/Pure-MatSU

Pure-Matlab Simulation for UAVs

Primary LanguageMATLABGNU General Public License v3.0GPL-3.0

Pure-MatSU

Pure-Matlab Simulation for UAVs

Currently designed for fixed-wing UAVs only

This simulator is written in pure Matlab, i.e. does not make use of Simulink.

It is a free-running simulation, meaning that it will execute as fast as it can, according to the capabilities of your system. It is not locked to a real-time clock.

It provides for fixed time-step with Forward-Euler integration, or the use of Matlab's ode45 and ode15s solvers.

Please use the version releases to ensure that your downloaded version is as bug-free as possible and has updated documentation.

User Instructions

Follow these instructions to use this simulator:

  1. Load path: Run the load_path script to setup the execution path

  • Edit simulation options: Edit the simulation_options function and adjust the simulation parameters to your needs.

    Ample documentation is provided inline. This is also where you can set the initialization options.

    You can enable live plots and graphic output, but keep in mind that this places a significant burden to the system and slows the simulation down.

    Finally, this is where you can select the simulated vehicle, by setting the sim_options.vehicle parameter to one of the contents of the vehicles folder, but only one model is currently available.

  • (optional) Edit vehicle parameters: To change the model parameters of a vehicle, open its corresponding function from the vehicles folder. Currenlty only the skywalker_2013 model is provided. Change any parameter value to your liking.

    Specifically the skywalker_2013 is modeled like the corresponding model from last_letter and more modeling information can be found in the uav-modeling documenation.

  • Run simulation: Run the simulation function (sim_output = simulation(simulation_options())) to generate the simulation output in the sim_output workspace struct variable. The vehicle states trajectory is saved in the array_states member. The input history is saved in the array_inputs member.

  • Plot resulting trajectory: Run plot_trajectory(sim_output, sim_options) to plot the resulting trajectory. Read the function documentation for more graphics options.

Writing a new controller

You can select the controller which will be applied on the simulated vehicle in the simulation_options function.

A controller is a function which is given the VehicleState instance and generates a 4x1 vector with the control inputs to the vehicle. The contents of this vector are:

  1. Aileron input - range: [-1,1]
  2. Elevator input - range: [-1,1]
  3. Throttle input - range: [0,1]
  4. Rudder input - range: [-1,1]

and obviously correspond to vehicles with the airplane form.

By default, the option is set to 0, corresponding to the StaticOutput controller. This controller generates a constant, zero output for all four inputs.

To write your own controller, first create a new controller function in the controllers folder. Use the StaticOutput controller as a template. Your controller class must have at least a constructor to initialize itself and a method ctrl_output = calc_output(sim_vehicle, t) with the same signature as the one in StaticOutput.

Then, edit the Controller class to assign an index to your controller and add an option for it in the class constructor. This is also where you will setup any initialization options for your controller. Finally, edit the simulation_options function and set your new controller index in the sim_options.controller.type field, along with any other required parameters.