vertiport_autonomy.agents package

Subpackages

• vertiport_autonomy.agents.drl package
  • Module contents

Submodules

vertiport_autonomy.agents.base module

Abstract base class for all agents.

class vertiport_autonomy.agents.base.BaseAgent(name: str)

Bases: ABC

Abstract base class for vertiport autonomy agents.

__init__(name: str)

Initialize the agent.

Parameters:

name – Human-readable name for the agent

abstractmethod act(observation: Dict[str, Any]) → ndarray

Select actions based on the current observation.

Parameters:

observation – Current environment observation

Returns:

Array of actions for each drone

abstractmethod reset() → None

Reset agent state for a new episode.

__str__() → str

String representation of the agent.

vertiport_autonomy.agents.heuristic module

Simplified Heuristic Baseline Agent

Provides a working baseline for comparison with DRL agents. Uses simple FCFS strategy with basic conflict avoidance.

class vertiport_autonomy.agents.heuristic.SimpleHeuristicAgent(name: str = 'SimpleHeuristic')

Bases: BaseAgent

Simple heuristic agent that uses FCFS strategy.

__init__(name: str = 'SimpleHeuristic')

Initialize the simple heuristic agent.

act(observation: Dict[str, Any]) → ndarray

Select actions based on simple heuristic strategy.

Parameters:

observation – Current environment observation

Returns:

Array of actions for each drone

reset() → None

Reset agent state for a new episode.

vertiport_autonomy.agents.heuristic.run_simple_heuristic(scenario_path: str, max_steps: int = 200)

Run simple heuristic agent on given scenario

vertiport_autonomy.agents.heuristic.main()

Test simple heuristic on all scenarios

Module contents

Agent implementations.

class vertiport_autonomy.agents.BaseAgent(name: str)

Bases: ABC

Abstract base class for vertiport autonomy agents.

__init__(name: str)

Initialize the agent.

Parameters:

name – Human-readable name for the agent

__str__() → str

String representation of the agent.

abstractmethod act(observation: Dict[str, Any]) → ndarray

Select actions based on the current observation.

Parameters:

observation – Current environment observation

Returns:

Array of actions for each drone

abstractmethod reset() → None

Reset agent state for a new episode.

class vertiport_autonomy.agents.SimpleHeuristicAgent(name: str = 'SimpleHeuristic')

Bases: BaseAgent

Simple heuristic agent that uses FCFS strategy.

__init__(name: str = 'SimpleHeuristic')

Initialize the simple heuristic agent.

act(observation: Dict[str, Any]) → ndarray

Select actions based on simple heuristic strategy.

Parameters:

observation – Current environment observation

Returns:

Array of actions for each drone

reset() → None

Reset agent state for a new episode.