DE102018204006A1 - Method and system for the secure collision avoidance of vehicles monitored on the infrastructure side - Google Patents

Abstract

Disclosed is a method for controlling at least one vehicle on an infrastructure, wherein: an occupancy map based on measurement data of an infrastructure-side sensor system is created based on the created occupancy map by a control unit of at least one vehicle collision-free passable area is determined, starting from a through the sensor system determined position of the at least one vehicle and the determined collision-free passable area is calculated driving permission for the at least one vehicle by the control unit and transmitted via a communication link to the at least one vehicle, based on the driving release the collision-free passable area is traveled by the at least one vehicle, Deviations of the at least one vehicle from the driving release by an in-vehicle path control are checked and at a detected by the path control deviation of the at least one vehicle ugs of the drive release, the at least one vehicle is placed in a safe state. Furthermore, a system is disclosed.

Description

The invention relates to a method for controlling at least one vehicle on an infrastructure, and to a system for infrastructure-guided driving of at least one vehicle.

State of the art

In a so-called AVP (Automated Valet Parking) -enabled parking garage, information is transferred from the parking garage infrastructure to the vehicle to be parked automatically. In a well-known AVP implementation, the infrastructure handles much of the system's intelligence. The infrastructure monitors the vehicle to be parked and the surroundings of the vehicle to be parked by sensors installed in the parking garage. Such sensors may be cameras, for example.

The sensor-based ascertained information is processed outside the vehicle and control signals are sent via a wireless communication link to the vehicle to be parked. The vehicles to be parked should preferably have the smallest possible number of additional components for executing the control commands and for implementing the parking operation. Thus, the additional cost of producing AVP-enabled vehicles or related conversions may be low.

The vehicle can be through an appropriately equipped infrastructure in a car park without its own sensors on the road and is constantly on current information from the parking garage infrastructure dependent. Due to the lacking or low number of in-vehicle security systems, high security requirements have to be imposed on such AVP systems.

Disclosure of the invention

The object underlying the invention can be seen to propose a method and a system for safe collision avoidance of an automated moving vehicle.

This object is achieved by means of the subject matter of the independent claims. Advantageous embodiments of the invention are the subject of each dependent subclaims.

According to one aspect of the invention, a method for controlling at least one vehicle on an infrastructure is provided. In one step, an occupancy map is created based on measurement data of an infrastructure-side sensor system.

Based on the created occupancy card, a control unit determines a range which can be traveled by a collision-free vehicle by at least one vehicle.

Subsequently, starting from a position of the at least one vehicle determined by the sensor system and the ascertained collision-free passable area, a driving release for the at least one vehicle is calculated by the control unit and transmitted to the at least one vehicle via a communication link. The release can be configured, for example, in the form of a control instruction with defined actuator parameters of the vehicle. The defined actuator parameters can be adapted to a safe and permissible operating range for maintaining the collision-free passable range.

Based on the transmitted release, the collision-free passable area is traveled by the at least one vehicle. During the compliance of the control instruction or the drive release by the vehicle, deviations of the at least one vehicle from the drive release are checked by an in-vehicle path control. In the case of a deviation of the at least one vehicle from the driving release detected by the path control, the at least one vehicle is put into a safe state. The path control may alternatively or additionally be designed as an internal monitoring function. The internal monitoring function of the vehicle compares the current actuator parameters with the actuator parameters transmitted with the vehicle release. In the event of a deviation of a parameter of the at least one vehicle detected by the monitoring function, the at least one vehicle is put into a safe state.

The method can ensure that, for example, an AVP vehicle does not cause any collisions with either static or moving objects and obstacles while this vehicle is driverlessly controlled by an infrastructure. The infrastructure may be, for example, a parking lot, a pick-up zone, a drop-off zone, a charging station and the like.

For autonomous or semi-autonomous vehicles, safety is a relevant topic. Since the vehicles to be parked an AVP system driverless and drive with a minimum of sensors through a parking garage, such vehicles may have an accident and injury risk. By integrated into the at least one vehicle monitoring function, an efficient monitoring and control device can be used, which ensures that the at least one vehicle to be parked in terms of increased Safety requirements is put in a timely manner in a safe state. As a result, especially injuries to persons can be excluded or at least minimized. The vehicle to be parked can alternatively or additionally also be controlled or directed without a parking intent to a destination. For example, the vehicle may be a transporter, which can be guided to different target positions by the infrastructure.

