Patent Application
20190088133
This patent application claims priority to European Patent Application No. 17191248.8, filed 15 Sep. 2017, the disclosure of which is incorporated herein by reference in its entirety.
Illustrative embodiments relate to a method for planning a collision avoidance maneuver by use of a car-to-car communication method. Illustrative embodiments further relate to a corresponding control unit, and a transportation vehicle equipped with such a control unit and a corresponding computer program.
An exemplary embodiment is illustrated in the drawings and is explained in more detail below with reference to the figures. In the drawings:
Autonomous driving (sometimes referred to as automatic driving, automated driving or piloted driving) is the movement of transportation vehicles, mobile robots and driverless transport systems which are largely autonomous. There are different degrees of autonomous driving. In this case, autonomous driving is also spoken at certain levels even if a driver is still present in the transportation vehicle, who possibly only takes over the monitoring of the automatic driving operation. In Europe, the various transport ministries (in Germany, the Federal Institute for Road Systems (Bundesanstalt für Straßenwesen) was involved) worked together and defined the following autonomous stages.
Future cooperative driving applications are envisioned to significantly improve efficiency and comfort of automated driving. Increase in cooperation between transportation vehicles pose specific requirements on communication system and the way messages are exchanged among cooperating transportation vehicles. The field of use is from Level 2 onwards and concerns the important facet of collision avoidance. Such applications more in particular the cooperative collision avoidance pose very strict requirements on the overall time to build a cooperative decision and require simple, reliable and efficient solution.
In these systems, an exact position determination of the transportation vehicle from stage to stage is important as well as a precise observation of the surroundings of the transportation vehicle. Position determination techniques are known which take into account absolute position data and odometry position data in determining the estimated position of the transportation vehicle.
In this case, the absolute position data indicate the measured position of the transportation vehicle at a certain time in absolute values, for example, in a UTM or WGS84 reference coordinate system. Optionally, the absolute position data may also be provided with an orientation, e.g., a current direction of movement of the transportation vehicle. A combination of position and orientation is often referred to as a pose. The position is usually expressed two-dimensionally in a Cartesian coordinate system.
The odometry position data, on the other hand, indicate a self-movement or relative movement of the transportation vehicle, for example, in an arbitrarily defined reference coordinate system or in the so-called transportation vehicle coordinate system (“body frame”). In this case, the self-movement, is indicated in each case relative to a preceding position of the transportation vehicle.
The cooperative collision avoidance application is one of the key applications in the field of cooperative vehicular systems. There are multiple approaches known in the literature which propose methods in realizing the message exchange flow.
US 2013/0325306 A1 describes a cooperative driving and collision avoidance system with distributed receding horizon control. A controller in a coordinating transportation vehicle comprises a communication terminal configured to receive trajectory messages from a plurality of second coordinating transportation vehicles in a communication range, the trajectory messages including transportation vehicle trajectory information for a predetermined update interval. A computer in the controller performs the following tasks:
Task 1: Generate a nominal “assumed trajectory” for the next planning horizon, which may involve solving an optimization problem.
Task 2: Exchange (transmit and receive) assumed trajectories with each neighbor.
Task 3: Check assumed trajectories (self against neighbors) for conflicts, and resolve any conflicts as necessary. Computing a conflict-resolving maneuver involves solving an optimization problem.
Task 4: Solve an optimization problem to generate the next implemented maneuver, if not already done in tasks 1 or 3. This is required when initial maneuvers are simple to compute (i.e., no optimization required in task 1) and such maneuvers do not result in a conflict (no optimization required in task 3).
From WO 2016/020290 A1 a system for automated cooperative driving is known. This system is comprising a position determining unit for determining the position of a transportation vehicle, a reception unit for receiving surrounding data from at least one surrounding transportation vehicle, a signal processing unit for calculating a trajectory of the own transportation vehicle by information about the own transportation vehicle, wherein, by the surrounding data, the signal processing unit calculates whether the trajectory of the own transportation vehicle cuts a trajectory of an object or of the second transportation vehicle, and hence whether the transportation vehicle (Vx1) collides with this object and/or the second transportation vehicle.
