Apparatus, system, and method for planning intersection turn

Abstract

To provide an apparatus, a system, and a method for a track planning.SOLUTION: A disclosed apparatus, a system, and a method are directed to track planning. An example apparatus comprises one or more processors operable to receive an instruction to turn within an intersection, plan a track for an autonomous vehicle to turn within the intersection on the basis of map data comprising one or more optimal turn paths associated with the intersection, and instruct the autonomous vehicle to pass through the intersection according to the track. The one or more optimal turn paths are generated on the basis of the turn paths of one or more human driving vehicles.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] FIELD OF THE DISCLOSURE The present disclosure relates to devices, systems, and methods for vehicle control, and more particularly, to devices, systems, and methods for vehicle control by planning intersection turns. [Background technology]

[0002] Intersections present challenges for vehicles because vehicles compete for space within the same area. Intersections include multiple lanes, opposing traffic, and traffic signs that mandate right-of-way decisions. Relying on real-time decision-making at intersections can introduce uncertainty. Therefore, there is a need to establish planned intersection turns for vehicles with the goal of reducing potential traffic risks and improving the efficiency of navigating intersections. Summary of the Invention

[0003] In one embodiment, an apparatus for trajectory planning comprises one or more processors, the one or more processors operable to: receive an instruction to turn within an intersection; plan a trajectory for an autonomous vehicle to turn within the intersection based on map data comprising one or more optimal turn paths associated with the intersection, where the one or more optimal turn paths are generated based on the turn paths of one or more human-operated vehicles; and instruct the autonomous vehicle to pass through the intersection according to the trajectory.

[0004] In another embodiment, a method for trajectory planning includes receiving an instruction to turn within an intersection; planning a trajectory for an autonomous vehicle to turn within the intersection based on map data comprising one or more optimal turn paths associated with the intersection, where the one or more optimal turn paths are generated based on the turn paths of one or more human-operated vehicles; and instructing the autonomous vehicle to pass through the intersection according to the trajectory.

[0005] These and further features provided by the embodiments of the present disclosure will be more fully understood when the following detailed description is considered in conjunction with the drawings. [Brief explanation of the drawings]

[0006] The embodiments set forth in the drawings are illustrative and exemplary in nature and are not intended to limit the present disclosure. The following detailed description of illustrative embodiments can be understood when read in conjunction with the following drawings, in which like structure is designated with like reference numerals and in which:

[0007] [Figure 1] FIG. 1 is a diagram that schematically depicts an exemplary system for trajectory planning for intersection turns of the present disclosure, in accordance with one or more embodiments shown and described herein. [Figure 2] FIG. 1 is a diagram that schematically depicts example components of an apparatus and system for trajectory planning for intersection turns of the present disclosure, according to one or more embodiments shown and described herein. [Figure 3] FIG. 1 depicts an example block diagram for generating a planned trajectory at an intersection of the present disclosure, according to one or more embodiments shown and described herein. [Figure 4] FIG. 10 depicts a flowchart of example steps for generating a planned intersection maneuver at an intersection of the present disclosure, according to one or more embodiments shown and described herein. DETAILED DESCRIPTION OF THE INVENTION

[0008] Embodiments of the present disclosure include devices, systems, and methods for trajectory planning at intersections based on map data including one or more optimal turn paths. The optimal turn paths may be generated based on the turn paths of one or more human-driven vehicles. Embodiments of the present disclosure include devices, systems, and methods useful for planning trajectories for autonomous vehicles through intersections. Autonomous vehicles that rely on real-time decision-making at intersections may make undesirable wide turns that force the autonomous vehicle out into the intersection, making it difficult to navigate through the intersection. Such wide turns may disrupt traffic flow and increase the risk of collisions with other vehicles, pedestrians, or bicyclists in the intersection.

