The Artificial Immune Endocrine Algorithms Based Artificial Intelligence Conformity

In recent years, inspired by the biological mechanism, the information processing system has become the important content of the artificial intelligence. Biological immune system has some intelligence characteristics like memory, parallel and auto-adaptation, so at present it is a hot issue to develop the intelligent computation and its application. However, the endocrine system, as one of the three natural regulation systems (neural, endocrine and immune systems), also includes a great deal of information processing system. In this thesis, immune endocrine system and algorithms are exploited, which base on the principle of homeostasis, immune memory, and antibody recognition and hormone regulation. The main contents are as follows:(1) Based on the principle of the homeostasis, a new artificial intelligence conformity (AIC) theory is proposed. Through the further analysis for the homeostasis mechanism (one of the three biology regulation systems), and the characteristics of the three systems mutual relations, the information molecule, information storage and memory, positive and negative feedback mechanism, we propose an artificial intelligence conformity concept and principle, which also provides a new research idea on biological intelligence algorithm and its application.(2) Based on the principle of the immune response and the hormone regulation, a new artificial immune and endocrine algorithm (AIEA) is proposed. According to the adjustment mechanism of hormone for the immune system, the concept of the immune body density and the hormone density is introduced, and the individual density is measured by the vector distance between the fitness of the individual. The hormone density and the individual sufficiency can drive the evolution process of community, which guarantee the algorithm rapid convergence and avoid precocity effectively.(3) The algorithm research of the artificial immune and endocrine system. A set of functions are chosen to test its performance; moreover, this algorithm can solve typical combination optimization problems: the TSP questions in 31 cities are investigated, and the result shows that this algorithm has good rapidity and overall situation astringency in combination optimization problems.(4) Based on the hormone regulation principle of endocrine system, a novel artificial hormone regulation algorithm is presented. Through deeply analysis on the functions, characteristics and regulation principle of hormone, the every stage in endocrine and regulation processes of hormone is simulated from perspective of molecular dynamics, and the mathematic model is founded according to the useful mechanism. At last, the algorithm was proved to be successful in the mathematic simulation.