An adaptive learning system yielding unbiased parameter estimates

A learning system involving model reference adaptive control (MRAC) algorithms is studied in which the Lyapunov function and its associated time derivative are simultaneous quadratic functions of both the position tracking error and the parameter estimation error. An implementation method is described which expands results of Repperger et al. (1999). Initially it appears that the condition of persistent excitation need not be explicitly satisfied, however, this condition is actually implicit in the requirements for a solution.