For the cooperative or autonomous driving, the exchange of certain messages of the transportation vehicles among each other is very important. Such information included in messages may be position information, surroundings information, trajectory information, warning information, control information and more.
All these types of information require an extensive message exchange. Moreover, to come to an agreement about a maneuver, further messages need to be exchanged to build the common cooperative decision, about the coordinated maneuver. There is an article from Kai Franke; Mark Gonter; Michael During; Karsten Lemmer; Reza Balaghiasefi; Ferit Küçükay, which describes this technique. The article has the title “A reference architecture for CISS/CDAS within the field of cooperative driving,” and was published on 2014 International Conference on Connected Vehicles and Expo (ICCVE), 2014. The article describes the following approach: In this case first one transportation vehicle detects the potential collision problem and triggers the start of the cooperative collision avoidance maneuver planning by sending a corresponding message. After this, the transportation vehicles which are selected or have agreed on cooperation, exchange the information about the environment among each other. As a next operation the involved transportation vehicles based on the exchanged information calculate their own and the trajectories for other involved transportation vehicles. Once the suggested trajectories were calculated, each transportation vehicle rates each solution based on its own ego preferences and sends ranks and trajectories to other involved partners. Once the exchange process has been completed, each transportation vehicle compares the ranks of all received suggested trajectories and selects the one with the maximum overall ranking. The selected choice is then being sent to other partners to make all partners understand the selected own trajectory and the overall maneuver.
In DE 10 2015 221 817 A1 another proposed idea is described, which assumes an implicit information exchange without any acknowledgment on the selected path at all. In this case each transportation vehicle periodically sends the planned as well as one alternative planned trajectory on the information about environment and information received from other partners. In this approach each transportation vehicle assumes that other transportation vehicles, once they receive information about a planned trajectory of its neighbor, would correspondingly adopt their own plan such that the maneuver of all transportation vehicles is optimized.
The disclosure identified different problems with these approaches described above. The main challenge with the approach described in the article “A reference architecture for CISS/CDAS within the field of cooperative driving” is that it requires extensive message exchange among partners within a very short time over an error prone communication network.
The problem with the other approaches is that the algorithm assumes that every new sent trajectory update ultimately leads to the direction of the global optima for the given co-operative maneuver. In reality, it may lead to a local optima or even result in a constant oscillation of a next operation at the same level without any move into the direction of global optima.
There is therefore a need for an improved approach for a collision avoidance system which avoids the above mentioned drawbacks.
Disclosed embodiments provide a method for planning a collision avoidance maneuver, a computing unit, a transportation vehicle, and a computer program.
The solution according to the proposal comprises the following operations of observing the surroundings of a transportation vehicle and determining the risk of a collision. If there is a risk of a collision, starting an information exchange phase for the involved transportation vehicles where the information about the observations from the other transportation vehicles is collected. During the information exchange phase an involved transportation vehicle submits its so-called own driving hose to the involved transportation vehicles which includes decisive information from its own observations. Then an evaluation phase is started in which a transportation vehicle evaluates its own and the received driving hoses from the other involved transportation vehicles and determines a mutual safe region for the involved transportation vehicles to go for avoiding the collision.
In a disclosed embodiment, this mutual safe region is submitted to the involved transportation vehicles, which perform an operation of checking if the mutual safe region received from the other involved transportation vehicles fulfills their own safety requirements. If this is not the case, such transportation vehicle will change its state from collision avoidance state to collision minimization state and inform the other involved transportation vehicles about the change of state. This proposal allows the amount of information exchange about suggested trajectories to be subjectively reduced. There are far less messages with all their overhead to be exchanged, i.e., the overall performance of the communication system is subjectively increased. This is certainly a great benefit since with all the traffic on the roads the car-to-car communication could easily outperform the mobile communication system otherwise. If the safety requirements are fulfilled, the transportation vehicle selects a desired trajectory from the mutual safe region.