The method enables an infrastructure-side or parking-side generation of occupancy maps. The occupancy maps may be, for example, grid maps with marked obstacles. The unmarked grids can potentially be driven without collision. The obstacles and the occupancy map are determined based on measurement data of the infrastructure-side sensor system and can preferably be valid only for a defined period of time.

The sensor system may consist of a variety of sensors, such as cameras, radar sensors, LIDAR sensors, light barriers, ultrasonic sensors and the like. The sensors are coupled directly or via interfaces to the infrastructure-side control unit. The sensor system can also have its own control unit for generating occupancy maps and for predicting future trajectories of dynamic objects. Alternatively, the measurement data determined by the sensor system can be completely evaluated by the control unit.

On the basis of the "free" or not occupied by obstacles grid driving permits for different vehicles can be generated and then transmitted wirelessly to the respective vehicles.

In the respective vehicles, the drive release can be interpreted by the vehicle-side path control and converted into a vehicle movement by direct or indirect actuation of actuators. This process from the creation of the occupancy card to the generation and transmission of the control instructions to the vehicles can be repeated at defined time intervals.

The driving permits can be sent via a wireless communication link to at least one vehicle. The wireless communication link may preferably be encrypted, so that a risk of third-party access is reduced. The wireless communication link may be based on a WLAN, UMTS, LTE, GSM and the like wireless communication standard.

Due to the simple structure of the chain of action of the method, the safety-relevant components can be mathematically detected and practically ensured in a simple manner, since the method behaves deterministically. All it takes is a simple on-board monitoring logic that can be easily implemented in an embedded controller. The actual path control (automated vehicle guidance) of the at least one vehicle can be implemented and applied decoupled from the monitoring logic. If a deviation of the vehicle from the release determined by the monitoring logic, the vehicle is immediately placed in the safe state. Such deviations may be, for example, steering angles or speeds exceeding the limits of release. Furthermore, a deviation may be that a temporal validity of the control command expires before a new control command has been sent from the control unit to the at least one vehicle. Furthermore, the components of the automated vehicle management control chain can be decoupled both in the infrastructure or in the parking garage and in the vehicles or can be independent of the security-relevant monitoring chain. As a result, the method can be configured independently of an architecture of the vehicle movement control of different vehicle manufacturers.

The complex calculations for generating the drive release including reliable actuator parameters remain on the parking garage side and can preferably be carried out on a control unit designed as a server with a high computing power. The signals of the drive release or the control instructions directly represent the physical degrees of freedom of the at least one vehicle. The potentially passable area is thus clearly defined at all times.

The at least one vehicle may be a passenger car, a truck, a forklift, transport robots and the like. The method is not limited to parking garages, parking lots or parking areas, but can be applied analogously in automated halls or premises, such as depots.

According to one embodiment of the method, an emergency braking maneuver is performed by the at least one vehicle during the transfer of the at least one vehicle to a safe state. A stationary vehicle can be defined as a safe state of an AVP system. The state and the transfer to the safe state can be referred to as a so-called "safety stop".

Preferably, the triggering of the safe state can be done by the path control. The path control can be embodied as a separate hardware component or software component integrated in a control unit of the vehicle. When the vehicle is released, the vehicle can receive limit values or operating frames which must be observed by the vehicle in order to remain within the range that can not be traveled over. Due to the continuous comparison of the actual operations carried out by the vehicle or actuator movements with the operation limits of the drive release, a full braking can be initiated when the value exceeds or falls below a value, such as a steering angle or a speed.

The driving task of the at least one vehicle and the vehicle guidance or the driving release are preferably functionally and / or spatially separated from one another and configured independently of one another. The at least one vehicle reacts to the infrastructure-generated driving license or the driving release and internally assumes the path control chain. This would allow vehicles to be safely and accurately controlled by the infrastructure without permission for highly automated driving due to precise infrastructural sensing.

