Methods and systems for radar localization in autonomous vehicles

Abstract

Localization procedures, including: receiving sensor data relating to an environment of a vehicle (10), the sensor data including a plurality of sensor returns associated with objects (511, 512, 521, 522, 530, 531) in the environment, each of the sensor returns having a plurality of corresponding attributes; creating a first plurality of sensor data groups, each of which includes a self-consistent subset of the plurality of sensor returns based on their corresponding attributes; defining a first set of features for each of the first plurality of sensor data groups, each feature being based on at least one of the corresponding attributes and each having an associated feature location, and the first set of features including a histogram of one of the corresponding attributes; determining a feature correlation between the first set of features and a second, previously determined set of features with a processor (44); and estimating a position of the vehicle (10) based on the feature correlation.

Description

TECHNICAL FIELD

[0001] The present invention relates generally to autonomous vehicles and, more particularly, to systems and methods for determining the position of an autonomous vehicle relative to its environment.

[0002] For general background information, please refer to the article by Erik WARD and John FOLKESSON entitled “Vehicle localization with low cost radar sensors” (Published in: Proceedings / 2016 IEEE intelligent vehicles symposium, 19-22 June 2016, 7 pp. - ISBN 978- 1-5090-1822-2) and to DE 10 2015 100 134 A1.

[0003] For further information, please refer to the WIKIPEDIA article "RANSAC algorithm" as of August 4, 2017 (URL: https: / / de.wikipedia.org / w / index.php?title=RANSAC-Algorithmus&oldid=167873865). BACKGROUND

[0004] An autonomous vehicle is a vehicle capable of sensing its environment and navigating with little or no user input. This is achieved through the use of sensors such as radar, lidar, image sensors, and the like. Autonomous vehicles also utilize information from global positioning systems (GPS), navigation systems, vehicle-to-vehicle communications, vehicle infrastructure technologies, and / or wired systems to navigate the vehicle.

[0005] While significant progress has been made in autonomous vehicles in recent years, these vehicles could still be improved in several aspects. For example, localization—that is, determining a vehicle's geographical position—is often achieved using a combination of global positioning system (GPS) data and direct observation of structures and objects in the vehicle's surroundings using optical, radar, and lidar sensor data. Such localization can be difficult in certain situations, such as atmospheric conditions, compromising the effectiveness of the sensors used for localization.

[0006] Accordingly, it is desirable to provide systems and methods for localization in autonomous vehicles. Furthermore, other desirable functions and features of the present invention will become apparent from the following detailed description and the appended claims, taken in conjunction with the accompanying drawings, as well as the foregoing technical field and background. SUMMARY

[0007] According to the invention, a method for controlling a first vehicle is presented, which is characterized by the features of claim 1.

[0008] In one embodiment, the plurality of corresponding attributes includes at least one of Doppler shift, return power, and neighborhood similarity.

[0009] In one embodiment, the sensor data includes radar data.

[0010] In one embodiment, the first set of features is a convex hull of the histogram.

[0011] In one embodiment, the first set of features includes a summary statistic of one of the corresponding attributes.

[0012] In one embodiment, determining the feature correlation includes performing an outlier removal process with respect to the first and second sets of features.

[0013] In one embodiment, the outlier removal method is a random sample consensus (RANSAC) method.

[0014] In one embodiment, each of the sensor data groups is assigned to a dynamic object, a statically moving object, or a statically non-moving object, and the feature correlation is determined based only on the sensor data groups assigned to statically non-moving objects.

[0015] Furthermore, according to the invention, a system for controlling a vehicle is presented, which is characterized by the features of claim 5.

[0016] In one embodiment, the plurality of corresponding attributes includes at least one of Doppler shift, return power, and neighborhood similarity; wherein the sensor data is at least one of radar data and lidar data.

[0017] In one embodiment, the first set of features includes a summary statistic of one of the corresponding attributes.

[0018] In one embodiment, the feature correlation module performs a method for removing outliers with respect to the first and second sets of features.

[0019] In one embodiment, the outlier removal method is a random sample consensus (RANSAC) method.

[0020] In one embodiment, the feature determination module classifies each of the sensor data groups as associated with a dynamic object, a statically moving object, or a statically non-moving object, and the feature correlation module determines the feature correlation based only on the sensor data groups associated with statically non-moving objects.

[0021] We also describe an autonomous vehicle. The autonomous vehicle includes at least one sensor that provides sensor data relating to an environment of the autonomous vehicle, wherein the sensor data includes a plurality of sensor returns associated with the objects in the environment, wherein the respective sensor returns have a plurality of corresponding attributes.The autonomous vehicle further includes a controller that, through a processor: receives the sensor data; creates a first plurality of sensor data groups, each including a self-consistent subset of the plurality of sensor data returns based on their corresponding attributes; determines, for each of the first plurality of sensor data groups, a first set of features, each feature based on at least one of the corresponding attributes and each having an associated feature position; determines, with a processor, a feature correlation between the first set of features and a second predetermined set of features; and estimates a position of the vehicle based on the feature correlation.

