A lane going in straight line.

A network for actor

An multi-agent learned with PPO reference: https://github.com/ChenglongChen/pytorch-DRL

Replay memory buffer

Replay memory buffer

This is used to record videos of evaluations

An actor-critic network that shared lower-layer representations but have distinct output layers

A dynamical bicycle model, with tire friction and slipping. See Chapter 2 of Lateral Vehicle Dynamics. Vehicle Dynamics and Control. Raj

An continuous action space for throttle and/or steering angle. If both throttle and steering are enabled, they are set in this order: [throt

A network for critic

An discrete action space of meta-actions: lane changes, and cruise control set-point.

A viewer to render a highway driving environment.

An observation class that collects directly what the simulator renders Also stacks the collected frames as in the nature DQN. Specific k

Observe the kinematics of nearby vehicles.

An multi-agent learned with DQN reference: https://github.com/ChenglongChen/pytorch-DRL

Obstacles on the road.

Observe an occupancy grid of nearby vehicles.

A road is a set of lanes, and a set of vehicles driving on these lanes.

A sinusoidal lane.

A pygame Surface implementing a local coordinate system so that we can move and zoom in the displayed area.

A generic environment for various tasks involving a vehicle driving on a road. The environment contains a road populated with vehicles, and

A lane on the road, described by its central curve.

A type of action specifies its definition space, and how actions are executed in the environment

A unified agent interface: - interact: interact with the environment to collect experience - _take_one_step: take one step -

A lane going in circle arc.

A vehicle piloted by two low-level controller, allowing high-level actions such as cruise control and lane changes. - The longitudinal contr

A vehicle using both a longitudinal and a lateral decision policies. - Longitudinal: the IDM model computes an acceleration given the preced

Estimator for the interval-membership of a LinearVehicle under parameter uncertainty. The model trajectory is stored in a model_vehicle, and

An multi-agent learned with ACKTR reference: https://github.com/ChenglongChen/pytorch-DRL

Landmarks of certain areas on the road that must be reached.

A visualization of a lane.

A lane side line type.

A Vehicle whose longitudinal and lateral controllers are linear with respect to parameters.

A controlled vehicle with a specified discrete range of allowed target speeds.

A highway merge negotiation environment. The ego-vehicle is driving on a highway and approached a merge, with some vehicles incoming on the

Estimator for the parameter of a LinearVehicle.

A visualization of a road lanes and vehicles.

Common interface for objects that appear on the road, beside vehicles. For now we assume all objects are rectangular. TODO: vehicles and

A visualization of objects on the road.

A moving vehicle on a road, and its kinematics. The vehicle is represented by a dynamical system: a modified bicycle model. It's state i