According to a further embodiment of the method, the drive release has a steering angle range, a speed to be maintained by the at least one vehicle, and a time specification. As a result, the drive release can have a small amount of information and can be transmitted quickly and reliably with a low bandwidth. For example, a maximum permitted steering angle in one direction, a maximum permitted steering angle in an opposite direction, a speed, and a time specification or a maximum validity of the control command in the form of a drive release to the at least one vehicle can be transmitted. The maximum steering angle can be configured in the form of allowed areas of curvature in the generation of the driving release based on the occupancy map. The steering angle range and the corresponding drive release can preferably take into account both a front axle steering and a rear axle steering.

According to a further embodiment of the method, the drive release has a gear path range or a rotation range of a steering motor and a wheel rotation speed or a travel distance. Alternatively, the drive release based on physical vehicle models, which are stored in the control unit, influence or define the direct action variables of the actuators or odometer. As a result, the vehicle trajectory can be controlled more precisely or remotely controlled by the control unit.

According to a further embodiment of the method, the drive release has a curvature region. In this way, the release area can be abstractly and clearly defined from the point of view of the infrastructure in order to define the permitted steering angle position limits. The vehicle can thereby convert the maximum and minimum curvature range in-vehicle into corresponding maximum and minimum steering angles, as a result of which no model of the specific vehicle steering mechanism must be known on the infrastructure side.

According to a further embodiment of the method, the area which can be driven by the at least one vehicle is calculated on the basis of a vehicle model stored in the control unit. The vehicle model may be, for example, a simplified physical vehicle model or a vehicle model of the vehicle uniquely identified by the sensor system. Based on the vehicle model, the operation limits or actuator limits of the drive release can be calculated in such a way that the vehicle can drive without collision based on the drive release.

In accordance with a further embodiment of the method, the travel enable and / or the area which can be traveled without collision have a time-limited validity. On the basis of the occupancy map or occupancy grid, the permitted, collision-free range of movement of the vehicle is calculated within a defined time range. The occupancy card can be a simple or an extended occupancy card with consideration of dynamic objects. The time range can be one or more seconds or less than one second. If a range which can be traveled by the vehicle without collision has been found by the algorithm of the sensor system or the control unit, a time-limited driving release is issued based on this and the vehicle is transmitted wirelessly. Otherwise, no driving clearance will be granted. The vehicle initiates full braking in the absence of drive release and remains in the safe state until released.

In accordance with a further embodiment of the method, a new occupancy card is created by the infrastructure-side sensor system before a time-limited validity expires, and a collision-free passable area based on the new occupancy card is determined by the control unit. Depending on the repetition frequency of the process, new updated occupancy maps with correspondingly adapted travel approvals on the infrastructure side are produced at defined time intervals created and sent to the vehicle in the form of driving permits. As a result, the at least one vehicle can be guided gradually to a destination of the infrastructure. Each step sets a time segment or a defined limited section in a defined direction with a defined possible steering deviation.

According to a further embodiment of the method, the at least one vehicle is put into the safe state after a time expiration of validity. In this way, it is possible to prevent damage being caused by the vehicle in the event of a faulty communication connection between the at least one vehicle and the infrastructure or the parking space. Without current driving permits, the vehicle remains in safe condition.

According to a further embodiment of the method, movements of dynamic obstacles by the sensor system are taken into account by estimating or performing a simulation in the preparation of the occupancy map. As a result, static obstacles or objects as well as dynamic and moving objects, such as other vehicles or people, can be taken into account when creating occupancy maps.

The occupancy card contains only statically recorded objects at time t0 when being transmitted to the parking garage server or to the control unit. The possible range of movement of dynamic obstacles within a time period t, such as, for example, 1 s, can preferably be calculated and entered into the occupancy map or taken into account there during the determination of a collision-free passable area. Dynamic obstacles can be enlarged, for example, in a defined direction or in several directions and thus further cells or grids of the occupancy map can be marked as "occupied".

According to a further embodiment of the method, the collision-free passable region is delimited on at least two sides by identically or unequally shaped radii of curvature or clothoids. Depending on a speed of the vehicle to be controlled, the radii of curvature can be selected larger or smaller starting from a vehicle center or a vehicle steering axle. On a straight line, the radii of curvature may be oriented away from the vehicle to define maximum steering angles. For driving on curves, the radii of curvature can be directed in a same direction and preferably according to the curve. The use of clothoids as time-limited valid collision-free passable areas or so-called "safety zone", these areas can be implemented physically precise.