[0022] In one embodiment, the controller performs outlier removal via a random sample consensus (RANSAC) method.

[0023] In one embodiment, the plurality of corresponding attributes includes at least one of Doppler shift, return power, and neighborhood similarity; wherein the sensor data is at least one of radar data and lidar data.

[0024] In one embodiment, the first set of features includes one of a histogram or a summary statistic of one of the corresponding attributes. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The exemplary embodiments are described below in conjunction with the following drawings, wherein like reference numerals denote like elements, and wherein: Fig. 1 is a functional block diagram illustrating an autonomous vehicle with a positioning system according to various embodiments; Fig. 2 is a functional block diagram illustrating a transport system with one or more autonomous vehicles from Fig. 1 according to various embodiments; Fig. 3 is a functional block diagram illustrating an autonomous driving system (ADS) in conjunction with an autonomous vehicle according to various embodiments; Fig. 4 is a conceptual top view of an exemplary sensor configuration for an autonomous vehicle according to various embodiments; Fig. 5 illustrates an autonomous vehicle traveling along a roadway according to an exemplary embodiment; Fig. Figure 6 is a top view of exemplary sensor returns associated with the scenario of Fig. 5; The Fig. 7 and Fig. 8 show selected sensor returns from Fig. 6 divided into different network configurations; Fig. 9 is a data flow diagram illustrating an autonomous vehicle positioning system according to various embodiments; Fig. 10 is a data flow diagram illustrating a classifier according to various embodiments; Fig. 11 illustrates exemplary sensor returns collected by an autonomous vehicle while traveling along an exemplary roadway according to various embodiments. Fig. 12 is a flowchart illustrating a control method for controlling the autonomous vehicle according to various embodiments. DETAILED DESCRIPTION

[0026] The following detailed description is provided for illustrative purposes only. Furthermore, there is no intention to be bound by any theory, expressed or implied, in the foregoing technical field, background, brief summary, or the following detailed description. As used herein, the term "module" refers to all hardware, software, firmware products, electronic control components, processing logic, and / or processor devices, individually or in any combination, including, but not limited to, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), an electronic circuit, a processor (shared, dedicated, or group processor), and memory that executes one or more software or firmware programs, combinational logic circuitry, and / or other suitable components that provide the described functionality.

[0027] Embodiments of the present invention may be described herein as functional and / or logical block components and various processing steps. It should be noted that such block components may be constructed from any number of hardware, software, and / or firmware components configured to perform the required functions. For example, an embodiment of the present invention of a system or component may employ various integrated circuit components, such as memory elements, digital signal processing elements, logic elements, look-up tables, or the like, capable of performing multiple functions under the control of one or more microprocessors or other control devices.Furthermore, those skilled in the art will recognize that the exemplary embodiments of the present invention may be used in connection with any number of systems, and that the system described herein represents merely one exemplary embodiment of the present invention.

[0028] For the sake of brevity, conventional techniques associated with signal processing, data transmission, signaling, control, machine learning models, radar, lidar, image analysis, and other functional aspects of the systems (and the individual operating components of the systems) cannot be described in detail herein. Furthermore, the connecting lines depicted in the various figures are intended to represent exemplary functional relationships and / or physical connections between the various elements. It should be noted that many alternative or additional functional relationships or physical connections may be present in an embodiment of the present invention.

[0029] With reference to Fig. 1, a positioning system, generally depicted at 100, is associated with a vehicle 10 according to various embodiments. In general, the positioning system (or simply "system") 100 enables a positioning method in which sensor data sets are created using self-consistent subsets of sensor returns based on their corresponding attributes (such as return power, Doppler shift, neighborhood similarity, and the like). Features are specified for the corresponding attributes and each have an associated feature location, and a processor determines a feature correlation between the first set of features and a second predefined set of features.

[0030] As in Fig. 1, the vehicle 10 generally includes a chassis 12, a body 14, front wheels 16, and rear wheels 18. The body 14 is disposed on the chassis 12 and substantially encloses the other components of the vehicle 10. The body 14 and the chassis 12 may together form a frame. The wheels 16-18 are each pivotally connected to the chassis 12 near a respective corner of the body 14.

[0031] In various embodiments, the vehicle 10 is an autonomous vehicle, and the positioning system 100 is integrated into the autonomous vehicle 10 (hereinafter referred to as the autonomous vehicle 10). The autonomous vehicle 10 is, for example, a vehicle that is automatically controlled to transport passengers from one location to another. The vehicle 10 is depicted as a passenger car in the illustrated embodiment, but it should be noted that any other vehicle, including motorcycles, trucks, sport utility vehicles (SUVs), recreational vehicles (RVs), vessels, aircraft, etc., may be used.

