DE102018117830A1 - Digital environment map with sensor ranges - Google Patents

Abstract

The present invention relates to a method for providing a sensor-data-based environment model of a vehicle, which comprises the steps: a) scanning the current environment of a vehicle by means of at least one sensor; b) determining the areas of the vehicle environment that can be detected by the sensor or sensors for the current position of the vehicle Vehicle from the sensor data obtained in step a); in a further step c) the vehicle environment determined in step b), which can be detected by sensors, is stored in a memory as a function of the current position and the sensor used. Furthermore, the present invention relates to a computer program product to carry out the method according to the invention, a driver assistance system having such a computer program product and a vehicle with such a driver assistance system.

Description

1. a) Abtasten der aktuellen Umgebung eines Fahrzeugs mittels mindestens eines Sensors;
2. b) Ermitteln des über den oder die Sensoren erfassbaren Bereiche der Fahrzeugumgebung für die aktuelle Position des Fahrzeuges aus den in Schritt a) erhaltenen Sensordaten; wobei in einem weiteren Schritt
3. c) die in Schritt b) ermittelte, durch Sensoren erfassbare Fahrzeugumgebung als Funktion der aktuellen Position und des verwendeten Sensors in einem Speicher abgelegt wird. Des Weiteren betrifft die vorliegende Erfindung ein Computerprogrammprodukt zur Ausführung des erfindungsgemäßen Verfahrens, ein Fahrassistenzsystem aufweisend ein solches Computerprogrammprodukt sowie ein Fahrzeug mit einem solchem Fahrassistenzsystem.

The present invention relates to a method for providing a sensor data-based environment model of a vehicle, which comprises the steps:

1. a) scanning the current environment of a vehicle by means of at least one sensor;
2. b) determining the areas of the vehicle surroundings that can be detected via the sensor or sensors for the current position of the vehicle from the sensor data obtained in step a); being in a further step
3. c) the vehicle environment determined in step b), which can be detected by sensors, is stored in a memory as a function of the current position and the sensor used. Furthermore, the present invention relates to a computer program product for executing the method according to the invention, a driver assistance system comprising such a computer program product and a vehicle with such a driver assistance system.

The reliability and safety of partially or fully autonomous driving systems essentially depends on the technical design and the interaction of the different sensor types. This is understandable because fail-safe and more powerful sensors should in principle be able to provide "better" data for the current position and the surroundings of a vehicle. The generation of a sufficient amount of data in principle is no longer the limiting factor due to the improvement of the hardware in recent years, be it on the part of the sensors themselves or on the part of the processing of the data supplied by them. Rather, there is a need for increasingly networked, "integral" data that is able to map the real driving situation as accurately as possible and to meet the increased demands on safety, even in complex and confusing driving situations. For these reasons, in addition to data collection as such, the processing of the delivered data is increasingly in the foreground.

For example, the DE 10 2014 212 216 A1 A method for detecting a vehicle environment with the aid of a vehicle sensor, the vehicle sensor recording image information from objects in the vehicle environment and objects in the vehicle environment being recognized on the basis of the image information. The reliability of the sensor information is improved in that a minimum sensor range of the vehicle sensor is determined on the basis of at least one object most distant from the vehicle and recognized. This method can be implemented as a computer program in a driver assistance system.

Furthermore, the DE 10 2016 203 833 A1 a method for controlling a motor vehicle comprising the steps of scanning an environment of the motor vehicle by means of a sensor and controlling the motor vehicle as a function of a scanning result on a predetermined route. A range of the sensor on an upcoming route section is determined depending on an external influencing factor and a planned speed on the route section is limited depending on the range of the sensor.

In another patent document, the DE 10 2013 215 098 A1 , A method of providing an environmental model for a vehicle is disclosed, comprising: providing an environmental model that is obtained based on sensor measurements from sensors of the vehicle; the environment model provides information regarding the occupancy of the environment by objects and the type of objects in the environment of the vehicle; Determining an area of the environment for which the environment model does not provide any information regarding occupancy by objects; wherein the area lies within a distance limit for which the environment model could provide information regarding the occupancy by objects based on sensor measurements; check whether a phantom object is to be added to the environment model in the area, in particular on the basis of predefined regulations; If a phantom object is to be added: determining the occupancy by the phantom object in the area of the environment; Generating an expanded environment model by adding information regarding the occupancy by the phantom object to the environment model; Provision of the extended environment model.

