Novel Recurrent Neural Network for Time-Varying Problems Solving [Research Frontier

Abstract

By following the inspirational work of McCulloch and Pitts [1], lots of neural networks have been proposed, developed and studied for scientific research and engineering applications [2][18]. For instance, one classical neural network is Hopfield neural network (HNN) which was proposed by Hopfield in the early 1980s [2]. Another classical neural network is based on the error back-propagation (BP) algorithm, i.e., BP neural network, which was developed by Rumelhart, McClelland and others in the mid-1980s [3]. Generally speaking, according to the nature of connectivity, these neural networks can be classified into two categories: feedback neural networks (or termed recurrent neural networks, RNN) and feed forward neural networks. Recently, due to the in-depth research on neural networks, the artificial neural-dynamic approach based on RNN has been viewed as a powerful alternative to online solution of mathematical problems arising in numerous fields of science and engineering, such as matrix inversion in robots redundancy resolution (as an essential part of the pseudoinversetype solution) [16], [18].