[0032] In an exemplary embodiment, the autonomous vehicle 10 corresponds to a level four or level five automation system according to the Society of Automotive Engineers (SAE) standard taxonomy of automated driving levels, "J3016." Using this terminology, a level four system indicates "high automation," referring to a driving mode in which the automated driving system assumes all aspects of the dynamic driving task, even if a human driver does not adequately respond to a request for intervention. A level five system, on the other hand, exhibits "full automation," referring to a driving mode in which the automated driving system performs all aspects of the dynamic driving task under all road and environmental conditions that a human driver can handle.It should be understood that the embodiments according to the present subject matter are not limited to any particular taxonomy or heading of automation categories. Furthermore, systems according to the present embodiment may be used in connection with any vehicle in which the present subject matter can be implemented, regardless of its autonomy.

[0033] As illustrated, the autonomous vehicle 10 generally includes a propulsion system 20, a transmission system 22, a steering system 24, a braking system 26, a sensor system 28, an actuator system 30, at least one data storage device 32, at least one controller 34, and a communication system 36. The propulsion system 20, in various embodiments, may include an internal combustion engine, an electric machine such as a traction motor, and / or a fuel cell propulsion system. The transmission system 22 is configured to transfer power from the propulsion system 20 to the vehicle wheels 16 and 18 according to selectable gear ratios. According to various embodiments, the transmission system 22 may include a stepped-ratio automatic transmission, a continuously variable transmission, or other suitable transmission.

[0034] The braking system 26 is configured to provide braking torque to the vehicle wheels 16 and 18. In various embodiments, the braking system 26 may include friction brakes, bake-by-wire, a regenerative braking system such as an electric machine, and / or other suitable braking systems.

[0035] The steering system 24 influences a position of the vehicle wheels 16 and / or 18. While shown in some embodiments for illustrative purposes as a steering wheel 25, the steering system 24 may not include a steering wheel.

[0036] The sensor system 28 includes one or more sensor devices 40a-40n that can detect observable conditions of the external environment and / or the internal environment of the autonomous vehicle 10 (e.g., the condition of one or more occupants). The sensor devices 40a-40n may include radars (e.g., long-range, medium-range, short-range), lidars, global positioning systems, optical cameras (e.g., forward-looking, 360-degree, rear-looking, side-looking, stereo, etc.), thermal imaging cameras (e.g., infrared), ultrasonic sensors, velocity measurement sensors (e.g., encoders), an inertial measurement unit (IMU) (IMUs), and / or other sensors that can be used in connection with systems and methods according to the present subject matter.

[0037] The actuator system 30 includes one or more actuator devices 42a-42n that control one or more vehicle features, such as the drive system 20, the transmission system 22, the steering system 24, and the braking system 26. In various embodiments, the autonomous vehicle 10 may also include vehicle interior and / or exterior features not included in Fig. 1, such as various doors, trunk and cabin features, such as air, music, lighting, touchscreen display components (such as those used in conjunction with navigation systems) and the like.

[0038] The data storage device 32 stores data for use in automatically controlling the autonomous vehicle 10. In various embodiments, the data storage device 32 stores defined maps of the navigable environment. In various embodiments, the defined maps are predefined and retrieved from a remote system (in further detail with respect to Fig. 2). For example, the defined maps may be assembled by the remote system and communicated to the autonomous vehicle 10 (wirelessly and / or wired) and stored in the data storage device 32. Route information may also be stored in the data storage device 32—i.e., a series of road segments (geographically linked to one or more of the defined maps) that together define a route the user may take from a starting location (e.g., the user's current location) to a destination location. As can be appreciated, the data storage device 32 may be part of the controller 34, separate from the controller 34, or part of the controller 34 and part of a separate system.

[0039] The controller 34 includes at least one processor 44 and a computer-readable storage device or media 46. The processor 44 may be a custom-made or off-the-shelf processor, a central processing unit (CPU), a graphics processing unit (GPU), an application-specific integrated circuit (ASIC) (e.g., an ASIC for implementing a neural network), a field-programmable gate array (FPGA), an auxiliary processor among multiple processors connected to the controller 34, a semiconductor-based microprocessor (in the form of a microchip or chipset), a combination thereof, or generally any device for executing instructions. The computer-readable storage device or media 46 may include volatile and non-volatile memory in read-only memory (ROM), random access memory (RAM), and keep-alive memory (KAM).KAM is persistent or non-volatile memory that can be used to store various operating variables while the processor 44 is powered off. The computer-readable storage device or media 46 can be implemented using any of a number of known storage devices, such as PROMs (programmable read-only memory), EPROMs (electrical PROMs), EEPROMs (electrically erasable PROMs), flash memory, or any other electrical, magnetic, optical, or combination storage devices capable of storing data, some of which represent executable instructions used by the controller 34 in controlling the autonomous vehicle 10. In various embodiments, the controller 34 is configured to implement the location system 100, discussed in detail below.

[0040] The instructions may include one or more separate programs, each of which includes an ordered list of executable instructions for implementing logical functions. The instructions, when executed by processor 44, receive and process signals from sensor system 28, perform logic, calculations, methods, and / or algorithms to automatically control the components of autonomous vehicle 10, and generate control signals that are transmitted to actuator system 30 to automatically control the components of autonomous vehicle 10 based on the logic, calculations, methods, and / or algorithms. Although in Fig. 1 only one controller 34 is shown, embodiments of the autonomous vehicle 10 may include any number of controllers 34 that communicate and cooperate via a suitable communication medium or combination of communication media to process the sensor signals, perform logic, calculations, methods and / or algorithms, and generate control signals to automatically control the functions of the autonomous vehicle 10.