Such solutions known from the prior art can offer further potential for improvement, in particular with regard to an intelligent method in which data are not only collected but also linked to further data, so that a more complete and better image of the current driving situation can be generated.

It is the object of the present invention to at least partially overcome the disadvantages known from the prior art. It is in particular the object of the present invention to provide a solution by which partially or fully autonomous driving can be made safer, since sensor experience data can be stored as a function of the route position and reproducibly made available to other road users.

According to the invention, the object is achieved by a method having the features of claim 1. The object is also achieved by a computer program product with the features of claim 9, a driver assistance system with the features of claim 11, and a vehicle with the features of claim 12 Preferred embodiments of the invention are reproduced in the subclaims, in the description or in the figures, further features described or shown in the subclaims, in the description or in the figures, individually or in any combination, being able to constitute an object of the invention, if the opposite does not clearly result from the context.

1. a) Abtasten der aktuellen Umgebung eines Fahrzeugs mittels mindestens eines Sensors;
2. b) Ermitteln des über den oder die Sensoren erfassbaren Bereiche der Fahrzeugumgebung für die aktuelle Position des Fahrzeuges aus den in Schritt a) erhaltenen Sensordaten; wobei in einem weiteren Schritt
3. c) die in Schritt b) ermittelte, durch Sensoren erfassbare Fahrzeugumgebung als Funktion der aktuellen Position und des verwendeten Sensors in einem Speicher abgelegt wird.

The present invention relates to a method for providing a sensor data-based environment model of a vehicle, comprising the steps:

1. a) scanning the current environment of a vehicle by means of at least one sensor;
2. b) determining the areas of the vehicle surroundings that can be detected via the sensor or sensors for the current position of the vehicle from the sensor data obtained in step a); being in a further step
3. c) the vehicle environment determined in step b), which can be detected by sensors, is stored in a memory as a function of the current position and the sensor used.

Such a method can have significant advantages over the solutions from the prior art, in particular with regard to the susceptibility to errors and the safety in partially or fully autonomous driving processes, as well as when checking the sensors themselves. Surprisingly, it has been shown that the detectable sensor areas of the vehicle environment depend very much on the respective driving environment and represent a highly specific parameter for the present sensor (s) in this situation, which parameter can not only be used expediently to check the sensors themselves. Rather, the advantage of the method lies in the fact that a location-specific storage / storage of this data can be used to create a learning system which, when repeatedly moving to known positions or paths, means a significant gain in safety, since differences to sensor areas that have already been stored are recognized and errors in the system are detected Sensors or changes in the route are noticed. In addition, better and safer trajectories and speeds of fully or partially automatic systems can be determined by storing this information, since the physical range of the sensors and thus their “field of vision” have been determined and validated several times for this point. In this respect, speeds or trajectories that are based on assumed data or ranges that cannot be reached at precisely these positions can be excluded in these situations. This reduces the susceptibility to errors and the likelihood of possible dangerous situations. For example, the method according to the invention can represent a decision-making aid for overtaking processes. That Based on the stored data, it is known whether a vehicle in front can be overtaken, even if the road is not completely visible. This decision can be made on the basis of the information that the sensors should in principle be able to “see” an oncoming car on the road with a sufficient distance. The storage of the sensor position and the accumulation of the data in a memory also make it possible to provide estimates for sensor positions that have not yet been stored at this point. The latter can be done, for example, by extrapolating the last data of a point directly in front of the current position or data that has already been recorded in the past. In this respect, the method provides a learning system that can be used both for monitoring the sensors and for making decisions for safe driving trajectories.

A sensor data-based environment model is provided within the process. This means that location-specific information is recorded using the data stored in a memory. The area around the vehicle is, for example, individual information such as position, sensor ranges and / or sensor ranges as a function of the measurement angle and possibly further information, such as road guidance in front of or behind the vehicle, intersections or street names. In addition to the route, the vehicle environment can also include other information such as traffic signs, traffic control elements, speed limits, landmarks or the like. Information about dynamic, i.e. only objects appearing transiently, such as other vehicles or pedestrians.

The environment model is created for a vehicle. Vehicles can be objects driven by muscle power as well as by motors. Known vehicles are, for example, motor vehicles such as cars, trucks, motorcycles, bicycles, rail vehicles, hovercraft, airplanes or the like.