[0009] By encoding the optimal turn path of a human-driven vehicle into an intersection in map data, an autonomous vehicle can make turns within the intersection more naturally and desirably. Embodiments of the present disclosure further include devices, systems, and methods that have an artificial intelligence function continuously refine the optimal turn path based on the generated planned trajectory. Thus, the devices, systems, and methods of the present disclosure are useful for mitigating problems associated with navigating a vehicle through an intersection, such as wide turns that may extend into the intersection, thereby improving overall traffic flow and vehicle interaction. Furthermore, the artificial intelligence function continuously refines the optimal turn path based on the generated planned trajectory to ensure the trajectory planning system evolves, contributing to efficient and desirable safe navigation of intersections by autonomous vehicles in response to changing traffic dynamics and improving system capability by ensuring that the trajectory plan remains robust and responsive to evolving road conditions, traffic patterns, and other dynamic factors. The trajectory planning system not only prioritizes safety and natural maneuvers, but also demonstrates the efficiency of its use of map data. The integration of optimal turn paths into map data facilitates accurate decision-making for autonomous vehicles, leveraging spatial information to effectively navigate intersections.

[0010] Various embodiments of methods and systems for trajectory planning are described in more detail herein. Wherever possible, the same reference numbers are used throughout the drawings to refer to the same or like parts. As used herein, the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to "a" component includes aspects having two or more such components unless the context clearly dictates otherwise.

[0011] 1 and 2 schematically depict an example trajectory planning system 100. The trajectory planning system 100 may include one or more controllers 201, which may further include one or more modules, such as a trajectory generation module 222 and an optimal turn path module 232.

[0012] The trajectory generation module 222 may include one or more first machine learning algorithms, such as a first neural network 322. The trajectory generation module 222 may generate the planned trajectory 105. The planned trajectory 105 may include a future turn path and velocity, time, and kinematics of the vehicle 101 associated with the future turn path.

[0013] The optimal turn path module 232 may include one or more second machine learning algorithms, such as a second neural network 432. The second machine learning algorithm may generate one or more optimal turn paths 315 (as shown in FIG. 3 ) associated with the intersection 135. Information about the optimal turn path 315 and the intersection 135 may be stored in map data 217. The map data 217 may include pedestrian crosswalks, traffic lights, traffic signs, barriers, road lanes, road edges, shoulders, dividers, paint marks, poles, or combinations thereof. In an embodiment, the map data 217 may be high definition (HD) map data or standard definition (SD) map data. SD maps, including SD map data, are based on road curves, elevation, and coordinates. SD maps may have a meter resolution. HD maps, including HD map data, may include more detail than SD maps, including SD map data. HD maps may have sub-meter to centimeter resolution. HD map data may include all the information contained in SD map data and much more, such as road geometry, road markings, traffic signs, poles, guardrails, walls, and barriers.

[0014] Trajectory planning system 100 may include one or more vehicles 101, which may be autonomous vehicles. In some embodiments, one or more controllers 201 are included in vehicle 101. In some embodiments, a portion of vehicle 101 may include a communication device, such as vehicle network interface hardware, operable to communicate wirelessly with controller 201. In some embodiments, controller 201 may be included in one or more servers, including a server communication device, such as network interface hardware 206, operable to communicate with vehicle 101.

[0015] Each of the vehicles 101 may be an automobile or any other passenger or non-passenger vehicle, such as a land vehicle, an underwater vehicle, and / or an air vehicle. Each of the vehicles 101 may be an autonomous vehicle that navigates its environment with limited or no human input. Each of the vehicles 101 may be driven on a road and, for example, perform vision-based lane centering using one or more sensors. Each of the vehicles 101 may include an actuator that drives the vehicle, such as a motor, an engine, or any other powertrain. The vehicles 101 may travel on various surfaces, such as, but not limited to, roads, highways, streets, freeways, bridges, tunnels, parking lots, garages, off-road trails, railroad tracks, or any surface on which a vehicle may operate.

[0016] In an embodiment, vehicle 101 may travel on a road 137 that includes one or more intersections 135. Intersection 135 may include one or more lanes. Intersection 135 may include traffic signs, signals, roundabouts, and other structures to control traffic flow. Intersection 135 may be four-way, cross-street, three-way (such as T-junctions and Y-junctions), or five-way or more. For example, as shown in FIG. 1 , road 137 may include longitudinal and transverse roads. The longitudinal road may include two directional lanes, i.e., southbound lane 121 and northbound lane 122. The transverse road may include two directional lanes, i.e., westbound lane 131 and eastbound lane 132. The longitudinal and transverse roads may intersect at intersection 135. Intersection 135 may include one or more curbs 125 at the junction of the longitudinal and transverse roads.

