Perception-Based -Norm Minimization Approach for Nonlinear System Identification in GGD Noise

摘要

Nonlinear system identification is an important and fundamental problem in many practical applications. It becomes more challenging when the noise is non-Gaussian. Inspired by the cognitive dynamic system concept, we propose a perception-based -norm minimization approach for nonlinear system identification in generalized Gaussian distribution noise environments. Volterra model is utilized to describe the nonlinear system. The proposed cognitive algorithm incorporates a closed feedback loop between perceptions and actions to the environments. Computer simulations have been carried out to illustrate the effectiveness of the proposed method.