Experimental Results with Simplified Algorithm of Estimation the Parameters of a Dynamical System

Abstract The paper deals with the three estimation methods of parameters, i.e. non-recursive least-squares method, recursive least-square method, and Kaczmarz's projection algorithm. These methods were experimentaly tested on estimation the parameters of a fast DC motor. The tests showed that the simple projection algorithm gives a little less precise estimates in comparison to the least-squares methods, but is significantly faster. This algorithm can be successfully applied in those situations where occur rapid changes in the description of a system dynamics, what is typical for example in robotics. It can be useful in some applications of indirect adaptive control schemes.

[1]  Mohammed M'Saad,et al.  Parameter estimation aspects in adaptive control , 1991, Autom..

[2]  Carlos Canudas de Wit,et al.  Adaptive friction compensation in DC motor drives , 1986, Proceedings. 1986 IEEE International Conference on Robotics and Automation.

[3]  P. Kumar,et al.  Convergence of adaptive control schemes using least-squares parameter estimates , 1990 .

[4]  Rafael Kelly,et al.  A Linear-State Feedback Plus Adaptive Feed-Forward Control for DC Servomotors , 1987, IEEE Transactions on Industrial Electronics.