Dynamic Clonal Selection Algorithm And Design Of Artificial Immune System

Natural immune system is the identification of biological cells and distinguishes all the external intimate harmful antigens and their own organizations, thus allowing the removal of organisms and maintaining stability,while Artificial immune system makes use of natural immune system as the prototype, taking advantage of its main principles and mechanisms for the various immune models, algorithms and their applications in engineering and science. Further more , Basic artifitial immune algorithms based on natural immune mechanisms such as somatic and network theory inspiration, antigen recognition, cell differentiation, memory and self-regulatory function and so on, have developed into three main types : Negative Selective Algorithm(NSA) Clone Selective Algorithm(CSA) Dynamic Clone Selective Algorithm(DCSA).CSA is a simulation of the evolution of the natural immune system's learning process, and also its optimization processors,while DCSA is a dynamic learning algorithm to expand the traditional CSA under the dynamic data feature to be able to maintain the appropriate detect efficiency. On the mean while, new problem arises from the lack of correct detection rate's stability under the fluctuant outer data context, and this is where the direction of this paper's research and improvement.The improvement of DCSA ,originating from the decisive antigen parts and immune dominant point theory ,combines the immune dominant point evolution and immune dominant point library, with the advanced clone mutation process reverse parsing process and reverse arranging process respectively to improve both the first initial identical correct rate and second identical correct rate , thereby enhances the overall identification algorithm accuracy rate, keeping its stability in a dynamic environment. Finally we construct a corresponding software platform to implement the above algorithms and compare the test results to prove the efficacy of this paper's improvement.