[0017] Referring to FIG. 2, exemplary components of a controller 201 are schematically depicted. While FIG. 2 shows one controller 201, in some embodiments, the trajectory planning system 100 may include two or more controllers 201. The controller 201 or the vehicle 101 may include one or more vision sensors 208 and vehicle sensors 212. The vision sensors 208 may be used to capture images or video of the environment surrounding the vehicle 101. In some embodiments, the one or more vision sensors 208 include one or more imaging sensors configured to operate in the visible and / or infrared spectrum and detect visible and / or infrared light. Furthermore, while certain embodiments described herein are described with reference to hardware that detects light in the visible and / or infrared spectrum, it should be understood that other types of sensors are contemplated. For example, the systems described herein may include one or more LiDAR sensors, radar sensors, sonar sensors, or other types of sensors that collect data that may be incorporated into or supplement the data collection described herein. A ranging sensor, such as radar, may be used to obtain rough depth and speed information about the field of view of the vehicle 101. The one or more vision sensors 208 may include a forward-facing camera mounted on the vehicle 101. The one or more vision sensors 208 may be any device having an array of sensing devices capable of detecting radiation in the ultraviolet, visible, or infrared wavelength bands. The one or more vision sensors 208 may have any resolution. In some embodiments, one or more optical components, such as a mirror, a fisheye lens, or any other type of lens, may be optically coupled to the one or more vision sensors 208. In the embodiments described herein, the one or more vision sensors 208 may provide image data to one or more processors 204 or another component communicatively coupled to the communication path 203. In some embodiments, the one or more vision sensors 208 may also provide navigation support. That is, data captured by the one or more vision sensors 208 may be used to navigate the vehicle autonomously or semi-autonomously.

[0018] The controller 201 or vehicle 101 may include one or more vehicle sensors 212. Each of the one or more vehicle sensors 212 is connected to the communication path 203 and communicatively coupled to the one or more processors 204. The one or more vehicle sensors 212 may include one or more speed or motion sensors that detect and measure motion and changes in motion of the vehicle, e.g., the vehicle 101. The motion sensors may include an inertial measurement unit. Each of the one or more motion sensors may include one or more accelerometers and one or more gyroscopes. Each of the one or more motion sensors converts sensed physical vehicle movement into signals indicative of the vehicle's orientation, rotation, speed, or acceleration. The acquired data from the vehicle sensors 212 may be used to determine the vehicle kinematics of the vehicle 101.

[0019] The vision sensors 208 and vehicle sensors 212 may be used to collect vehicle control data, road condition data, and vehicle kinematics data. The vehicle control data, road condition data, and vehicle kinematics data may be used to monitor the actual trajectory of the vehicle 101 as it traverses the intersection 135. The vehicle control data may include the throttle position, braking situation, steering angle, and gear selection of the vehicle 101. The road condition data may include road type, coefficient of friction, and surface irregularities (e.g., bumps). The vehicle kinematics data may include the speed, acceleration, position, and orientation of the vehicle 101.

[0020] The controller 201 may include one or more processors 204. Each of the one or more processors 204 may be any device capable of executing machine-readable executable instructions. The instructions may be in the form of a machine-readable instruction set stored in the data storage component 207 and / or the memory component 202. Accordingly, each of the one or more processors 204 may be a controller, an integrated circuit, a microchip, a computer, or any other computing device. The one or more processors 204 are connected to a communication path 203 that provides signal interconnection between various modules of the system. Thus, the communication path 203 may communicatively connect any number of processors 204 to each other, enabling the modules connected to the communication path 203 to operate in a distributed computing environment. Specifically, each module may operate as a node that can send and / or receive data. As used herein, the term "communicatively connected" means that the connected components are capable of exchanging data signals with each other, such as electrical signals over conductive media, electromagnetic signals over air, optical signals over optical waveguides, and the like.

