• Problem: Schedule the task for UAV.
• Target:
  • Minimize the length of path;
  • Minimize the execution time.
• Presumption:
  • UAV moves in a 2D plane;
  • All UAV are equal;
  • UAV is capable for all the task;
  • No limitation for the maximum number of tasks executed by a single UAV;
  • No time is needed for task execution and target is still during the task execution;
  • No consideration for the following action after the UAV finish all its task;
  • No interval between tasks.
• Constrains:
  • UAV follows the Dubins Path;
  • The task in sequence must be executed in order;
  • No collision between UAV.

We choose CGA as the baseline. The fitness is the sum of length of all the UAVs.

We use the average gene-distance to evaluate the ability of our algorithm to explore the entire space, the higher, the better.

• NaiveGA

  A simple implementation for the general GA algorithm.

  • Purpose:
    • Test the GA algorithm;
    • Explore the resolve the bottleneck of the performance;
    • An example for the GA interface;

• SchGA

  The solution for the project.

  • Build the Project

    $ mkdir build 
    $ cd build
    $ cmake ..
    $ make
    

    • cmake configuration

      -DPRINT_INFO: Print the information to the terminal, ON by default.

      -DADVANCED_FEATURE: Select the algorithm, ON -> our algorithm, OFF -> baseline algorithm, ON by default.

  • Evaluate the Project

    ./GAHW recorder.csv
    

    Program will generate a random task configuration and ask for you confirmation.

    The statistical data will be recorded to the csv file.

  • Task Descriptor

    To be filled.

  • Gene Coding

    To be filled.

  • Fitness Function

    • Dubins Path
    • Collision Detection
    • Correction for the gene violating constrains

  • Interbred

    To be filled.

  • Mutation

    To be filled.

  • Selection

    To be filled.

  • Advanced Features

    To be filled.

• bug detecting...