Since each transportation vehicle calculates the mutual safety region, it is not important anymore which exact trajectory the transportation vehicle actually selects for the next driving moment as long as this trajectory is within the safety region. The main problem with the exchange of unique trajectories is that they are discrete and if one is changed by any little value, it is already a new trajectory and other transportation vehicles have to be informed about it. That's one important property, why according to the proposal the safety region is taken into accounts as the most important parameter and the exchange of actual trajectories is an option behind this which may be utilized as an additional feature (nice to have, but not must have).
In other words the proposal allows finding a common global solution for the collision avoidance problem wherein a minimum number of messages need to be exchanged. This benefit stems from the fact that driving hoses will be exchanged which contain information about longitudinal and lateral position of the transportation vehicle over time and over the probability that the transportation vehicle reaches the point in the regarded time slot. Therefore, the whole of the possible trajectories is comprised in this driving hose such that it is somewhat guaranteed that the global minimum, i.e., the maneuver with safest driving conditions to avoid collision is included in the data set.
In summary, the proposed method comprises the following major benefits:
It is beneficial if the driving hose of a transportation vehicle is represented as a 4-dimensional tensor with information about longitudinal and lateral position of the transportation vehicle over time and over the probability that the transportation vehicles reaches the corresponding point in time.
For performing the evaluation of the driving hoses with a computer it is beneficial to calculate a correlation of the driving hoses received from the other involved transportation vehicles over space and time with the own driving hose. Then the global minimum in the correlation calculation result determines where the mutual safe region is located.
It is beneficial for the maneuver planning that each transportation vehicle selects a desired trajectory for itself out of the mutual safe region, and to submit the desired trajectory to the involved transportation vehicles for further checking. Thus the involved transportation vehicles can check again whether there is a risk of a collision in the planned maneuver.
The efficiency of the information exchange phase can be increased if the technique of linear network coding is applied when submitting the own driving hose. Even if a message with a driving hose could not be received due to an error prone channel, there is information present about the lost driving hose in the submissions from the other involved transportation vehicles such that in the evaluation phase the lost diving hose is not totally neglected. This increases subjectively the robustness of the information exchange phase communication.
Optionally, for linear network coding, the own driving hose will be linearly combined with the previously received driving hoses from the other involved transportation vehicles and the generation matrix for the linear combination is also included in the message submitting the own driving hose such that station receiving this message can easily decode the message.
As one example, the technique of random linear network coding RLNC may be used for linear network coding.
An efficient way to distribute the information of a driving hose is when the involved transportation vehicles broadcast the driving hose in the information exchange phase. Thus with one message only all the involved transportation vehicles will get the information without the need to send it multiple times.
It is beneficial that the minimum allowed distance between two transportation vehicles at a time in the future and/or the probability that two transportation vehicles end up at the same place at a time in the future are used as a safety requirement.
Moreover, it is beneficial that the transportation vehicle which detects the risk of a collision starts the information exchange phase by broadcasting a warning message to the surrounding transportation vehicles.
Further on, it is beneficial, that a timer for the information exchange phase is started in the transportation vehicle which detected the risk of a collision and wherein the warning message includes information about the timer for the information exchange phase. This way, all involved transportation vehicles know in which time they should react sending their own driving hose before the information exchange phase is closed by the alarming transportation vehicle.
Then it is beneficial for the implementation that also a timer for the evaluation phase is started and when this timer expires without having received a desired trajectory from one of the involved transportation vehicles that the timer setting transportation vehicle changes state from collision avoidance state to collision minimization state and informs the other involved transportation vehicles about the change of state.
If one of the involved transportation vehicles in the maneuver does not send any message within the time interval of the first timer, then this transportation vehicle is treated as non-cooperative by other the other involved transportation vehicles. But among the other transportation vehicles there can be still a cooperative collision avoidance maneuver carried out.
Then, if after correlation of all exchanged driving hoses no global solution is found which fulfills the “safety” criteria at least by one of the cooperating transportation vehicles, then such transportation vehicle changes its state to collision minimization and informs the other transportation vehicles about it.
Corresponding benefits are apparent for a control unit according to the proposal and a computer program according to the proposal as claimed.
The present description illustrates the principles of the present disclosure. It will thus be appreciated that those skilled in the art will be able to devise various arrangements that, although not explicitly described or shown herein, embody the principles of the disclosure.