In step a), the current environment of a vehicle is scanned. This means that one or more sensors are used to deliver measured values at regular intervals about the immediate surroundings of the vehicle. Usually, different sensors have different ranges due to the design and method, and the spatial extent of the current environment is determined by the number, orientation and type of sensors used. It is possible that only the front of the vehicle counts to the current vehicle environment. This is the case, for example, in cases where the vehicle only has forward-facing sensors. In this context, the scanning means that a measurement is taking place by the sensor. A current position is the physical location at the specified time.

The current environment is scanned using at least one sensor. Possible and sensible sensors are, for example, ultrasound, radar or LIDAR sensors. The use of optical and laser distance sensors is also conceivable. Optical sensors can be cameras, for example, which are able to deliver two- or three-dimensional images of the vehicle environment.

In step b), the sensor area (s) for the current vehicle position is determined from the sensor data. The detectable sensor areas are the areas that can be detected by the sensors, so to speak, the detectable sensor field. The detectable sensor areas can include, for example, the angle-dependent range of the respective sensor. Further parameters of the detectable sensor area are, for example, the maximum and / or the minimum sensor range at the respective vehicle position. The maximum sensor range is the distance between a detected object and the sensor / vehicle, which is detected at this point in time and at this location by the specific sensor and which is the furthest away from the vehicle in comparison to all other objects. The maximum sensor range can be a function of the sensor type used and the ambient conditions such as weather and light conditions (day / night). The minimum sensor range is accordingly the distance between the next detected object and the vehicle or sensor position. These values can be determined continuously or within defined repetition frequencies. These frequencies can be selected as a function of the sensors used and / or the current environmental conditions. It is possible that this data is averaged over a certain period of time or a certain distance traveled, or that it is mathematically summarized or aggregated in some other way. This can improve the statistics of the process. Due to the object dependency of the method, minimum sensor ranges, maximum sensor ranges or both values can preferably be stored. It is also possible to add additional values as a function of the min./max. Determine and store sensor ranges. This includes, for example, the mean sensor range as a function of an angle segment or as a function of the weather conditions. Possible further data would be, for example, the type of vehicle that carries the sensors and the type of sensor used.

The sensor data are determined for the current position of the vehicle and stored as a function of this. The current position of the vehicle can be determined and stored using an external coordinate system, for example GPS coordinates. However, it is also possible for the current position to be determined only as a function of the position data supplied by the vehicle.

The sensor ranges determined in step b) are stored in a memory as a function of the current position and the sensor used. The ranges and sensor ranges that can be measured and determined from them are stored in a memory, usually an electronic memory. The storage takes place as a function of the current position and in this respect an entire data set for a specific location is stored in the memory. The memory can be the main memory of a computer of the vehicle, a hard disk or another memory that is not in or on the vehicle. In the latter case, the transmission of the data by the vehicle or by the sensors themselves to the external memory must of course be guaranteed. This can be done using conventional, for example wireless, methods such as GSM, Bluetooth or WLAN.

In a preferred embodiment of the method, the sensor areas determined in step b) can be stored in the form of the minimum and maximum sensor ranges. In particular, the minimum and maximum sensor ranges, as a function of the current vehicle position, have proven to be very reproducibly measurable and reliable. These parameters in particular can be stored easily and with little data expenditure and represent significantly the entire measurable sensor range. In this way, highly specific statements about the current vehicle environment can be made with small amounts of data.

In a preferred embodiment, the data stored in step c) can be combined with the data of a digital road map. A digital road map contains other location-specific sensor data location-specific data and features for localization (traffic signs, roadsides, lines, etc.) and possibly other information such as speed limits or the like. The map information can be formed, for example, by coordinates of the respective map. The digital card can be a high-precision card (HD card), for example. Furthermore, the map can contain other structural conditions or landmarks such as tunnels, mountains, bridges and other static information. This additional map information can optionally also be used to determine the position and / or to determine the sensor ranges. A landmark is understood to be an orientation point that can be recognized by a vehicle sensor and that allows a clear conclusion to be drawn about a unique coordinate within the map, in particular an electronic map. The map also expediently contains topographic information (height) in order to enable a particularly efficient combination with the detectable sensor areas. The combination of the determined sensor areas and the data of a digital map also enables an additional comparison between recognized objects and the objects on the digital map, which can contribute to a location of the vehicle position accurate to the cm. The latter option is based in particular on the further finding that it is possible, by means of the comparison, to make statements for the cases in which objects were recognized that the part of the street / map can be reached with the sensor.

