the vehicle information of the second detection device.

Advantages of the invention

Usually, data from different sensors are combined in a vehicle, for. B. to create an environment model of the vehicle environment. For this purpose, the individual sensor systems of the vehicle provide sensor data z. B. to a higher-level control unit, which performs a so-called. Sensor data fusion and generates the environment model.

The present invention is based on the recognition that it may be helpful at an earlier stage in the signal chain to obtain information from other sensors to a sensor.

Such data can z. B. provide important information for an internal sensor signal conditioning. For example, a camera may tell the radar sensor if it detects that it is raining or snowing. The radar sensor can then adapt its internal signal processing to these changed environmental conditions.

The invention thus derives context information from the vehicle information of a detection device, which can help to determine or correct the reduced efficiency of other detection devices. In this case, all information which is acquired or collected by the detection devices via the vehicle can be referred to as vehicle information. The vehicle information can be provided by the detection devices as raw data. Additionally or alternatively, the vehicle information may be preprocessed, e.g. At feature level or decision level. The feature level denotes vehicle information, the information about features, ie z. B. recognized edges and corners in an image, have. The decision level refers to vehicle information, the information about functions to be performed, so z. As the decision on a vehicle trajectory to drive, have. Here, the features can be processed into an object hypothesis and so z. B. recognized objects are identified. The vehicle information can also be edited, so z. B. filtered, shape and / or as raw data or predicted data or in each case as a local environment model.

The present invention thus enables z. B. an improved signal processing in the individual sensor systems. Of course, the individual sensor systems can have more than one detection device. For example, a surround view system may include a plurality of cameras.

In particular, individual sensors can be informed about environmental or vehicle conditions that are relevant for their signal acquisition or signal processing or influence them.

It is understood that the control device can also carry out further control tasks in the vehicle or can provide the edited vehicle information to other control devices.

Advantageous embodiments and further developments emerge from the dependent claims and from the description with reference to the figures.

In an embodiment, the first detection means and the second detection means may use different physical detection principles. For example, the first detection device may be a radar sensor, while the second detection device z. B. may be a camera. Other possible types of sensors are z. As ultrasonic sensors, lidar sensors, inertial sensors, GPS sensors or yaw rate sensors. Any combination of the mentioned or other sensors is possible.

In one embodiment, the computing device may be arranged with the first detection device in a first sensor system. The second detection device can be arranged in a second sensor system. The computing device thus forms, together with the first detection device, a single, separate sensor system of a vehicle. For example, this sensor system may be a radar system that provides other vehicle systems information about the objects that a radar sensor has detected.

The computing device can perform an internal measurement value processing and from the Radarrohdaten z. B. generate corresponding information for the other vehicle systems. The computing device can, for. B. also generate a local environment model in the sensor system and provide it to other vehicle systems. If several computing devices in different sensor systems create their own models and a central environment model exists, this can also be referred to as a hybrid environment model.

The vehicle information of the second detection device, the computing device z. B. via a vehicle bus, z. For example, CAN or FlexRay.

In one embodiment, the computing device may retrieve the vehicle information from the individual detection devices independently of the first detection device and the second detection device. The computing device can thus z. B. may be a higher-level computing device that is not associated with any special sensor. Such a higher-level computing device can, for. B. serve for sensor data fusion in the vehicle and create an environment model or a model Eg, ie a physical vehicle model of the vehicle. For this purpose, the computing device can link a large number of sensor data, that is, fuse it.

The sensor data fusion can be adaptive, not static, based on the vehicle information of the individual detection devices. In this case, the computing device z. B. from a camera receive the information that it is snowing and this information in the fusion of the sensor data z. As a radar, a Lidarsensors or the like into account. For example, the information about a detected by the radar tunnel can be used to z. B. in lidar sensors or camera-based sensors to establish appropriate interference. The computing device can, for. B. the confidence levels, so a measure of confidence in the respective sensor values, adjust accordingly. This allows a more reliable self-assessment of the vehicle.

If the sensor fusion takes place centrally in the vehicle and a corresponding environmental model is generated, this can also be referred to as a central environmental model.