All examples and conditional language recited herein are intended for educational purposes to aid the reader in understanding the principles of the disclosure and the concepts to furthering the art, and are to be construed as being without limitation to such specifically recited examples and conditions.
Moreover, all statements herein reciting principles, properties, and embodiments of the disclosure, as well as specific examples thereof, are intended to encompass both structural and functional equivalents thereof. Additionally, it is intended that such equivalents include both currently known equivalents as well as equivalents developed in the future, i.e., any elements developed that perform the same function, regardless of structure.
Thus, for example, it will be appreciated by those skilled in the art that the diagrams presented herein represent conceptual views of illustrative circuitry embodying the principles of the disclosure.
The functions of the various elements shown in the figures may be provided through the use of dedicated hardware as well as hardware capable of executing software in association with appropriate software. When provided by a processor, the functions may be provided by a single dedicated processor, by a single shared processor, or by a plurality of individual processors, some of which may be shared. Moreover, explicit use of the term “processor” or “controller” should not be construed to refer exclusively to hardware capable of executing software, and may implicitly include, without limitation, digital signal processor (DSP) hardware, read only memory (ROM) for storing software, random access memory (RAM), and nonvolatile storage.
Other hardware, conventional and/or custom, may also be included. Similarly, any switches shown in the figures are conceptual only. Their function may be carried out through the operation of program logic, through dedicated logic, through the interaction of program control and dedicated logic, or even manually, the particular technique being selectable by the implementer as more specifically understood from the context.
In the claims hereof, any element expressed as a way for performing a specified function is intended to encompass any way of performing that function including, for example, a) a combination of circuit elements that performs that function or b) software in any form, including, therefore, firmware, microcode or the like, combined with appropriate circuitry for executing that software to perform the function. The disclosure as defined by such claims resides in the fact that the functionalities provided by the various recited ways are combined and brought together in the manner which the claims call for. It is thus regarded that any way that can provide those functionalities are equivalent to those shown herein.
However, transportation vehicle communication, in particular, transportation vehicle direct communication, is necessary for this purpose. Various systems for transportation vehicle communication are developed. Examples include WLAN-based transportation vehicle communication, also called “ad-hoc domain”, and transportation vehicle communication in the field of mobile radio networks. In the case of mobile radio-based technology, however, the base station has to transmit the messages from transportation vehicle to transportation vehicle. This is the area where communication takes place in the so-called “Infrastructure Domain” For the future mobile radio generation, the transportation vehicle direct communication is also made possible. In LTE, according to Long Term Evolution, this option is called LTE-V; in the case of the 5G initiative, this option is called D2D.
The transportation vehicles in
Such mobile radio technologies are standardized and reference is made here to the corresponding specifications of mobile radio standards. As a modern example of a mobile radio standard, reference is made to the 3GPP initiative and the LTE standard (Long Term Evolution). Many of the related ETSI specifications are currently available in the version 14. The following is mentioned as an example: ETSI TS 136 213 V13.0.0 (2016-05); Evolved Universal Terrestrial Radio Access (E-UTRA); Physical layer procedures (3GPP TS 36.213 version 13.0.0 Release 13).
For the present disclosure, a sensor system is used which allows the measurement of distances in longitudinal and axial direction as well as the relative speed in the longitudinal and transverse directions.
The movement between the observer transportation vehicle 31 and the preceding transportation vehicle 32 is determined as a measured relative longitudinal and lateral distance with sx and sy, and vx and vy are used as the measured relative longitudinal and transverse speed components, as shown in
However, the modern transportation vehicle can also have further components such as video cameras 105. As a reversing camera or as a driver monitoring camera, as well as a LIDAR or RADAR device for the realization of a radar system or for implementing a distance warning or collision warning/avoidance device.
