CONTINUED LEARNING IN ARTIFICIAL INTELLIGENCE TECHNIQUES

This book presents an intelligent system with continuous learning, inspired in the biological immune system operation, to perform integrity monitoring in aeronautical structures. This intelligent system is composed of techniques based on artificial immune systems (IAS), such as the negative selection algorithm (NSA) and the clonal selection algorithm (CSA). Artificial immune systems constitute powerful tools suitable for performing complex tasks such as diagnostics. Because it is a stable, reliable and adaptive architecture, it allows the inclusion of the continuous learning module. Thus, the system becomes more efficient as new information is processed and made available. This property is an advantage of EISs, because they allow the possibility of continuous learning. Finally, the proposed algorithm was evaluated using data obtained from a numerical model of an aeronautical structure.