ROMJIST Volume 23, No. T, 2020, pp. T57-T77
Febina CHRISTUDAS, Angeline VIJULA DHANRAJ System Identification Using Long Short Term Memory Recurrent Neural Networks for Real Time Conical Tank System
ABSTRACT: An essential prerequisite for designing model-based controllers is an accurate mathematical model. System Identification is a classical approach for getting a mathematical model from experimental data. Neural Networks are extensively deployed for the identification of systems as they are efficient function approximators. This paper proposes an intelligent technique for modelling a nonlinear Conical Tank System (CTS) using neural networks. Long Short Term Memory Recurrent Neural Networks (LSTM-RNN) are used for modelling the real- time CTS using input-output data. It is shown that LSTM-RNN models are effective in modelling in comparison with empirical models.KEYWORDS: Artificial neural networks, LSTM, nonlinear system, predictive models, RNN, system identificationRead full text (pdf)