[0021] Thus, communication path 203 may be formed of any medium capable of transmitting a signal, such as a conductive wire, a conductive trace, an optical waveguide, or the like. In some embodiments, communication path 203 may facilitate the transmission of wireless signals, such as WiFi, Bluetooth, near field communication (NFC), and the like. Furthermore, communication path 203 may be formed of a combination of media capable of transmitting a signal. In one embodiment, communication path 203 comprises a combination of conductive traces, conductive wires, connectors, and buses that cooperate to enable the transmission of electrical data signals to components such as processors, memories, sensors, input devices, output devices, and communication devices. Thus, communication path 203 may comprise a vehicle bus, such as a LIN bus, a CAN bus, a VAN bus, and the like. Furthermore, it should be noted that the term “signal” refers to a waveform (e.g., an electrical waveform, an optical waveform, a magnetic waveform, a mechanical waveform, or an electromagnetic waveform), such as DC, AC, a sine wave, a triangular wave, a square wave, a vibration, and the like, capable of traveling through a medium.

[0022] The controller 201 may include one or more memory components 202 coupled to the communication path 203. The one or more memory components 202 may comprise RAM, ROM, flash memory, a hard drive, or any device capable of storing machine-readable executable instructions such that the machine-readable executable instructions can be accessed by the one or more processors 204. The machine-readable executable instructions may comprise logic or algorithms written in any programming language of any generation (e.g., 1GL, 2GL, 3GL, 4GL, or 5GL), such as a machine language that can be executed directly by a processor, or in assembly language, object-oriented programming (OOP), scripting language, microcode, etc., that can be compiled or assembled into machine-readable executable instructions and stored in the one or more memory components 202. Alternatively, the machine-readable executable instructions may be written in a hardware description language (HDL), such as logic implemented via a field programmable gate array (FPGA) configuration or an application-specific integrated circuit (ASIC), or the like. Thus, the methods described herein may be implemented in any conventional computer programming language, as pre-programmed hardware elements, or as a combination of hardware and software components.

[0023] The one or more memory components 202 may include one or more modules, including a trajectory generation module 222 and an optimal turn path module 232. Each of the one or more modules may include, but is not limited to, routines, subroutines, programs, objects, components, data structures, and the like, that perform a particular task or execute a particular data type, as described below. The data storage component 207 may store map data, including intersections 135, roads 137, and optimal turn paths 315. The data storage component 207 may further store training data 227 for training the trajectory generation module 222 and the optimal turn path module 232. The training data 227 may include ground truth data related to trajectory planning at one or more intersections for human-driven and autonomous vehicles. The data storage component 207 may store historical data 237, such as historical vehicle kinematics data, historical vehicle control data, historical road condition data, historical trajectory data, and other historical data related to the operation of the vehicle 101, as well as historical sensor data of the human-operated vehicle 301 (e.g., as shown in FIG. 3 ). The trajectory generation module 222 and the optimal turn path module 232 may also be stored in the data storage component 207 during or after operation.

[0024] One or more modules, including the trajectory generation module 222 and the optimal turning path module 232, may include one or more machine learning algorithms, such as a neural network. The modules may be trained and provided with machine learning capabilities via the neural network as described herein. By way of example, and not limitation, the neural network may utilize one or more artificial neural networks (ANNs). In an ANN, connections between nodes may form a directed acyclic graph (DAG). The ANN may include node inputs, one or more hidden activation layers, and node outputs, and may utilize activation functions in the one or more hidden activation layers, such as linear functions, step functions, logistic (sigmoid) functions, tanh functions, rectified linear unit (ReLu) functions, or combinations thereof. The ANN is trained by applying the activation functions to a training dataset, determining an optimized solution from adjustable weights and biases applied to the nodes in the hidden activation layers, and generating one or more outputs as the optimized solution with minimized error. In machine learning applications, new inputs (such as one or more generated outputs) may be provided to the ANN model as training data to continually improve accuracy and minimize the ANN model's error. One or more ANN models may utilize one-to-one, one-to-many, many-to-one, and / or many-to-many (sequence-to-sequence) sequence modeling. One or more ANN models may employ a combination of artificial intelligence techniques, such as, but not limited to, deep learning, random forest classifiers, feature extraction from audio or images, clustering algorithms, or combinations thereof. In some embodiments, a convolutional neural network (CNN) may be utilized. For example, a convolutional neural network (CNN) may be used in the field of machine learning as a class of deep feedforward ANNs applied, for example, to audio analysis of recordings. A CNN may be shift- or space-invariant and utilize a shared weight architecture and transformations. Furthermore, each of the various modules may include one or more generative artificial intelligence algorithms.Generative artificial intelligence algorithms may include general adversarial networks (GANs) that have two networks: a generator model and a discriminator model. Generative artificial intelligence algorithms may also be based on variational autoencoders (VAEs) or transformer-based models.