In a further characteristic of the method, the sensor areas determined in step b) can be stored on the digital road map as a function of the vehicle position. A digital map expanded with the current, position-sensitive sensor areas can have a wide range of additional benefits. On the one hand, the sensors can be checked or calibrated by comparing them with known data. This can counteract the susceptibility of a sensor system to errors. On the other hand, the position of the vehicle can be determined from the sensor areas. For this purpose, for example, a maximum range of a sensor type can be determined in the digital map. This can be determined, for example, by the spatial environment (for example a tunnel). The position of the vehicle can then be determined by comparing the target value with the actual value.

In a further embodiment of the method, in addition to the data stored in step c), further data selected from the group consisting of the time, position of the sun, brightness, rain, fog, air humidity or combinations thereof can be stored. It has been found to be particularly advantageous to store, in addition to the sensor and position-specific stored data, also relevant environmental data such as weather conditions and or lighting conditions of this location in the memory. This additional information can increase the accuracy and reliability of the method by taking into account a possible influence of the sensor signals, and here in particular the ranges, by these dependencies. It is also possible to use this weather data, for example to store data that has been specially corrected or standardized to the lighting conditions.

In a further characteristic of the method, the maximum sensor range can be stored in step c). In particular, from the location-dependent maximum sensor range, many safety-relevant parameters can be derived, which can be helpful for a fully or partially autonomous process. This parameter defines the maximum distance in which information about the sensor can be supplied in the best case. For events outside this event horizon, the sensor is blind and reliable forecasts about possible trajectories or vehicle speeds can only be made with great reservations. Taking this feature into account in planning can therefore lead to a significant increase in security.

In a further aspect of the method, the digital road map expanded by the sensor areas can be stored in a digital memory which is designed to make the data available to other road users. The method according to the invention with learning cards / sensor data can thus be made more efficient by storing not only the experiences and data of a vehicle, but also of a vehicle community and, if necessary, already evaluating them statistically. In these cases, a vehicle no longer has to have passed a certain waypoint in order to be able to access sensor ranges. It is sufficient if another vehicle, preferably of the same type or with a similar sensor in a similar position, has already done this and stored the data. In this way, deviations from sensor ranges in different vehicles can also be determined. In the event of large deviations, for example, the recommendation can be made to have his sensors checked. This can reduce the susceptibility of sensors to errors, since the sensors can be practically permanently compared with comparable sensor data. In addition, the method can lead to a reduction in data, since when the sensor coverage is determined, only objects that are already within one must be passed on stop the determined "true" or "probable" sensor cover. Objects outside the detectable sensor area must then be based on incorrect measurements and can be excluded from the transfer. Appropriate storage would be, for example, cloud storage for these cases, and the determined ranges can be forwarded to such a central storage via wireless connections such as GSM or other transmitter / receiver models.

By means of a further design of the method, the data determined in step b) can be compared with existing, stored data before being stored in step c). The comparison of existing data with newly collected data can contribute to a further improvement of the data situation or to the collection of only situation-specific selected data. After calculating the detectable sensor areas, it can only be evaluated whether the ranges are outside a certain standard deviation of a large number of measured data. If this is the case, this data can be discarded or saved separately. The latter can help to detect new obstacles that change the sensor ranges from a certain point in time. However, it is also possible to store unusually short ranges separately, for example caused by moving objects. On the one hand, statements about the frequency of moving objects can be made, and on the other hand it is ensured that the maximum sensor range is only measured and stored in the absence of dynamic objects. A comparison can therefore include a mathematical change, a classification or a specific grouping of the data. The comparison can take place on the raw data or on statistical quantities of the raw data. Statistical variables are, for example, mean values as a function of certain angular ranges, minima, maxima, average ranges or also statistical data from the sensors as a function of the vehicle and the vehicle speed. Of course, it is also possible to store the data as raw data and to carry out the evaluations at recurring intervals or at a specific point in time, for example before the delivery of a vehicle.

In a further preferred embodiment of the method, the comparison between the data determined in step c) and the data already stored can take place on the basis of sensor areas that have already been determined and stored. It has proven to be particularly efficient that the data are compared to existing data before being stored in the memory and are either stored or discarded as a function of the comparison result. This process step can contribute to the fact that only statistically significant, that is really possible, data contribute to the construction of the model. This reduces the susceptibility of the method to errors and improves the fully or partially autonomous control options for vehicles.

