Adaptive online multi-phase neuro-identification method using virtual system generation

An adaptive multi-phase model identification method is proposed. The first phase identification, which is described as a real neuro-identification, is designed for estimating a coarse model while the second phase identification, described as polishing virtual neuro-identification, is utilized for determining a fine model. The presented approach utilizes the well-known backpropagation neural network. A remarkable characteristic is that virtual signals are artificially generated and virtual model identification is also performed using the newly generated series. The complementary approach, based on real and virtual model identification, could be utilized as an efficient model identification. Experimental results are given to verify the proposed approach