According to a further embodiment of the method, the shape of the radii of curvature or clothoids of the collision-free passable region is set as a function of a speed or a distance of the at least one vehicle covered during the validity of the control command. It can thus be determined on a route and a vehicle speed optimally adapted collision free passable areas and generated in the form of driving permits.

Alternatively or additionally, the driving permits may be designed as a stop request instead of as drive releases. In the process, corrections and brake commands can be sent to the at least one vehicle on the infrastructure side. Furthermore, a plurality of control instructions or releases can be sent to the at least one vehicle, whereby defined control operations, such as steering movements or braking operations are performed more precisely and with continuous gradation.

According to a further aspect of the invention, a system is provided for infrastructure-assisted guidance of at least one vehicle. The system includes an infrastructure having a sensor system for monitoring and measuring an infrastructure environment and creating occupancy maps of the infrastructure environment. A control unit of the system is used for vehicle-specific calculation of collision-free passable areas and for creating vehicle-specific driving permits. The system includes at least one communication device for establishing a wireless communication link to at least one vehicle and for transmitting control instructions to the at least one vehicle. Furthermore, the system comprises at least one vehicle with a path control for receiving and maintaining driving permits.

In this way, a system for guiding at least one vehicle along a freely passable area to at least one destination can be implemented, which performs all the computing-intensive planning tasks and calculation tasks on the infrastructure side and provides the vehicle with a minimum amount of necessary information. This information can be checked quickly and easily by a vehicle-side path control, whereby the vehicle can be put into a safe state in the event of a negative test of the driving releases.

• 1 eine schematische Darstellung eines Parkplatzes mit einem System gemäß einer Ausführungsform der Erfindung,
• 2 schematische Darstellungen unterschiedlicher Safety Zonen gemäß Ausführungsformen der Erfindung und
• 3 eine schematische Darstellung des Systems gemäß einer Ausführungsform der Erfindung.

In the following, preferred exemplary embodiments of the invention are explained in greater detail on the basis of greatly simplified schematic representations. Show here

• 1 a schematic representation of a parking lot with a system according to an embodiment of the invention,
• 2 schematic representations of different safety zones according to embodiments of the invention and
• 3 a schematic representation of the system according to an embodiment of the invention.

In the figures, the same constructive elements each have the same reference numerals.

In the 1 is a schematic representation of a parking lot 1 or a level of a parking garage 1 as an infrastructure 1 with a system 2 according to an embodiment of the invention. In particular, the infrastructure 1 with an occupancy card B superimposed so that the inventive method can be illustrated.

It is a trajectory T of a vehicle to be parked 4 shown. Furthermore, there is a safety zone 6 or a collision-free passable area 6 representing which the vehicle 4 leads around a bend. The walls or the boundaries of the parking lot 1 and the stationary vehicles pose static obstacles 8th and are accordingly as occupied grid 8th . 10 in the occupancy card B noted. Behind the curve was from the in 3 quantified sensor system a dynamic obstacle 10 in the form of a parked vehicle and the expected trajectory of the dynamic obstacle 10 estimated by the sensor system. Based on this, the occupancy card B became more occupied with grids 8th . 10 Mistake.

The 2a to 2c show schematic representations of different safety zones 6 according to embodiments of the invention. The respective safety zones 6 or the collision-free passable areas 6 be starting from a point on the vehicle 4 by two radii of curvature 11 . 12 laterally limited. The two radii of curvature 11 . 12 define the maximum steering angle of the vehicle 4 , which were transmitted via driving permits. It is illustrated that the radii of curvature 11 . 12 with increasing speed or increasing time validity of the drive release greater or the collision-free passable area 6 gets longer. The walls are exemplary here as static obstacles 8th shown. The 2a shows the calculated safety zones 6 during a regular straight ahead of the vehicle 4 with a distance of 2.77 m, 5 m and 15 m during the validity period of the driving clearance. The 2 B clarifies calculated safety zones 6 as part of a course correction and the 2c calculated possible safety zones 6 during a regular cornering.

In the 3 is a schematic representation of the system 2 illustrated according to an embodiment of the invention. In particular, the safety-relevant components are shown. The system 2 has an infrastructure 1 with a sensor system 14 and one with the sensor system 14 data-connected control unit 16 on.