A navigation system 120, which is also installed in the area of the cockpit, is often distinguished from this. The route, which is displayed on a map, can of course also be displayed on the display in the cockpit. Other components, such as a hands-free system, may be present but are not shown in detail. Reference numeral 110 denotes an on-board unit. This on-board unit 110 corresponds to a communication module via which the transportation vehicle can receive and transmit mobile data. Typically, this is a mobile radio communication module, e.g., according to the LTE standard. All these devices are assigned to the infotainment area. They are therefore networked via a bus system 102 designed for the special needs of this device category. A high speed CAN bus is one example, which could be applied. As a further example, an Ethernet-Bus 108 is shown which connects only the two components driver assistance controller 171 and RADAR device 172. The Ethernet-Bus is a choice also for this communication bus 108 due to its higher bandwidth for data transport. In addition or as an alternative to the RADAR sensor a LIDAR sensor or a number of cameras and/or ultrasonic sensors could be applied for surrounding observation. One Ethernet-Bus adapted to the special needs of car communication is standardized in the IEEE 802.1Q specification. The ultrasonic sensor typically is used for a short distance observation, e.g., 3 to 5 m. The RADAR and LIDAR sensors could be used for scanning a range up to 250 m or 150 m and the cameras cover a range from 30 to 120 m.
For the purpose of transmitting the transportation vehicle-relevant sensor data via the communication interface 110 to another transportation vehicle or to a central computer, the gateway 140 is provided. This is connected to the different bus systems 102, 104 and 108. The gateway 140 is adapted to convert the data it receives via the Ethernet-Bus 108 to be converted into the transmission format of the infotainment CAN-bus 102 so that it can be distributed in the packets specified there. For the forwarding of this data to the outside, i.e., to another transportation vehicle or to central computer 50, the on-board unit 110 is equipped with the communication interface to receive these data packets and, in turn, to convert them into the transmission format of the correspondingly used mobile radio standard. As shown, the gateway 140 is connected as a central device to the busses 102, 104 as well as 108. It therefore takes all the necessary format conversions if data are to be exchanged between the different bus systems and if required.
To be exact, the 3D representation is not fully correct. More for maneuver planning, a 4D representation of the driving hose is used. One dimension corresponds to the longitudinal direction and the second dimension to the lateral direction. The third dimension corresponds to the time and the fourth dimension to the so-called ego-probability, i.e., the probability that the observer transportation vehicle will reach the corresponding point at the given time. Each cooperating transportation vehicle in the disclosed collision avoidance system will calculate this sort of 4D driving hose. Another way of interpreting the 4D driving hose is to view it as a trajectory space containing a great variety of possible trajectories the car may take.
Since the broadcast message will be communicated wirelessly in a broadcast channel of the mobile communication sidelink traffic in case of an LTE mobile communication system or in a WLAN channel for WLAN based car to car communication, the linear combination corresponding to linear network coding should definitely improve the robustness.
For sidelink traffic the broadcast messages will be submitted in the SL-PSBCH channel. The methodology how transportation vehicles access the medium for broadcast can be arbitrary.
When a UE (user equipment) joins a network, it goes through a network entry phase. The purpose of the network entry phase is for the current UE to transmit a Network Entry Packet (NEP), a one-time special transmission which is meant to inform all the other UEs within range of the current one that it is about to join the network. The slot for the transmission of the NEP is chosen using the Random Access Time-Division Multiple Access (RATDMA) protocol, which “is used when a station needs to allocate a slot, which has not been pre-announced”. After the UE has entered the network, for the sidelink traffic an STDMA (self-organized TDMA) MAC layer protocol has been adopted where TDMA is the well-known time division multiple access method where time slots will be assigned to participants. If for the considered scenario the RATDMA access is not reliable enough, one possible approach could be that when a transportation vehicle broadcasts its status, it points out who will broadcast next (until all transportation vehicles in the vicinity have broadcasted). This is possible if one bears in mind that under the considered scenario of cooperative or autonomous driving the transportation vehicles anyhow broadcast so-called Cooperative Awareness Messages CAM periodically such that they are aware which other transportation vehicles are in the vicinity.
After the deadline time has passed, no further trajectories spaces are being considered. Since all transportation vehicles influence each other on the road to some extent the exchanged trajectories spaces will also have certain correlation, which means that some level of information about one road participant will be already present in the trajectory space sent by another transportation vehicle. This additionally adds further redundancy, when some of the sent trajectory spaces were not received by some of the involved transportation vehicles.
