An ADALINE neural network with truncated momentum for system identification of linear time varying systems

This paper presents a new version of online identification method for linear time varying systems based on the ADaptive LINear Element - ADALINE (Widrow and Lehr, 1990 ) neural network with a truncated momentum term added for the purpose of reducing fluctuation while sudden parameter change happens thus offers a smoother transition in tracking the parameter. It is well known ADALINE is slow in convergence which is not appropriate for online application and identification of time varying system. To speed up convergence of learning and thus increase the capability of tracking time varying system parameters, our previous work added a momentum term to the weight adjustment. While the momentum does speed up convergence, it also shows over-shooting or oscillating and also tracks noise closely. To help to reduce this effect, we propose a truncated version of the momentum term to track variable parameters better and track noise less. Simulation results show that the proposed method provides indeed fast yet smoother convergence and better tracking of time varying parameters.