[0025] The controller 201 may include input / output hardware 205 connected to the communication path 203. The input / output hardware 205 may include a monitor, keyboard, mouse, printer, camera, microphone, speaker, and / or other device that receives, transmits, and / or presents data. The controller 201 may include network interface hardware 206 that communicatively connects the controller 201 to external resources (e.g., the vehicle 101 or smart devices), the Internet of Things (IoT), and / or a server. The network interface hardware 206 may be any device that can be communicatively connected to the communication path 203 and that is capable of transmitting and / or receiving data over a network. Thus, the network interface hardware 206 may include a communication transceiver that transmits and / or receives any wired or wireless communication. For example, the network interface hardware 206 may include an antenna, a modem, a LAN port, a WiFi card, a WiMAX card, mobile communication hardware, near-field communication hardware, satellite communication hardware, and / or any wired or wireless hardware that communicates with other networks and / or devices. In one embodiment, the network interface hardware 206 includes hardware configured to operate according to the Bluetooth® wireless communication protocol. For example, the network interface hardware 206 of the trajectory planning system 100 may receive and / or transmit map data 217, planned trajectory 105, and sensor data of the vehicle 101 (e.g., speed data, acceleration data, steering data, yaw data, wheel slip data, lane departure data, time of day, weather conditions, vehicle type, vehicle size, minimum and maximum vehicle turning radii) to a server or the vehicle 101.

[0026] 1 and 3 , an example trajectory planning system 100 is depicted that uses a trajectory generation module 222 to generate a planned trajectory 105 for a vehicle 101 (e.g., an autonomous vehicle) approaching an upcoming intersection 135 to traverse the intersection 135 based on one or more optimal turn paths 315 included in map data 217. FIGS. 1 and 3 further depict an example trajectory planning system 100 that uses an optimal turn path module 232 to generate one or more optimal turn paths 315 associated with intersections 135 in map data 217.

[0027] In an embodiment, the vehicle 101 may detect an upcoming intersection 135 on the road 137 and send a command to the controller 201 to assist in a turn within the intersection 135. After receiving the command to turn, the controller 201 may generate a planned trajectory 105 based on one or more optimal turn paths 315 stored in the map data 217 for the vehicle 101, such as an autonomous vehicle, and command the vehicle 101 to pass through the intersection 135 according to the trajectory. The planned trajectory 105 may include the upcoming turn path and the speed, time, and kinematics of the vehicle 101 associated with the upcoming turn path. Each upcoming turn path may be an entrance point 151 at a boundary of the intersection 135, an exit point 153 at a boundary of the intersection 135, and a geometric path 152 between the entrance point 151 and the exit point 153. In some embodiments, the upcoming turn path may be one of the one or more optimal turn paths 315.

[0028] In some embodiments, the vehicle 101 may use vision sensors 208 and vehicle sensors 212 to generate and send sensor data, such as vehicle kinematics data, vehicle control data, and road condition data, to the controller 201. In an embodiment, the vehicle kinematics data may be vehicle position, vehicle speed, and vehicle acceleration of the vehicle 101. The vehicle control data may be vehicle steering, vehicle throttle, and brake inputs of the vehicle 101. The road condition data may include road surface conditions, road geometry, and traffic conditions of the road 137. Upon receiving the sensor data and other input data, such as local weather conditions, the controller 201 may use a trajectory generation module 222 to generate a planned trajectory 105 for the vehicle 101.

[0029] In some embodiments, the planned trajectory 105 may further be generated based on parameters of the vehicle 101 relative to parameters of the human-driven vehicle 301 associated with one or more optimal turn paths 315. The parameters of the vehicle 101 and the human-driven vehicle 301 may include vehicle length, minimum turning radius, steering system, acceleration and deceleration performance, or other parameters related to turning performance at an intersection 135. The trajectory planning system 100 may compare the similarity of the parameters between the vehicle 101 and the human-driven vehicle 301 and further adjust the planned trajectory 105 inversely proportional to the similarity. In some embodiments, the trajectory planning system 100 may determine whether the similarity is greater than or equal to a threshold similarity. When generating the planned trajectory 105, the trajectory planning system 100 may select optimal turn paths 315 associated with a similarity greater than or equal to the threshold similarity and ignore optimal turn paths 315 associated with a similarity less than the threshold similarity.

