Solving discrete time-varying complex division using recurrent neural network with disturbance suppression

Discrete time-varying complex division has perplexed researchers in many fields for many years. How to solve such problem has actually been discussed for a long time. In recent years, lots of researchers have studied the recurrent neural network (RNN) model for solving time-varying problems. As an extended mathematical problem, we introduce the discrete time-varying complex division from the traditional division and reducing disturbance by using the 5-instant discretization formula type advanced discrete time-varying recurrent neural network (5IDF-type advanced DTRNN) model. First, the continuous time-varying complex division is proposed based on the traditional Euclid division. Then, inspired by achievements of previous researches, we equivalently transform the continuous time-varying complex division into a simple matrix-vector equation in the continuous time-varying environment. Accordingly, the discrete time-varying complex division is transformed into the discrete time-varying matrix vector equation. In addition, we also show the 5-instant discretization formula and propose the 5IDF-type advanced DTRNN model. Numerical experiments further prove that validity and superiority of such 5IDF-type advanced DTRNN model.