[0041] The communication system 36 is configured to wirelessly transmit information to and from other entities 48, such as other vehicles (“V2V” communication), infrastructure (“V2I” communication), networks (“V2N” communication), pedestrians (“V2P” communication), remote transportation systems, and / or user devices (described in more detail with respect to Fig. 2). In an exemplary embodiment, the wireless communication system 36 is configured to communicate via a wireless local area network (WLAN) using the IEEE 802.11 standard, via Bluetooth, or via mobile data communication. However, additional or alternative communication methods, such as a dedicated short-range communication (DSRC) channel, are also contemplated within the scope of the present invention. DSRC channels refer to one-way or two-way short- to mid-range radio communication channels specifically designed for automotive applications and a corresponding set of protocols and standards.

[0042] With further reference to Fig. 2 in various embodiments, the autonomous vehicle 10 described with reference to Fig. 1, may be suitable for use as part of a taxi or shuttle service within a specific geographic area (e.g., a city, a school or business campus, a shopping mall, an amusement park, an event center, or the like). For example, the autonomous vehicle 10 may be associated with an autonomous vehicle-based transportation system. Fig. Figure 2 illustrates an exemplary embodiment of an operating environment, shown generally at 50, including an autonomous vehicle-based remote transportation system (or simply “remote transportation system”) 52, which, as described with reference to Fig. 1, is associated with one or more autonomous vehicles 10a-10n. In various embodiments, the operating environment 50 (which may fully or partially correspond to the Fig. 1) further comprises one or more user devices 54 that communicate with the autonomous vehicle 10 and / or the remote transportation system 52 via a communications network 56.

[0043] The communications network 56 supports communication between devices, systems, and components supported by the operating environment 50 (e.g., via physical communication links and / or wireless communication links). For example, the communications network 56 may include a wireless carrier system 60, such as a cellular phone system, including a plurality of cellular towers (not shown), one or more mobile switching centers (MSCs) (not shown), and any other network components necessary to connect the wireless carrier system 60 to the fixed network. Each cellular tower includes transmit and receive antennas and a base station, with the base stations of various cellular towers connected to the MSCs, either directly or through intermediary devices, such as a base station controller.The wireless carrier system 60 may implement any suitable communications technology, for example, digital technologies such as CDMA (e.g., CDMA2000), LTE (e.g., 4G LTE or 5G LTE), GSM / GPRS, or other current or emerging wireless technologies. Other cell tower / base station / MSC arrangements are possible and could be used with the wireless carrier system 60. For example, the base station and cell tower could be co-located or remote from each other, each base station could serve a single cell tower, a single base station could serve multiple cell towers, or multiple base stations could be coupled to a single MSC, to name a few of the possible arrangements.

[0044] In addition to using the wireless carrier system 60, a second wireless carrier system in the form of a satellite communications system 64 may be used to provide unidirectional or bidirectional communications with the autonomous vehicle 10a-10n. This may be accomplished using one or more communications satellites (not shown) and an uplink transmitting station (not shown). Unidirectional communications may include, for example, satellite radio services, where programmed content data (news, music, etc.) is received from the transmitting station, packaged for upload, and then transmitted to the satellite, which broadcasts the programming to subscribers. Bidirectional communications may include, for example, satellite telephone services that use the satellite to relay telephone communications between the vehicle 10 and the station.Satellite telephony can be used either in addition to or instead of the mobile operator system 60.

[0045] A fixed-line communication system 62 may include a conventional fixed-line telecommunications network connected to one or more landline telephones and connecting the wireless carrier system 60 to the remote transport system 52. For example, the fixed-line communication system 62 may be a public switched telephone network (PSTN) such as those used to provide wired telephony, packet-switched data communications, and the Internet infrastructure. One or more segments of the fixed-line communication system 62 could be implemented using a standard wired network, a fiber optic or other optical network, a cable network, power lines, other wireless networks such as wireless local area networks (WLANs) or networks providing wireless broadband access (BWA), or any combination thereof.Furthermore, the remote transport system 52 need not be connected via the fixed network communication system 62, but could include radiotelephone equipment so that it can communicate directly with a wireless network, such as the wireless carrier system 60.

