Reliable multi-objective optimal control of batch processes based on stacked neural network models

Abstract

Abstract This paper presents a stacked neural network based multi-objective optimal control method for batch processes. Stacked neural networks not only give better generalisation performance than single neural networks but also provide model prediction confidence bounds. In addition to the process operation objectives, the reliability of model prediction is incorporated in multi-objective optimisation in order to improve the reliability of the obtained optimal control policy. The standard error of the individual neural network predictions is taken as the indication of model prediction reliability. The proposed method is demonstrated on a simulated fel-batch process.