IMPLEMENTATION OF DIGITAL FILTERS USING NEURAL NETWORKS WHEN THE OBJECT MOVEMENT DYNAMICS ARE NOT KNOWN
Zinovii Liubun, V. Nesterenko, V. Mandziy, O. Karpin · Electronics and Information Technologies · 2022
Considered: Neural network-based algorithms to obtain an optimal digital filter. Recurrent neural networks can be used to implement an adaptive digital filter considering the characteristics of both interference and useful signal. The developed program trains recurrent neural networks and as a result obtains the coefficients of recursive digital filters. The efficiency of the proposed algorithm is confirmed with numerical experiments results. Keywords : digital filter, recursive filters, recurrent neural networks