[0046] Although in Fig. 2, embodiments of the operating environment 50 may support any number of user devices 54, including multiple user devices 54 owned, operated, or otherwise used by a person. Each user device 54 supported by the operating environment 50 may be implemented using any suitable hardware platform. In this regard, the user device 54 may be implemented in a common form factor, including: a desktop computer; a mobile computer (e.g., a tablet computer, a laptop computer, or a netbook computer); a smartphone; a video game device; a digital media player; a component of a home entertainment device; a digital camera or video camera; a wearable computing device (e.g., a smart watch, smart glasses, smart clothing); or the like.Each user device 54 supported by operating environment 50 is embodied as a computer-implemented or computer-based device having the hardware, software, firmware, and / or processing logic necessary to perform the various techniques and methods described herein. For example, user device 54 includes a microprocessor in the form of a programmable device that includes one or more instructions stored in an internal memory structure and is operable to receive binary inputs and generate binary outputs. In some embodiments, user device 54 includes a GPS module that can receive GPS satellite signals and generate GPS coordinates based on those signals.In further embodiments, user device 54 includes cellular communication functionality such that the device performs voice and / or data communications over communication network 56 using one or more cellular communication protocols, as explained herein. In various embodiments, user device 54 includes a visual display, such as a graphical touchscreen display or other display.

[0047] The remote transportation system 52 includes one or more backend server systems (not shown) that may be cloud-based, network-based, or resident at the particular campus or geographic location served by the transportation system 52. The remote transportation system 52 may be staffed with a live advisor, an automated advisor, an artificial intelligence system, or a combination thereof. The remote transportation system 52 may communicate with the user devices 54 and the autonomous vehicles 10a-10n to schedule trips, dispatch autonomous vehicles 10a-10n, and the like. In various embodiments, the remote transportation system 52 stores account information, such as subscriber authentication data, vehicle license plates, profile records, biometric data, behavior patterns, and other related subscriber information.

[0048] According to a typical use case workflow, a registered user of the remote transportation system 52 may create a ride request via the user device 54. The ride request typically specifies the passenger's desired pickup location (or current GPS location), the desired destination (which may identify a predefined vehicle stop and / or a user-defined passenger destination), and a pickup time. The remote transportation system 52 receives the ride request, processes the request, and dispatches a selected one of the autonomous vehicles 10a-10n (when and if available) to pick up the passenger at the designated pickup location and time. The transportation system 52 may also generate and send an appropriately configured confirmation message or notification to the user device 54 to notify the passenger that a vehicle is en route.

[0049] As can be seen, the subject matter disclosed herein provides certain enhanced features and functionality for what may be considered a standard or baseline autonomous vehicle 10 and / or an autonomous vehicle-based transportation system 52. To this end, an autonomous vehicle-based transportation system may be modified, enhanced, or otherwise supplemented to provide the additional functionality described in more detail below.

[0050] According to various embodiments, the controller 34 implements an autonomous drive system (ADS) 70, as shown in Fig. 3. That is, suitable software and / or hardware components of the controller 34 (e.g., the processor 44 and the computer-readable storage medium 46) are used to provide an autonomous propulsion system 70 used in conjunction with the vehicle 10.

[0051] In various embodiments, the instructions of the autonomous driving system 70 may be structured according to function or system. For example, the autonomous driving system 70 may, as shown in Fig. 3, include a computer vision system 74, a positioning system 76, a guidance system 78, and a vehicle control system 80. As can be seen, in various embodiments, the instructions may be divided into any number of systems (e.g., combined, further subdivided, etc.).

[0052] In various embodiments, the computer vision system 74 synthesizes and processes sensor data and predicts the presence, location, classification, and / or history of objects and features of the environment of the vehicle 10. In various embodiments, the computer vision system 74 may include information from multiple sensors (e.g., the sensor system 28), including cameras, lidars, radars, and / or any number of other types of sensors.

[0053] The positioning system 76 processes sensor data along with other data to "locate" a position (e.g., a local position relative to a map, an exact position relative to the lane of a road, vehicle heading, etc.) of the vehicle 10 relative to the surrounding environment, as further explained below. As can be appreciated, various techniques may be employed to perform localization, such as simultaneous localization and mapping (SLAM), particle filters, Kalman filters, Bayesian filters, and the like.

[0054] The guidance system 78 processes sensor data along with other data to determine a route for the vehicle 10 to follow. The vehicle control system 80 generates control signals for controlling the vehicle 10 according to the determined route.

[0055] In various embodiments, the controller 34 implements machine learning techniques to support the functionality of the controller 34, such as feature detection / classification, obstacle mitigation, route traversal, mapping, sensor integration, ground truth determination, and the like.

[0056] In advance, the embodiments disclosed herein may be implemented using a variety of sensor types within the sensor system 28. Briefly referring to Fig. 4, the AV 10 may, for example, include a number of directional sensors 410 distributed throughout the AV 10 (e.g., a front group of sensors 411-415 and a rear group of sensors 416-418), each having a respective field of view as shown, and each configured to generate sensor data of a type. In a particular embodiment, the sensors 410 correspond to radar sensors of a type with particular effectiveness in severe weather conditions (e.g., rain and snowstorms) and that receive "returns" (or "sensor returns") corresponding to return radar signals characterized by a position within space (relative to the AV 10). These sensor returns, as explained in more detail below, may also be characterized by their respective attributes, such as return power, neighborhood similarity, Doppler shift, and the like.

[0057] As briefly mentioned at the beginning, the positioning system 100 is Fig. 1 is designed to create a set of “features” from self-consistent groups of sensor returns (e.g., radar returns) and then track the position of these features in subsequent time steps.

