A Predictive Control System for Concrete Plants. Application of RBF Neural Networks for Reduce Dosing Inaccuracies

In industry, a comprehensive control process is necessary in order to ensure the quality of a manufactured product. In the manufacturing process of concrete, the variables are dependent on several factors, some of them external, which require very precise estimation. To resolve this problem we use techniques based on artificial neural networks. Throughout this paper we describe an RBF (Radial Basis Function) neural network, designed and trained for the prediction of radial in concrete manufacturing plants. With this predictive algorithm we have achieved results that have significantly improved upon those obtained to date using other methods in the concrete industry.