Of course, a local computing device of a sensor system can fulfill these tasks and provide the vehicle model to other systems via a vehicle bus.

In one embodiment, the computing device may be configured to create a confidence model for the detection devices, in which the computing device calculates the confidence values for one of the detection devices based on the vehicle information of at least one other of the detection devices. In addition to the environment model, the computing device additionally creates a confidence model in which a type of confidence level for the individual detection devices is determined. The confidence values provide a measure of trust in the company respective detection device on. Of course, the confidence model can also be integrated into a corresponding environment model.

In one embodiment, the first detection device can be designed as a camera-based detection device. The second detection device can be designed as a radar-based detection device. Furthermore, the computing device may be configured to retrieve data about detected objects from the second detection device and to analyze the image sections in the data of the camera-based detection device, which correspond to the detected objects, with increased accuracy. A radar usually has a greater range than a camera system and can identify objects directly from its measurements without using complex signal or image processing algorithms. On the other hand, a camera with complex image analyzes enables an exact statement about the type and type of the object. If the presence of an object is thus detected by a radar system, this information can be used to more intensively process those image sections in the camera image, eg. With higher resolution or more complex algorithms corresponding to the detected object.

In an embodiment, the first detection device may be designed as a radar-based detection device. Furthermore, the second detection device can be designed as a camera-based detection device. The computing device can be configured to retrieve data about the weather situation from the second detection device and to correct the noise level in the data of the radar device in accordance with the weather situation. Radar sensors can be affected by snowflakes, rain drops or mist drops in the air. In particular, this increases their noise level. If the camera detects such a weather situation, this can be used in the evaluation of the radar data in order to correct it or at least to explain its high noise level.

The above embodiments and developments can, if appropriate, combine with each other as desired. Further possible refinements, developments and implementations of the invention also include combinations of features of the invention which have not been explicitly mentioned above or described below with regard to the exemplary embodiments. In particular, the person skilled in the art will also add individual aspects as improvements or additions to the respective basic form of the present invention.

Brief description of the drawings

The present invention will be explained in more detail with reference to the exemplary embodiments indicated in the schematic figures of the drawings. It shows:

1 a block diagram of an embodiment of a control device according to the invention;

2 a block diagram of another embodiment of a control device according to the invention; and

3 a flow diagram of an embodiment of a method according to the invention.

In all figures, the same or functionally identical elements and devices - unless otherwise stated - have been given the same reference numerals.

Embodiments of the invention

1 shows a block diagram of an embodiment of a control device according to the invention 1 in a vehicle 2 is arranged. The control device 1 is used in particular for sensor data processing, but can also be used for more extensive control tasks such. As an autonomous vehicle control run.

In the vehicle 2 is one as a camera 3 trained first detection device arranged which camera images 4 and to the computing device 5 transfers. The camera 3 and the computing device 5 together form a first sensor system 9 , In the vehicle 2 is also a radar sensor 6 trained second detection device arranged radar data 7 captured and sent to a computing device 8th outputs. The radar sensor 6 and the computing device 8th together form a second sensor system 10 , The computing device 8th processes the radar data 7 , thus generates z. B. an environment model 12 ,

The radar sensor 6 transmits the radar data 7 in addition also to the computing device 5 , which from the camera pictures 4 and the radar data 7 also an environment model 11 generated. The environment models 11 and 12 can z. B. as local environment models 11 and 12 to a central control device (not shown separately), which performs a sensor fusion and provides a central environment model for other vehicle systems.

The computing device 5 can the radar data 7 received as raw data and evaluate, so z. B. Detecting objects. Alternatively, the computing device 5 but also already preprocessed radar data 7 which only contain information about detected objects. Based on the radar data 7 extracted or contained in this information, the computing device 5 then the camera pictures 4 to process. For example, the computing device 5 the camera pictures 4 at those positions undergo an intensive analysis that in the camera images 4 in the radar data 7 correspond to detected objects.

