Stability Analysis in Normalized Right Coprime Factorization Using Neural Network

Currently, control strategies based on right coprime factorization (RCF) have been applied into numerous fields. However, the RCF structure mainly focuses the input-output stability, which is thought to be potentially insufficient while considering complicated state-space realized systems. In this work, the stability analysis for input-output and states in normalized right coprime factorization (NRCF) feedback system will be accomplished. The NRCF of given unstable plant is conducted using neural network, aiming at derivation of approximating solution to Hamilton-Jacobi-Bellman (HJB) equation. In addition, the stability after NRCF is analyzed, which is also proved applicable to stable feedback system. The simulation results suggest desired stability while using the proposed approach.