Αn online identification experiment of a 2-DOF robotic manipulator.

This work is a system identification study, and it's main purpose is the identification of a non-linear conitnuous system on a region of interest. The work is based on Prescribed Performance Control (PPC) [1], Radial Basis Neural Networks (RBF-NNs) [2] and the results of [3] on the satisfaction of the Persistancy of Excitation condition for RBF-NNs.

To run this simulation, you will need MATLAB (at least R2015b). So, clone the folder and run the script as follows:

cd demo
robot_example
% ... wait a lot ...

To plot the weights convergence you can edit and run the sig_recreation.m script inside demo/.

In this repository you will find:

• util/ : Folder containg the utility functions like implementation of the PPC functions and RBF-NNs
• demo/: Identification experiment files
  • demo_experiment/: Results folder
  • robot_example.m: Experiment definition and tunable parameters
  • sim_controller.m: Invokes ODE and handels signal storage and reinitiallization in every iteration
  • robot_plant.m: The simulated system equations
  • robotic_2dof_idnt.m: The control/identification loop
  • index_calcluator.m: Active states calculator for accelerating the simulation
  • sig_recreation.m: Plotting tool to evaluate the weights convergence

[1] Robust adaptive control of feedback linearizable mimo nonlinear systems with prescribed performance
[2] Universal approximation using radial-basis-function networks
[3] Persistency of excitation in identification using radial basis function approximants