[0030] In some embodiments, the trajectory planning system 100 may collect sensor data from the vision sensors 208 and the vehicle sensors 212 and further operate the vehicle 101 to follow the planned trajectory 105 by controlling or adjusting the steering, throttle, and braking inputs of the vehicle 101. The trajectory planning system 100 may monitor the trajectory 154 of the vehicle 101 while the vehicle 101 passes through the intersection to determine whether the trajectory 154 deviates from the planned trajectory 105. In response to determining that the trajectory 154 deviates from the planned trajectory 105 (e.g., the trajectory 154 differs from the geometric path 152 as shown in FIG. 1 ), the trajectory planning system 100 may generate an updated trajectory for the vehicle 101 to pass through the intersection 135 based on the past turning paths 303 and 305 of the human-driven vehicle 301.

[0031] 3 , the trajectory planning system 100 may use the optimal turn path module 232 to generate one or more optimal turn paths 315 associated with an intersection 135 based on the turn paths 303 and 305 of one or more human-driven vehicles 301. In some embodiments, the optimal turn path 315 may be the average turn path of the input turn paths 303 and 305 of the human-driven vehicle 301. In some embodiments, the optimal turn path 315 may be the shortest turn path of the input turn paths 303 and 305 of the human-driven vehicle 301. In some embodiments, the optimal turn path 315 may be the most energy-efficient turn path of the input turn paths 303 and 305 of the human-driven vehicle 301. The one or more optimal turn paths 315 may also be generated based on historical sensor data of the human-driven vehicle 301 over time. The sensor data may include speed data, acceleration data, steering data, yaw data, wheel slip data, lane departure data, time of day, weather conditions, vehicle type, vehicle size, minimum and maximum vehicle turning radii, or combinations thereof.

[0032] 2 and 3 , in an embodiment, the trajectory generation module 222 and the optimal turn path module 232 may include one or more neural networks 322 and 432. Each of the neural networks 322 and 432 may include an encoder, one or more hidden layers, and a decoder. The neural networks 322 and 432 may feed the training data 227 into the encoder during a pre-training process to generate lower-dimensional representations of target input-output pairs. For example, the lower-dimensional representations may include inputs of sensor data collected using the vision sensors 208 and vehicle sensors 212 paired with the planned and detected trajectories of the vehicle 101. The first neural network 322 may output the planned trajectory 105. The second neural network 432 may output the optimal turn path 315, which is incorporated into the map data 217.

[0033] In some embodiments, the trajectory generation module 222 and the optimal turn path module 232 may include one or more neural networks 322 and 432 trained with the training data 227 and the historical data 237. The neural networks 322 and 432 may include an encoder and / or decoder associated with a layer normalization operation and / or an activation function operation. The encoded input data may be normalized and weighted through an activation function before being fed to a hidden layer. The hidden layer may generate a representation of the input data in a bottleneck layer. After neural network processing data to the final layer of the neural network, global layer normalization may be performed to normalize the planned trajectory and the optimal turn path. As described in more detail below, the output may be normalized and transformed using an activation function for training and validation purposes. The activation function may be linear or nonlinear. The activation function may be, without limitation, a sigmoid function, a softmax function, a hyperbolic tangent function (Tanh), or a rectified linear unit (ReLU). The neural networks 322 and 432 may provide historical data 237, such as historical vehicle kinematics data, historical vehicle control data, historical road condition data, historical planned trajectories, historical detected trajectories, historical turning paths, and other historical vehicle operation data, of a human-driven vehicle to the encoder for continuous training.