[0058] For illustration purposes, Fig. 5 shows, in a perspective view, an AV 10 moving along a path 502 towards a series of objects in the environment, namely a traffic sign 530, stationary masts 511 and 512, a parked vehicle 531 and moving vehicles 521 and 522 (moving in the same direction as the AV 10). Fig. Figure 6 illustrates the corresponding sensor data in the form of sensor returns 600 that can be received by the AV 10 (corresponding to the symbol 610 in Fig. 6). It should be noted that although the sensor returns 600 are conceptually represented as points on the surfaces of objects sensed by the corresponding sensors, these sensor returns actually correspond to the sensor data received at the AV 10 and include information about the position of the returns (e.g., distance from the AV 10) as well as various attributes of the return signal determined by the sensor system 28.

[0059] With further reference to Fig. 6 in conjunction with Fig. 5, the sensor returns 621 correspond to the returns from the rear of the moving vehicle 521, the sensor returns 631 to the returns of the parked vehicle 531, the sensor returns 630 to the returns of the traffic sign structure 530, the sensor returns 622 to the returns from the rear of the moving vehicle 522, and the sensor returns 611 and 612 to the stationary objects 511 and 512, respectively.

[0060] In addition to the spatial position of each sensor return 600, the AV 10 also has (via the sensor system 28) a set of attribute values ​​for each return, which can typically vary depending on the type of return and the type of sensor used to generate the sensor data. For example, in one embodiment, sensor returns 600 are radar sensor returns characterized by the attributes of return power, Doppler shift (which indicates the relative velocity of the object with respect to the AV 10), and neighborhood similarity (e.g., how similar each return is to the nearest neighbor(s)). For example, the returns 600 in Fig. 6 is illustrated by different shading levels corresponding to different Doppler shift values. Thus, the moving vehicle 521 and 522 yields returns 621 and 622, which differ from the returns 611, 612, 630, and 631 of the stationary objects 511, 512, 530, and 531.

[0061] Accordingly, one or more “sensor data groups” can be created based on self-consistent attributes of the sensor returns 600. For example, taking into account the Doppler shift attribute, the sensor returns 621 can be Fig. 6 elements of one sensor data group, the sensor returns 630 elements of another sensor data group, etc. As used herein, the term "self-consistent" with respect to the attributes of the sensor return refers to a certain degree of similarity. For example, the sensor data groups may be determined based on a machine learning method.

[0062] In various embodiments, the returns 600 are divided into a geometric "grid" to facilitate the processing of the sensor data as described in more detail below. With reference to Fig. 7, for example, a square grid 700 with any side length (e.g., 1.0 m) can be defined to divide the sensor returns (in this illustration, returns 630 and 612) into a matrix of cells 710. For example, returns 612 are represented as elements of a cell 702, while returns 630 are represented as spanning multiple cells 710.

[0063] As in Fig. As further illustrated in Figure 8, the sensor returns within individual cells of grid 700 may be processed to define "features" of those sensor returns. These features may be represented as individual layers (e.g., layers 801-804) parallel to grid 700, with each layer corresponding to a separate feature. Thus, a vector of features for a particular cell may be visualized as a column, such as column 820 in Fig. 8.

[0064] As used herein, the term "feature," with respect to the attributes of a subset of sensor returns, refers to a mathematical function applied to those attributes to derive a simplified value that characterizes that subset of features in some way. Depending on the embodiment, any number of features may be determined, and each feature may be a function of one or more of the corresponding attributes.

[0065] In some embodiments, the features include one or more summary statistics, such as mean, standard deviation, skewness, range, and the like. In the case of radar sensor returns, these may be, for example, mean return power, Doppler shift range, average similarity, or the like. In further embodiments, the feature is a histogram of an attribute, such as a histogram of Doppler shift, a histogram of return power, a histogram of return similarity, etc. In some embodiments, the histogram is normalized—e.g., to include frequencies within the range (0,0, 1,0). In some embodiments, a histogram feature may be characterized by the vertices of its convex hull (i.e., the convex polygon forms a convex hull bounding the histogram).

[0066] Regardless of the number and type of features defined for each of the sensor data groups (which in some embodiments are further divided into individual grid cells 700), the positioning system 100, as described in more detail below, is configured to correlate these features in successive time steps as the AV 10 travels along the roadway 502, thus providing information that can be used to determine the position of the AV 10 (e.g., by the positioning system 76 of Fig. 3).

[0067] With reference to Fig. 9, an exemplary location system generally includes a feature determination module 930 and a feature correlation module 940. The feature determination module is generally configured to receive sensor data 901, which in various embodiments includes sensor returns (e.g., radar and / or lidar sensor returns) associated with objects in the environment (as shown in the Fig. 5 and Fig. 6). The feature determination module 930 is further configured to create a sensor data group (such as groups 630 and 621 in Fig. 6), each containing a self-consistent subset of the plurality of sensor returns based on their corresponding attributes. The feature determination module 930 then determines a set of features for each of the sensor data groups based on the attributes of the sensor returns.

