Neural Network Predictive Control of Systems with Faster Dynamics using PSO

Abstract

An effective application of Model Predictive Control by means of a multi-layer feed forward neural network as the nonlinear model of the process is discussed. The main draw back in the use of Nonlinear programming for optimization is the complexity in the calculation of Hessian matrix and its inverse. There are many derivative free evolutionary algorithms, inspired by biological evolution. In this paper the optimization problem is solved using particle swarm optimization (PSO). Simulation results show convergence to a good solution within fewer numbers of iterations which makes it suitable for real time applications with faster sampling.