[0034] In an embodiment, one or more vehicle modules may be pre-trained using training data 227 including ground truth examples and scenarios in which multiple entities (e.g., vehicle 101 and human-driven vehicle 301) are driven on a road 137 including multiple lanes 121, 122, 131, and 132, intersections 135, and curbs 125. Pre-training may include labeling entities and desired planned trajectories and optimal turn paths based on the entities and intersections 135 in the examples and scenarios, and using one or more neural networks 322 and 432 to predict desired and undesired trajectories and optimal turn paths based on the training data 227. Pre-training may further include fine-tuning, evaluation, and testing steps. The modules may be continuously trained using real-world collected data as historical data 237 to adapt to changing conditions and factors and improve performance over time. The neural networks may be further trained based on the activation functions described above. The encoder may generate encoded input data h=(Wx+b) that is transformed from input data of one or more input channels. The encoded input data of one of the input channels is transformed from raw input data x ij From h ij =g(Wx ij +b), which can then be expressed as

number

[0035] 4, a flowchart of a method 400 for trajectory planning is depicted. At block 401, the method 400 includes receiving an instruction to turn within an intersection 135. At block 403, the method 400 includes planning a trajectory 105 for the vehicle 101 to turn within the intersection 135 based on one or more optimal turn paths 315 associated with the intersection 135 in the map data 217. The one or more optimal turn paths 315 are generated based on the turn paths 303 and 305 of one or more human-operated vehicles 301.

[0036] In some embodiments, map data 217 may include pedestrian crosswalks, traffic lights, traffic signs, barriers, road lanes, road edges, shoulders, dividers, paint marks, poles, or combinations thereof. Map data 217 may be HD map data or SD map data.

[0037] In some embodiments, one or more optimal turn paths 315 may further be generated based on historical sensor data of the human-driven vehicle 301 over time. The sensor data may include speed data, acceleration data, steering data, yaw data, wheel slip data, lane departure data, time of day, weather conditions, vehicle type, vehicle size, minimum and maximum vehicle turning radii, or combinations thereof. The planned trajectory 105 may include a future turn path and the speed, time, and kinematics of the vehicle 101 associated with the future turn path.

[0038] In some embodiments, the planned trajectory 105 may be generated using a first machine learning algorithm (such as the first neural network 322 in FIGS. 2 and 3 ) that is trained based on one or more optimal turn paths 315 and parameters of the vehicle 101 relative to parameters of the human-driven vehicle 301 associated with the one or more optimal turn paths 315. The parameters of the vehicle 101 and the parameters of the human-driven vehicle 301 may include vehicle length, minimum turning radius, steering system, acceleration and deceleration performance, or a combination thereof.

[0039] In some embodiments, the one or more optimal turning paths 315 may be generated by a second trained machine learning algorithm (such as the second neural network 432 in FIGS. 2 and 3) configured to reduce the path length of the optimal turning paths.

[0040] At block 405, the method 400 includes commanding the vehicle 101 to pass through the intersection 135 according to the planned trajectory 105. The method 400 may further include operating the vehicle 101 to follow the planned trajectory 105 by controlling or adjusting steering, throttle, braking input, or a combination thereof, of the vehicle 101. The method 400 may further include monitoring a trajectory 154 of the vehicle 101 while passing through the intersection 135, determining whether the trajectory 154 deviates from the planned trajectory 105, and, in response to determining that the trajectory 154 deviates from the planned trajectory 105, generating an updated trajectory for the vehicle 101 to pass through the intersection 135 based on the past turning paths 303 and 305 of the human-driven vehicle 301.

[0041] It should be noted that the terms "substantially" and "about" may be used herein to express the degree of inherent uncertainty that may result from any quantitative comparison, value, measurement, or other representation. These terms are also used herein to express the degree to which a quantitative representation may vary from the stated basis without resulting in a change in the basic functionality of the subject matter at issue.

[0042] While particular embodiments have been shown and described herein, it should be understood that various other changes and modifications can be made without departing from the spirit and scope of the claimed subject matter. Moreover, although various aspects of the claimed subject matter are described herein, such aspects need not be utilized in combination. Accordingly, the appended claims are intended to cover all such changes and modifications that are within the scope of the claimed subject matter.

Claims

1. 1. An apparatus for trajectory planning comprising one or more processors, the one or more processors comprising: receive instructions to turn within the intersection, planning a trajectory for an autonomous vehicle to turn within the intersection based on map data comprising one or more optimal turn paths associated with the intersection, the one or more optimal turn paths being generated based on turn paths of one or more human-operated vehicles; commanding the autonomous vehicle to pass through the intersection according to the trajectory; The apparatus is operable to:

2. The apparatus of claim 1 , wherein the one or more optimal turning paths are further generated based on historical sensor data of the human-driven vehicle over time.

