/hytoperm

Hybrid Trajectory Optimization fro Persistent Monitoring

Primary LanguagePythonMIT LicenseMIT

Hybrid Trajectory Optimizer for Persistent Monitoring

This repository contains a collection of tools that can be used for the design of trajectories in Persistent Monitoring (PM) settings with partitioned mission spaces. The partitioning is assumed to be based on the agent's dynamics, which are allowed to vary throughout the environment. This repository contains all code used for this preprint.

Installation

  1. Setup & activate virtual environment, then install.
$ virtualenv env --python=python3
$ source env/bin/activate
$ pip install -e .

Usage Example

Utilizing the random experiment generator, we can run a simple test experiment

import matplotlib.pyplot as plt
from hytoperm import *

# generate and plot the expriment setup
ex = Experiment.generate()
fig, ax = ex.plotWorld()
ex.agent().plotSensorQuality()

# let the agent compute the visiting sequence and optimize the cycle
ex.agent().computeVisitingSequence()
ex.agent().optimizeCycle()

# plot the optimal cycle
ex.agent().plotCycle()
plt.show()

Coding Conventions

  • All classes use the CamelCase convention
  • All methods use the camelCase convention
    • exceptions are unit test methods, which use test_<test_name> convention
  • Properties and methods that have the prefix _ are considered private
  • All setters and getters are defined in the traditional sense (not the Python properties way)
  • All plot functions are expected to work as follows:
    • they return a PlotObject containing all plot objects that were added by the function
    • they take at least the arguments ax and **kwargs. ax is assumed to be either None or a matplotlib Axes. If ax is None then the matplotlib function gca is invoked to retrieve the current active Axes.
  • Each class definition is seperated by two lines
  • Respect the 80 column limit in all code files. Exceptions:
    • class property comments, which start at the 80 column mark
    • error and warning messages