Research of an immune clone selection algorithm and its application in heating furnace state recognition

Abstract

The immune clone selection algorithm presented in this paper is inspired by the mechanism exhibited in biological immune response. It is made up of four procedures: antigen-antibody matching, clone proliferation, mutation and memory cell selection. It is a self-adaptive learning system, antibody diversity coming from clone and mutation process, and which make the system has self-adaptive capability. It accumulates knowledge from antibody memory cell selection process. It has few parameter and easy to stable. We discuss the application of the algorithm on state recognition of the heating furnace, the results show that the model has good ability on pattern recognition and data compress.