In a preferred embodiment of the method, the comparison between the data determined in step c) and the data already stored can take place on the basis of the data of the digital map. The data comparison can be carried out not only with data of the same type, such as sensor ranges already stored, but also with other map data. This comparison with "foreign" stored data can increase the reliability of the process and contribute to a significant reduction in possible sources of error. The adjustment of the sensor ranges can be based, for example, on an adjustment with distance values between immovable objects on the map. However, it is also possible to allow a clearer deviation of the maximum or minimum sensor range by comparing it with zebra crossings listed on the map, for example, since, as expected, dynamic objects are to be expected more frequently at these points.

Furthermore, according to the invention is a computer program product that causes a computer to execute the method according to the invention when the computer program is executed. A computer program product comprises program code means for carrying out the method described above if the computer program product runs on a processing device or is stored on a computer-readable data carrier. The advantages of the computer program product according to the invention result as described above for the method according to the invention.

Also within the meaning of the invention is a digital road map, the digital road map comprising position-dependent sensor areas of different sensor types. A digital road map with location-specific or location-selective sensor ranges, which may still be clustered according to vehicle and / or sensor type, can contribute to a significant increase in the safety of the systems in the area of fully or partially autonomous driving. The same advantages apply to the digital card as described for the method according to the invention.

Furthermore, a driver assistance system according to the invention has a digital memory which has at least the program code of the computer program product according to the invention. Driver assistance systems increasingly relieve drivers of tasks or allow drivers to make decisions to review or support early or simultaneously. Basically, sensors and sensor systems are used to capture the vehicle surroundings using image information and to draw conclusions about driving behavior. These systems in particular can be better checked according to the invention. Furthermore, as described above, it is possible that these systems can submit safe driving suggestions or control them based on the information comparison.

Furthermore, the invention comprises a vehicle comprising sensors for detecting objects in the vehicle's surroundings and the driver assistance system according to the invention. For the advantages of the vehicle according to the invention, which has the driver assistance system according to the invention, reference is made explicitly to the advantages of the method according to the invention. It can be particularly preferred that the vehicle according to the invention is a motor vehicle.

QUOTES INCLUDE IN THE DESCRIPTION

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

Patent literature cited

• DE 102014212216 A1 [0003]
• DE 102016203833 A1 [0004]
• DE 102013215098 A1 [0005]

Claims (14)

Method for providing a sensor data-based environment model of a vehicle, comprising the steps: a) scanning the current environment of a vehicle by means of at least one sensor; b) determining the areas of the vehicle surroundings that can be detected via the sensor or sensors for the current position of the vehicle from the sensor data obtained in step a); characterized in that in a further step c) the vehicle environment determined in step b) and detectable by sensors is stored in a memory as a function of the current position and the sensor used. Procedure according to Claim 1 , the sensor areas determined in step b) being stored in the form of the minimum and maximum sensor ranges. Method according to one of the preceding claims, wherein the data stored in step c) are combined with the data of a digital road map. Procedure according to Claim 3 , wherein the sensor areas determined in step b) are stored as a function of the vehicle position on the digital road map. Method according to one of the preceding claims, wherein in addition to the data stored in step c), further data selected from the group consisting of the time, position of the sun, brightness, rain, fog, air humidity or combinations thereof are stored. Method according to one of the preceding claims, wherein in step c) the maximum sensor range is stored. Procedure according to one of the Claims 3 - 6 , wherein the digital road map expanded by the sensor areas is stored in a digital memory which is designed to make the data available to other road users. Method according to one of the preceding claims, wherein the data determined in step b) are compared with existing, stored data before being stored in step c). Procedure according to Claim 8 , whereby the adjustment is based on already determined and stored sensor areas. Procedure according to Claim 8 or 9 , whereby the comparison is based on the data of the digital map. Computer program product which causes a computer, when executing the computer program, to carry out a method according to one of the preceding claims. Digital road map, characterized in that the digital road map comprises position-dependent sensor areas of different sensor types. Driver assistance system comprising a digital memory containing at least the program code of the computer program product Claim 11 , Vehicle comprising sensors for detecting objects in the vehicle's surroundings and a driver assistance system Claim 13 ,