[0068] The output 931 of the feature determination module 930 is thus a set of "placed" features, i.e., features that have a corresponding geometric position in the environment with respect to the AV 10. This geometric position may, for example, correspond to the position of the grid cell containing these features (as in Fig. 8). In further embodiments, the geometric position of the feature is a function of the positions of the sensor returns in the corresponding sensor data group (e.g., center of mass of the sensor returns or the like).

[0069] The feature correlation module 940 is configured to receive placed features 931 and determine a feature correlation between these placed features and a predetermined set of placed features. That is, at each time step (which may be, for example, between 0.1 s and 1.0 s), the feature correlation module attempts to find the best mapping (e.g., homography) of the outputs 931. In other words, the module 940 attempts to find a one-to-one correspondence between the placed features across successive time steps to estimate the distance and direction traveled by the AV 10 during that time step.

[0070] For example, if you look at the Fig. 11, in which the AV 10 travels on a roadway twice: t0 (left) and t1 (right). The AV 10 approaches two objects: the traffic sign 530 and the moving vehicle 521. At t0, the AV 10 observes the sensor data group 1141 and the sensor data group 1121. As described above, these sensor data groups can be created based on the similarity of their mean power outputs, which are Fig. 11 are shown as respective histogram features 1131 and 1122.

[0071] At t1, the AV 10 has moved some distance along its lane, bringing it closer to the traffic sign 530 and the moving vehicle 521, which has changed lanes as shown. At this point, the AV 10 observes the sensor data group 1142 (with a corresponding histogram feature 1132) and the sensor data group 1152 (with a corresponding histogram feature 1122). Although the positions and shapes of the histogram features 1122 and 1132 are not identical to the corresponding histogram features 1122 and 1131, they are similar enough that the feature correlation module 940 can infer (through a suitable feature correlation method) that they correspond to the same objects (i.e., 521 and 530, respectively). The output 941 of the feature correlation module 940 can then be provided to the positioning system 76 of Fig. 3 to further support the localization of the AV 10.

[0072] In this context, it should be understood that stationary objects may represent the best measure of the distance between time steps. In this context, in some embodiments, the feature correlation module (or feature determination module 930) first removes sensor data groups corresponding to those that could be classified as moving objects. For example, with reference to Fig. 10, a machine learning classifier 1020 (e.g., a random forest classifier) ​​may be trained to receive an input 1001 corresponding to the attributes of a particular set of sensor returns and to generate a classification output including a vector of the probability that the sensor returns were received from a dynamic (1021), static-moving (1022), or static-non-moving (1023) object. With reference to Fig. 5, for example, the objects 511, 512 and 530 can be classified as statically immobile objects, the vehicles 521 and 522 as dynamic objects and the parked vehicle 531 as statically mobile objects.

[0073] Module 940 may further refine its correlation by employing outlier detection. According to various embodiments, a random sample consensus (RANSAC) technique is applied by module 940. In general, the RANSAC method iteratively estimates parameters of a mathematical model from a set of observed data—e.g., assignments of putative feature pairs, such as histogram features 1131 and 1132 in Fig. 11, if these assignments include outliers, which are assumed to be data points that lie outside the model itself, for example due to erroneous measurements, extreme noise levels or the like.

[0074] In various embodiments, module 940 uses the RANSAC method to randomly sample the observed data (e.g., feature correlations) and use a voting scheme to find the optimal fit. Data points effectively vote for one or more models, assuming that noisy features do not consistently vote for a particular model. A subset of feature assignments is randomly selected from the input data set, and a first mathematical model is then determined using this subset of data points. It is then determined which elements of the entire data set are consistent with the first mathematical model (i.e., the consensus set). A data point is considered an outlier if it does not fit the first model within an error threshold, which, for example, corresponds to the maximum deviation that can be attributed to the effect of noise.The above is repeated iteratively until the consensus set meets certain criteria, such as a minimum number of consensus data points.

[0075] It should be understood that various embodiments of the positioning system 100 according to the present invention include any number of sub-modules embedded in the controller 34, which may be combined and / or further subdivided with the systems and methods described herein. Furthermore, inputs to the positioning system 100 may be received from the sensor system 28, from other control modules (not shown) associated with the autonomous vehicle 10, from the communication system 36, and / or from other sub-modules (not shown) implemented within the controller 34 of Fig. 1. Furthermore, the inputs can also be subjected to preprocessing, such as partial sampling, noise reduction, normalization, feature extraction, missing data reduction, and the like.

[0076] Furthermore, the various modules described above can be implemented as one or more machine learning models subjected to supervised, unsupervised, semi-supervised, or boosted learning, performing classification (e.g., binary or multi-level classification), regression, clustering, dimensionality reduction, and / or other such tasks. Examples of these models include artificial neural networks (ANNs) (e.g., recurrent neural networks (RNNs) and convolutional neural networks (CNNs)), decision tree models (e.g., classification and regression trees (CARTs)), ensemble learning models (e.g., boosting, bootstrapped aggregation, gradient boost machines, and random forest), Bayesian network models (e.g., naive Bayes), principal component analysis (PCA), support vector machines (SVMs), clustering models (such as k-neighbors, k-means, expectation maximization, hierarchical clustering, etc.), and the like.