It is understood that the arrangement described above is merely exemplary. The camera 3 can the camera pictures 4 or z. B. information about detected rain, snow or fog, even to the computing device 8th to transfer. The computing device 8th This can then be used to z. B. increased noise in the radar data 7 to explain or eliminate, so z. B. to calculate. Furthermore, the sensor system 9 For example, have more cameras or the sensor system 10 can have more than one radar sensor.

2 shows a control device 20 in the vehicle 21 is arranged. The control device 20 also has two sensor systems 31 . 32 on. As in 1 has the sensor system 31 a camera 22 on, that of a computing device 24 camera images 23 provides. The sensor system 32 has a radar sensor 25 on top of a computing device 27 radar data 26 transmitted.

The two sensor systems 31 . 32 each give an environment model 33 . 34 and transmit this to a central computing device 28 , The computing device 28 generated from the environment models 33 . 34 a central environment model 29 and makes this available to other vehicle functions or control devices. At the same time, the computing device generates 28 a confidence model in which the computing device 28 in each case the data of the individual sensor systems 31 . 32 based on the data of at least one other sensor system 32 . 31 and assign them a confidence score. The confidence value can be z. B. indicate a likelihood that the respective vehicle information is correct.

It is understood that in the 2 shown arrangement is merely exemplary in nature. Other arrangements are also possible. For example, the central control device 28 be integrated into another vehicle system and perform the sensor data fusion in addition to the function of the vehicle system as described above.

3 shows a flowchart of a method for processing sensor data.

In a first step S1 vehicle information 4 . 7 . 23 . 26 about the vehicle 2 . 21 or the environment of the vehicle 2 . 21 with a first detection device 3 . 6 . 22 . 25 detected. In step S2, further vehicle information 4 . 7 . 23 . 26 about the vehicle 2 . 21 or the environment of the vehicle 2 . 21 with a second detection device 3 . 6 . 22 . 25 detected. In this case, the first and second detection means 3 . 6 . 22 . 25 based on different physical measuring principles. For example, one of the measuring devices 3 . 6 . 22 . 25 use an optical measuring method while the other uses an electromagnetic measuring method.

Finally, in step S3, the detected vehicle information 4 . 7 . 23 . 26 the first detection device 3 . 6 . 22 . 25 based on the vehicle information 4 . 7 . 23 . 26 the second detection device 3 . 6 . 22 . 25 processed.

The arrangement of the individual detection devices 3 . 6 . 22 . 25 is flexible. For example, the detection and processing of the vehicle information 4 . 7 . 23 . 26 the first detection device 3 . 6 . 22 . 25 in a first sensor system 9 . 10 be performed. The vehicle information 4 . 7 . 23 . 26 the second detection device 3 . 6 . 22 . 25 can in contrast in a second sensor system 9 . 10 be recorded. The individual sensor systems 9 . 10 So can local computing devices 5 . 8th . 24 . 27 exhibit.

Alternatively, the vehicle information 4 . 7 . 23 . 26 the first detection device 3 . 6 . 22 . 25 and the vehicle information 4 . 7 . 23 . 26 the second detection device 3 . 6 . 22 . 25 be retrieved centrally for processing. So z. B. in a central computing device 28 are processed.

In addition to pure data processing, z. B. for object recognition, can be a confidence model 30 for the detection devices 3 . 6 . 22 . 25 to be created. In this can provide confidence values for vehicle information 4 . 7 . 23 . 26 each one of the detection devices 3 . 6 . 22 . 25 based on the vehicle information 4 . 7 . 23 . 26 at least one other of the detection devices 3 . 6 . 22 . 25 be calculated.

The detection devices 3 . 6 . 22 . 25 can use different measuring principles. For example, the vehicle information 4 . 7 . 23 . 26 the first detection device 3 . 6 . 22 . 25 Camera-based capture. The vehicle information 4 . 7 . 23 . 26 the second detection device 3 . 6 . 22 . 25 can be detected, for example, radar-based. When processing the vehicle information 4 . 7 . 23 . 26 can then from the second detection device 3 . 6 . 22 . 25 Retrieve data about detected objects. This information can be used to display the image sections in the image data of the first detection device 3 . 6 . 22 . 25 which correspond to the detected objects to analyze with increased accuracy.