After having received the 4D driving hoses from all involved cars, each car evaluates the situation and calculates its own safe region and selects from this region an evasive trajectory for its own car to dissolve the dangerous situation. For this, each transportation vehicle decodes the linearly combined 4D driving hoses. At this point, it is assumed that all transportation vehicles will have the same set of 4D spaces. The way this is done, is illustrated in
Thus each transportation vehicle correlates all received trajectory spaces with its own over space and time to find the mutual global safe region which fulfills certain pre-defined safety requirements. Since each transportation vehicle has the same set of trajectory spaces, the solution found at each transportation vehicle will fulfill the requirements of each transportation vehicle. If the found solution at any transportation vehicle does not fulfill the pre-defined safety requirements, the transportation vehicle informs all other transportation vehicles, that it is not able to successfully complete the maneuver and that the maneuver of collision minimization has to be started instead of collision avoidance.
In the case if some messages were lost and the available redundancy and RLNC encoding was not sufficient for some transportation vehicle to receive some of the trajectory spaces, then this transportation vehicle has to find the common correlated optima based on limited available information. After this, the found solution is being compared with the final post-correlation exchanged solutions.
The found mutual safe region 31S, 32S, 33S based on correlated optima's is being then additionally broadcasted to ensure an additional cross-check whether the found solution is suitable for all transportation vehicles in the vicinity.
In the case if some messages were lost and the available redundancy and encoding was not sufficient for some transportation vehicle to receive some of the trajectory spaces, then this transportation vehicle has to find the common correlated optima based on limited available information. But since, the found solution is being compared with the final post-correlation exchanged solutions being broadcasted, no additional action is required if the solution found from limited information is still within the tolerable region. Otherwise, the transportation vehicle informs its neighbors that it is not able to successfully complete the maneuver and that the maneuver of collision minimization has to be started instead of collision avoidance.
It is to be understood that the proposed method and apparatus may be implemented in various forms of hardware, software, firmware, special purpose processors, or a combination thereof. Special purpose processors may include application specific integrated circuits (ASICs), reduced instruction set computers (RISCs) and/or field programmable gate arrays (FPGAs). Optionally, the proposed method and apparatus is implemented as a combination of hardware and software. Moreover, the software may be implemented as an application program tangibly embodied on a program storage device. The application program may be uploaded to, and executed by, a machine comprising any suitable architecture. Optionally, the machine is implemented on a computer platform having hardware such as one or more central processing units (CPU), a random access memory (RAM), and input/output (I/O) interface(s). The computer platform also includes an operating system and microinstruction code. The various processes and functions described herein may either be part of the microinstruction code or part of the application program (or a combination thereof), which is executed via the operating system. In addition, various other peripheral devices may be connected to the computer platform such as an additional data storage device and a printing device.
It should be understood that the elements shown in the figures may be implemented in various forms of hardware, software or combinations thereof. Optionally, these elements are implemented in a combination of hardware and software on one or more appropriately programmed general-purpose devices, which may include a processor, memory and input/output interfaces. Herein, the phrase “coupled” is defined to mean directly connected to or indirectly connected with through one or more intermediate components. Such intermediate components may include both hardware and software based components.
It is to be further understood that, because some of the constituent system components and method operations depicted in the accompanying figures may be implemented in software, the actual connections between the system components (or the process operations) may differ depending upon the manner in which the proposed method and apparatus is programmed Given the teachings herein, one of ordinary skill in the related art will be able to contemplate these and similar implementations or configurations of the proposed method and apparatus.
The disclosure is not restricted to the exemplary embodiments described here. There is scope for many different adaptations and developments which are also considered to belong to the disclosure.
The application possibilities for the described methods are not limited to use in the transportation vehicle. The approach for solving the problem of collision avoidance may also be applied in other areas. Examples include aircrafts and drones, as well as mobile robotics in a wide range of areas.
| Number | Date | Country | Kind |
|---|---|---|---|
| 17191248.8 | Sep 2017 | EP | regional |
