An autonomous vehicle can obtain sensor data. Upon determining that the autonomous vehicle is in a lane adjacent a shoulder, and there is an object in the shoulder, the autonomous vehicle can determine if performing a lane change maneuver out of the lane prior to the autonomous vehicle being positioned adjacent to the object is feasible. If it is, the lane change maneuver can be performed. If it is not, a nudge maneuver and/or a deceleration can be performed.

A computer-implemented method of determining a series of control signals for controlling an autonomous vehicle to implement a planned speed change maneuver comprises: receiving from a maneuver planner a position target for the planned speed change maneuver; selecting, from a predetermined family of kinematic functions, a kinematic function for carrying out the planned speed change maneuver, each kinematic function being a first or higher order derivative of acceleration with respect to time; and using the selected kinematic function to determine a series of control signals for implementing the planned speed change maneuver; wherein the kinematic function is selected in a constrained optimization process as substantially optimizing a cost function defined for the speed change maneuver, subject to a set of hard constraints that: (i) require a final acceleration, speed and position corresponding to the selected kinematic function to satisfy, respectively, an acceleration target, a speed target and the position target, given an initial speed and acceleration of the autonomous vehicle, and (ii) impose a jerk magnitude upper limit on the selected kinematic function.

The present disclosure relates to a method and system for integrated path planning and path tracking control of an autonomous vehicle. The method includes: obtaining five input control variables and eleven system state variables of an autonomous vehicle at current time; constructing a vehicle path planning-tracking integrated state model according to the obtained variables at the current time; enveloping external contours of two autonomous vehicles using elliptical envelope curves to determine elliptical vehicle envelope curves of the two autonomous vehicles, respectively; determining time to collision (TTC) between the vehicles according to elliptical vehicle envelope curves and vehicle driving states; establishing an objective function of a model prediction controller (MPC) according to the model; and solving the objective function based on the TTC, and determining input control variables to the MPC at the next time. Autonomous vehicle collision avoidance can be achieved according to the present disclosure.

A collision detection method, an electronic device, and a storage medium are provided, which relate to a field of artificial intelligence technology, and in particular to fields of intelligent transportation and autonomous driving technologies. The method includes: determining a predicted travel range of a target object based on a planned travel trajectory of the target object and a historical travel trajectory of the target object; determining, in response to a target obstacle being detected, a predicted travel range of the target obstacle based on a current travel state of the target obstacle; and determining whether the target object has a risk of colliding with the target obstacle, based on the predicted travel range of the target object and the predicted travel range of the target obstacle.

Object identification may be provided herein. A feature extractor may extract a first set of visual features, extract a second set of visual features, concatenate the first set of visual features, the second set of visual features, and a set of bounding box information, determine a number of object features and a global feature for a scene, and receive ego-vehicle feature information associated with an ego-vehicle. An object classifier may receive the number of object features, the global feature, and the ego-vehicle feature information, generate relational features with respect to relationships between each of the number of objects from the scene, and classify each of the number of objects from the scene based on the number of object features, the relational features, the global feature, the ego-vehicle feature information, and an intention of the ego-vehicle.

A deep neural network(s) (DNN) may be used to detect objects from sensor data of a three dimensional (3D) environment. For example, a multi-view perception DNN may include multiple constituent DNNs or stages chained together that sequentially process different views of the 3D environment. An example DNN may include a first stage that performs class segmentation in a first view (e.g., perspective view) and a second stage that performs class segmentation and/or regresses instance geometry in a second view (e.g., top-down). The DNN outputs may be processed to generate 2D and/or 3D bounding boxes and class labels for detected objects in the 3D environment. As such, the techniques described herein may be used to detect and classify animate objects and/or parts of an environment, and these detections and classifications may be provided to an autonomous vehicle drive stack to enable safe planning and control of the autonomous vehicle.

A vehicle includes a position recognition module that creates position information, a road information combining module configured to create first precise map information including a driving route of the vehicle, an object combining module that creates second precise map information including a driving route of a surrounding vehicle around the present vehicle, a lane link determination module that selects, from the second precise map information, a lane link at which a first lane and a second lane intersect or join each other, a target determination module configured to determine a target vehicle, a pass priority determination module that determines a pass priority at which each of the present vehicle and the target vehicle passes through the lane link, an object route creation module that creates a driving route of the target vehicle, and an adaptive route determination module that determines an adaptive driving route of the present vehicle.

Aspects of the disclosure may enable autonomous vehicles to respond to emergency vehicles. For instance, sensor data identifying an emergency vehicle approaching the autonomous vehicle may be received. A predicted trajectory for the emergency vehicle may be received. Whether the autonomous vehicle is impeding the emergency vehicle may be determined based on the predicted trajectory and map information identifying a road on which the autonomous vehicle is currently traveling. Based on a determination that the autonomous vehicle is impeding the emergency vehicle, the autonomous vehicle may be controlled in an autonomous driving mode in order to respond to the emergency vehicle.

Systems and methods for predicting a trajectory of a moving object are disclosed herein. One embodiment downloads, to a robot, a probabilistic hybrid discrete-continuous automaton (PHA) model learned as a deep neural network; uses the deep neural network to infer a sequence of high-level discrete modes and a set of associated low-level samples, wherein the high-level discrete modes correspond to candidate maneuvers for the moving object and the low-level samples are candidate trajectories; uses the sequence of high-level discrete modes and the set of associated low-level samples, via a learned proposal distribution in the deep neural network, to adaptively sample the sequence of high-level discrete modes to produce a reduced set of low-level samples; applies a sample selection technique to the reduced set of low-level samples to select a predicted trajectory for the moving object; and controls operation of the robot based, at least in part, on the predicted trajectory.

Disclosed are an apparatus for controlling a vehicle, a system including the apparatus, and a method for controlling the apparatus. The apparatus includes: a reference lane calculator to calculate a reference lane based on a traveling condition of the vehicle; a target determining device to determine a target of interest based on the reference lane and a predicted path of an object around the vehicle; and a control parameter calculating device to calculate a control parameter of the vehicle based on a traveling state of the target of interest.