But it can also the vehicle information 4 . 7 . 23 . 26 the first detection device 3 . 6 . 22 . 25 Radar-based capture and vehicle information 4 . 7 . 23 . 26 the second detection device 3 . 6 . 22 . 25 Camera-based capture. From the second detection device 3 . 6 . 22 . 25 can then z. B. data on the weather, z. As rain, snow, fog, be retrieved and the noise level in the data of the radar device are corrected according to the weather.

Although the present invention has been described above with reference to preferred embodiments, it is not limited thereto, but modifiable in a variety of ways. In particular, the invention can be varied or modified in many ways without deviating from the gist of the invention.

LIST OF REFERENCE NUMBERS

1, 20

 control device

2, 21

 vehicle

3, 6, 22, 25

 detector

4, 7, 23, 26

 vehicle information

5, 8, 24, 27, 28

 computing device

9, 10, 31, 32

 sensor system

11, 12, 29, 33, 34

 environment model

30

 Konfidenzmodell

S1-S3

 steps

Claims (14)

Control device ( 1 . 20 ) for a vehicle, with a first detection device ( 3 . 6 . 22 . 25 ), which is designed to receive vehicle information ( 4 . 7 . 23 . 26 ) over the vehicle ( 2 . 21 ) or the environment of the vehicle, and with a computing device ( 5 . 8th . 24 . 27 . 28 ), which is formed, the detected vehicle information ( 4 . 7 . 23 . 26 ) based on the vehicle information ( 4 . 7 . 23 . 26 ) a second detection device ( 3 . 6 . 22 . 25 ) to process. Control device ( 1 . 20 ) according to claim 1, wherein the first detection device ( 3 . 6 . 22 . 25 ) and the second detection device ( 3 . 6 . 22 . 25 ) use different physical detection principles. Control device ( 1 . 20 ) according to one of claims 1 and 2, wherein the computing device ( 5 . 8th . 24 . 27 . 28 ) with the first detection device ( 3 . 6 . 22 . 25 ) in a first sensor system ( 9 . 10 . 31 . 32 ) and wherein the second detection device ( 3 . 6 . 22 . 25 ) in a second sensor system ( 9 . 10 . 31 . 32 ) is arranged. Control device ( 1 . 20 ) according to one of the preceding claims, wherein the computing device ( 5 . 8th . 24 . 27 . 28 ) independent of the first detection device ( 3 . 6 . 22 . 25 ) and the second detection device ( 3 . 6 . 22 . 25 ) the vehicle information ( 4 . 7 . 23 . 26 ) of the individual collection facilities ( 3 . 6 . 22 . 25 ) retrieves. Control device ( 1 . 20 ) according to claim 4, wherein the computing device ( 5 . 8th . 24 . 27 . 28 ), a confidence model ( 30 ) for the detection devices ( 3 . 6 . 22 . 25 ) in which the computing device ( 5 . 8th . 24 . 27 . 28 ) the confidence levels for each of the collection facilities ( 3 . 6 . 22 . 25 ) based on the vehicle information ( 4 . 7 . 23 . 26 ) at least one other of the detection devices ( 3 . 6 . 22 . 25 ). Control device ( 1 . 20 ) according to one of the preceding claims 1 to 5, wherein the first detection device ( 3 . 6 . 22 . 25 ) as a camera-based detection device ( 3 . 6 . 22 . 25 ) and the second detection device ( 3 . 6 . 22 . 25 ) as a radar-based detection device ( 3 . 6 . 22 . 25 ), and wherein the computing device ( 5 . 8th . 24 . 27 . 28 ) is formed by the second detection device ( 3 . 6 . 22 . 25 ) Retrieve data about detected objects and the image sections in the data of the camera-based detection device ( 3 . 6 . 22 . 25 ), which correspond to the detected objects, to analyze with increased accuracy. Control device ( 1 . 20 ) according to one of the preceding claims 1 to 5, wherein the first detection device ( 3 . 6 . 22 . 