Immune Agent-Based Neural Networks Assembly for Soft-Sensor

Aiming at difficulty modeling of large amounts of industrial process data, a novel soft sensor model based on artificial immune agent-based multiple model Radial Basis Function (RBF) networks is proposed in this paper. The method is to predict the qualities of manufactured products of Crude Oil Tower. In the IMMST-Team system, some biological immune based operation and learning rules which can efficiently cluster the large amounts of data samples and adapt RBF submodels structure and parameter are established. Individuals in the immune system work cooperatively to accomplish the goal of model training. Meenwhile, immune memory and pattern recognition provide high efficiency of predicting. The prediction of dry point of naphtha produced in a practical industrial process is carried out as a case study. The results obtained indicate that the proposed method provides quality prediction with high efficiency and accuracy, which is capable of learning the relationships between process variables measured during the production cycle.