The sensor system 14 can, for example, relevant parts of the infrastructure 1 , such as drivable areas and routes, observe and survey. Preferably, the sensor system 14 a variety of distributed sensors, such as radar sensors, LIDAR sensors or cameras on. Thus, the sensor system 14 the at least one vehicle 4 , static obstacles 8th and dynamic obstacles 10 detect and as part of the generation of the occupancy map B consider.

The created occupancy card B is sent to the control unit 16 transfer. The control unit 16 can be based on the unoccupied grid of the occupancy map B Create a driving release and according to the embodiment via a secure WLAN connection as a wireless communication link 18 transfer.

In the vehicle 4 becomes the control instruction from a vehicle control 20 interpreted and translated into actuator controls. The limits defined in the drive enable, which are on collision-free passable areas 6 are continuously controlled. Becomes the defined safety zone 6 from the vehicle 4 left or exceeded by the vehicle control 20 an emergency stop which the vehicle 4 put in a safe state.

Claims (13)

Method for controlling at least one vehicle (4) on an infrastructure (1), wherein: - an occupancy map (B) is created based on measurement data of an infrastructure-side sensor system (14), - based on the created occupancy map (B) by a control unit (16 ) is determined by at least one vehicle (4) collision-free passable area (6), - based on a detected by the sensor system (14) position of the at least one vehicle (4) and the determined collision free passable area (6) a drive release for the at least one vehicle (4) calculated by the control unit (16) and is transmitted to the at least one vehicle (4) via a communication link (18), - the at least one vehicle (4) is driven on the basis of the driving release of the collision-free passable area (6), - deviations of the at least one vehicle (4) from the driving release by a vehicle-internal vehicle control (20) are checked and - at a detected by the vehicle controller (20) deviation of the at least one vehicle (4) from the driving release, the at least one vehicle (4) is placed in a safe state. Method according to Claim 1 wherein during the transfer of the at least one vehicle (4) to a safe state an emergency braking maneuver is performed by the at least one vehicle (4). Method according to Claim 1 or 2 wherein the drive release comprises a steering angle range, a speed to be maintained by the at least one vehicle (4), and a time indication. Method according to Claim 1 or 2 wherein the drive release has a gear path range or a rotation range of a steering motor and a wheel revolution number or a travel distance. Method according to one of Claims 1 to 3 wherein the drive release has a curvature area. Method according to one of Claims 1 to 5 wherein the at least one vehicle (4) collision-free passable area (6) is calculated based on a stored in the control unit (16) vehicle model. Method according to one of Claims 1 to 6 , wherein the drive release and / or the collision-free passable area (6) have a time-limited validity. Method according to Claim 7 , wherein a new occupancy map (B) is created by the infrastructure-side sensor system (14) and a determination on the new occupancy map (B) based collision free passable area (6) by the control unit (16) is determined before a expiry of the temporary validity. Method according to Claim 7 or 8th , wherein the at least one vehicle (4) is put into the safe state after a time expiration of the validity. Method according to one of Claims 1 to 9 wherein movements of dynamic obstacles (10) are taken into account by the sensor system (14) by estimating or performing a simulation in the preparation of the occupancy map (B). Method according to one of Claims 1 to 10 in which the collision-free passable region (6) is delimited on at least two sides (11, 12) by identically or unequally shaped radii of curvature (11, 12) or clothoids. Method according to Claim 11 in that the shape of the radii of curvature (11, 12) or clothoids of the collision-free passable area (6) is set as a function of a speed or a distance traveled by the at least one vehicle (4) during the validity of the control command. System (2) for infrastructure-guided guiding at least one vehicle (4) and for carrying out the method according to one of the preceding claims, comprising An infrastructure (1) having a sensor system (14) for observing and measuring an infrastructure environment and for creating occupancy maps (B) of the infrastructure environment, a control unit (16) for vehicle-specific calculation of collision-free passable areas (6) and for creating vehicle-specific driving permits, - At least one communication device for establishing a wireless communication link (18) to at least one vehicle (4) and for transmitting Fahrfreigaben to the at least one vehicle (4) and - At least one vehicle (4) with a vehicle controller (20) for receiving and completing driving permits.

Images