3. 3. The apparatus of claim 2, wherein the trajectory is generated using a first machine learning algorithm that is trained based on the one or more optimal turn paths and parameters of the autonomous vehicle relative to parameters of the human-driven vehicle associated with the one or more optimal turn paths.

4. 4. The apparatus of claim 3, wherein the parameters of the autonomous vehicle and the parameters of the human-driven vehicle comprise vehicle length, minimum turning radius, steering system, acceleration and deceleration performance, or a combination thereof.

5. 3. The apparatus of claim 2, wherein the sensor data comprises speed data, acceleration data, steering data, yaw data, wheel slip data, lane departure data, time of day, weather conditions, vehicle type, vehicle size, minimum and maximum vehicle turning radii, or combinations thereof.

6. The apparatus of claim 1 , wherein the one or more optimal turn paths are generated by a second trained machine learning algorithm configured to reduce a path length of the optimal turn paths.

7. 10. The apparatus of claim 1, wherein the trajectory comprises an upcoming turn path and a velocity, time, and kinematics of the autonomous vehicle associated with the upcoming turn path.

8. The apparatus of claim 7 , wherein the future turn path is one of the one or more optimal turn paths.

9. 10. The apparatus of claim 1, wherein the one or more processors are further operable to operate the autonomous vehicle to follow the trajectory by controlling or adjusting steering, throttle, braking inputs, or a combination thereof, of the autonomous vehicle.

10. The one or more processors further include: monitoring a trajectory of the autonomous vehicle while passing through the intersection; determining whether the trajectory deviates from the trajectory; 10. The apparatus of claim 1, operable, in response to determining that the trajectory deviates from the trajectory, to generate an updated trajectory for the autonomous vehicle through the intersection based on past turning paths of the human-driven vehicle.

11. The apparatus of claim 1 , wherein the map data comprises pedestrian crosswalks, traffic lights, traffic signs, barriers, road lanes, road edges, shoulders, dividers, paint marks, poles, or combinations thereof.

12. The apparatus of claim 1 , wherein the map data is high precision map data or standard precision map data.

13. 1. A method for trajectory planning, comprising: receiving a command to turn within the intersection; planning a trajectory for an autonomous vehicle to turn within the intersection based on map data comprising one or more optimal turn paths associated with the intersection, the one or more optimal turn paths being generated based on turn paths of one or more human-operated vehicles; commanding the autonomous vehicle to pass through the intersection according to the trajectory; A method comprising:

14. the one or more optimal turning paths are further generated based on historical sensor data of the human-driven vehicle over time; 14. The method of claim 13, wherein the sensor data comprises speed data, acceleration data, steering data, yaw data, wheel slip data, lane departure data, time of day, weather conditions, vehicle type, vehicle size, minimum and maximum vehicle turning radii, or combinations thereof.

15. the trajectory is generated using a first machine learning algorithm that is trained based on the one or more optimal turn paths and parameters of the autonomous vehicle relative to parameters of the human-driven vehicle associated with the one or more optimal turn paths; 14. The method of claim 13, wherein the parameters of the autonomous vehicle and the parameters of the human-driven vehicle comprise vehicle length, minimum turning radius, steering system, acceleration and deceleration performance, or a combination thereof.

16. The method of claim 13 , wherein the one or more optimal turn paths are generated by a second trained machine learning algorithm configured to reduce a path length of the optimal turn paths.

17. The method of claim 13 , wherein the trajectory comprises an upcoming turn path and a velocity, time, and kinematics of the autonomous vehicle associated with the upcoming turn path.

18. 14. The method of claim 13, further comprising operating the autonomous vehicle to follow the trajectory by controlling or adjusting steering, throttle, braking inputs, or a combination thereof, of the autonomous vehicle.

19. The method comprises: monitoring a trajectory of the autonomous vehicle while passing through the intersection; determining whether the trajectory deviates from the trajectory; In response to determining that the trajectory deviates from the trajectory, generating an updated trajectory for the autonomous vehicle through the intersection based on a past turning path of the human-driven vehicle; The method of claim 13 further comprising:

20. The method of claim 13 , wherein the map data is high precision map data or standard precision map data.

Images