Nonlinear dynamic system identification using Legendre neural network

We propose a computationally efficient Legendre neural network (LeNN) for identification of nonlinear dynamic systems. Due to its single-layer architecture, the LeNN offers much less computational complexity than that of a multilayer perceptron (MLP). By taking several plant models of increasing complexity and with extensive simulations we have shown superior performance of the LeNN-based plant model in comparison to that of an MLP model in terms of estimated output, mean square error (MSE) and computational complexity, in presence of additive noise.