Modelling and optimisation control of polymer composite moulding processes using bootstrap aggregated neural network models

This paper presents using bootstrap aggregated neural networks for the modelling and optimization control of reactive polymer composite moulding processes. Neural network models for the degree of cure are developed from process operational data. To improve model generalization capability, multiple neural networks are developed from bootstrap re-samples of the original data and are combined. Optimal heating profile is obtained by solving an optimization problem using the neural network model. The proposed method is applied to both simulated data and real industrial data.