[0077] In some embodiments, the learning of the models (e.g., 930, 940, and / or 1020) occurs within a system remote from the vehicle 10 (e.g., the system 52 in Fig. 2) and is subsequently downloaded to the vehicle 10 for normal operation of the vehicle 10. In further embodiments, the learning occurs at least partially within the controller 34 of the vehicle 10 itself, and the model is subsequently shared with external systems and / or other vehicles in a fleet (as in Fig. 2). Training data may be similarly generated by the vehicle 10 or collected externally and divided into training, validation, and test sets prior to learning.

[0078] With reference now to Fig. 12 and continued reference to the Fig. 1-11, the illustrated flowchart provides a control method 1200 that may be performed by the location system 100 according to the present invention. As will be appreciated in light of the invention, the sequence of operations within the method is not limited to sequential execution as depicted in the figure, but may be performed in one or more varying applicable orders according to the present invention. In various embodiments, the method may be scheduled for execution based on one or more predefined events and / or executed continuously during operation of the autonomous vehicle 10.

[0079] In various embodiments, the method begins at 1201, where the AV 10 receives sensor data in the form of, for example, sensor returns with a number of corresponding attributes (as in Fig. 6). As mentioned above, the individual attributes may vary depending on the type of sensor data.

[0080] Subsequently, at 1202, sensor data groups are constructed to contain self-consistent subsets of the sensor returns (based on their attributes). These sensor data groups and their respective sensor returns can be mapped into a spatial grid near the AV 10, as shown in Fig. 7 shown, divided.

[0081] At 1203, features are determined for each sensor data group, as in Fig. 8. The features may, for example, correspond to summary statistics of one or more attributes and / or a histogram or other visualization of one or more attributes. These features are effectively "placed" at a location relative to the AV 10 (e.g., due to the corresponding grid cell, as in Fig. 8).

[0082] Subsequently, starting at 1204, a correlation between the features is determined in subsequent time steps. This means that an association is established between the current set of placed features and a previously determined set of placed features (e.g., a set of placed features from the previous time step). To facilitate this correlation, an outlier removal technique, such as a RANSAC method, can be used. In some embodiments, only stationary features (e.g., traffic signs, pylons, etc.) are considered when determining this relationship.

[0083] Finally, at 1205, the localization of the AV 10 is updated based on the correlation determined at 1204. That is, the estimated movement and position change of the AV 10 during a time step is transmitted to the positioning system 76 ( Fig.3) to assist in determining the absolute position of the AV in its environment.

Claims

[1] Localization method, comprising: receiving sensor data relating to an environment of a vehicle (10), the sensor data including a plurality of sensor returns associated with objects (511, 512, 521, 522, 530, 531) in the environment, each of the sensor returns having a plurality of corresponding attributes; creating a first plurality of sensor data groups, each of which includes a self-consistent subset of the plurality of sensor returns based on their corresponding attributes; defining a first set of features for each of the first plurality of sensor data groups, each feature being based on at least one of the corresponding attributes and each having an associated feature location, and the first set of features including a histogram of one of the corresponding attributes; determining a feature correlation between the first set of features and a second, previously determined set of features with a processor (44); and estimating a position of the vehicle (10) based on the feature correlation. [2] The method of claim 1, wherein the plurality of corresponding attributes includes at least one of a Doppler shift, a return power, and a neighborhood similarity. [3] The method of claim 1, wherein the sensor data includes at least radar data. [4] The method of claim 1, wherein determining the feature correlation includes performing an outlier removal process with respect to the first and second sets of features. [5] System (100) for controlling a vehicle (10), comprising: a feature determination module (930) including a processor (44) configured to: - receiving sensor data relating to an environment of a vehicle (10), the sensor data including a plurality of sensor returns associated with objects (511, 512, 521, 522, 530, 531) in the environment, each of the sensor returns having a plurality of corresponding attributes; - Creating a first plurality of sensor data groups, each including a self-consistent subset of the plurality of sensor returns based on their respective attributes; and defining, for each of the first plurality of sensor data groups, a first set of features, wherein each feature is based on at least one of the respective attributes and each has an associated feature position, and wherein the first set of features includes a histogram of one of the respective attributes; and a feature correlation module (940) configured to determine, with a processor (44), a feature correlation between the first set of features and a second, predetermined set of features. [6] System (100) according to claim 5, wherein: the plurality of corresponding attributes includes at least one Doppler shift, return power, and neighborhood similarity; and the sensor data is at least one of the radar data and the lidar data. [7] The system (100) of claim 5, wherein the feature correlation module (940) performs a method for removing outliers with respect to the first and second sets of features. [8] The system (100) of claim 7, wherein the method for removing outliers comprises a random sample consensus (RANSAC) method.

Images