Artificial Neural Network (ANN) Based Process Modeling

Artificial Neural Networks (ANN) based on simulation data from validated models are a useful tool in system design and for process controls. This chapter provides an overview of the use of ANN in problem solving of many real-life engineering system design. It also discusses the effect of the number of hidden layer nodes, learning rate and momentum values on efficient ANN development. Some ANN examples discussed are: prediction of temperature and moisture contents of frankfurters (hot dogs) during thermal processing; prediction of food thermal process evaluation parameters; and prediction of freezing time for food products. The ANN development for meatball cooking is given in detail.