25 ) as a radar-based detection device ( 3 . 6 . 22 . 25 ) and the second detection device ( 3 . 6 . 22 . 25 ) as a camera-based detection device ( 3 . 6 . 22 . 25 ), and wherein the computing device ( 5 . 8th . 24 . 27 . 28 ) is formed by the second detection device ( 3 . 6 . 22 . 25 ) To retrieve data about the weather conditions and to correct the noise level in the data of the radar device according to the weather conditions. Method for processing sensor data for a vehicle, comprising the steps of: detecting (S1) vehicle information ( 4 . 7 . 23 . 26 ) over the vehicle ( 2 . 21 ) or the environment of the vehicle ( 2 . 21 ) with a first detection device ( 3 . 6 . 22 . 25 ), Detecting (S2) vehicle information ( 4 . 7 . 23 . 26 ) over the vehicle ( 2 . 21 ) or the environment of the vehicle ( 2 . 21 ) with a second detection device ( 3 . 6 . 22 . 25 ), and processing (S3) the captured vehicle information ( 4 . 7 . 23 . 26 ) of the first detection device ( 3 . 6 . 22 . 25 ) based on the vehicle information ( 4 . 7 . 23 . 26 ) of the second detection device ( 3 . 6 . 22 . 25 ). Method according to claim 8, wherein the vehicle information ( 4 . 7 . 23 . 26 ) of the first detection device ( 3 . 6 . 22 . 25 ) and the vehicle information ( 4 . 7 . 23 . 26 ) of the second detection device ( 3 . 6 . 22 . 25 ) are detected with different physical detection principles. Method according to one of claims 8 and 9, wherein the detection and processing of the vehicle information ( 4 . 7 . 23 . 26 ) of the first detection device ( 3 . 6 . 22 . 25 ) in a first sensor system ( 9 . 10 . 31 . 32 ) and capturing the vehicle information ( 4 . 7 . 23 . 26 ) of the second detection device ( 3 . 6 . 22 . 25 ) in a second sensor system ( 9 . 10 . 31 . 32 ) is carried out. Method according to one of the preceding claims 8 to 10, wherein the vehicle information ( 4 . 7 . 23 . 26 ) of the first detection device ( 3 . 6 . 22 . 25 ) and the vehicle information ( 4 . 7 . 23 . 26 ) of the second detection device ( 3 . 6 . 22 . 25 ) are retrieved for processing centrally. The method of claim 11, wherein a confidence model ( 30 ) for the detection devices ( 3 . 6 . 22 . 25 ) in which confidence values for vehicle information ( 4 . 7 . 23 . 26 ) one of the detection devices ( 3 . 6 . 22 . 25 ) based on the vehicle information ( 4 . 7 . 23 . 26 ) at least one other of the detection devices ( 3 . 6 . 22 . 25 ) be calculated. Method according to one of the preceding claims 8 to 12, wherein the vehicle information ( 4 . 7 . 23 . 26 ) of the first detection device ( 3 . 6 . 22 . 25 ) Camera-based capture and vehicle information ( 4 . 7 . 23 . 26 ) of the second detection device ( 3 . 6 . 22 . 25 ) Are detected radar-based, and wherein the second detection device ( 3 . 6 . 22 . 25 ) Data on detected objects are retrieved and the image sections in the image data of the first detection device ( 3 . 6 . 22 . 25 ), which correspond to the detected objects, are analyzed with increased accuracy. Method according to one of the preceding claims 8 to 12, wherein the vehicle information ( 4 . 7 . 23 . 26 ) of the first detection device ( 3 . 6 . 22 . 25 ) Radar-based capture and vehicle information ( 4 . 7 . 23 . 26 ) of the second detection device ( 3 . 6 . 22 . 25 ) Are detected on a camera-based basis, and wherein the second detection device ( 3 . 6 . 22 . 25 ) Data about the weather conditions are retrieved and the noise level is corrected in the data of the radar device according to the weather conditions.

Images