Vinyl Acetate Polymerization Rate Prediction Based on FOA GNN

The vinyl acetate polymerization rate is an important quality index in the production of polyvinyl alcohol. However, for it can not be measured online, the polyvinyl alcohol quality can not be controlled effectively. As a novel meta-heuristic and evolutional algorithm, the fruit fly optimization algorithm (FOA) has several merits such as having few parameters to be adjusted and able to achieve global optimum. Therefore, to improve the prediction performance, this paper proposes a grey neural network prediction mode that uses FOA to optimize the "whitening" parameters of this grey neural network. Simulation and experimental results show that the grey neural network prediction model combined with FOA (FOA_GNN) is an effective method to predict the vinyl acetate polymerization rate, and it outperforms other alternative methods, namely single the grey neural network model (GNN), the adaptive compete genetic neural network prediction model (ACGA) and radial basic function neural network model (RBF).