New Adaptive Immune Colonel Mix Algorithm And Its Application Research

In the past, people would be animals, plants and even human beings inspired by the characteristics, have many significant inventions and creative. Along with the development of science and technology, mankind has been not satisfied only from the surface structure of the organisms or unique institution individual to be enlightened, mankind find a new enlightenment from organisms information systems, opened up a new field, such as artificial neural networks, evolutionary computation (genetic algorithm) and the immune system. Which neural networks and evolutionary computation to be widely applied to various fields, the immune system because of its complexity without cause the same attention. Until recently the immune system attracted the attention of domestic and foreign researchers have rapid development, the project application has also been gradually surfaced, Immune Optimization Algorithm emerged.Starting with the basic theories and development process of artificial immune system and immune algorithm, the paper analysis the current development of the immune algorithm, study of the basic immune algorithm, cited several typical immune improved algorithms, analysis of immune algorithm's current strengths and its weaknesses. Research a new immune improved algorithm -- based on immune colonel theory adaptive mix algorithm.Based on immune colonel theory adaptive mix algorithm consists of three parts:(1) The algorithm based on the most widely used information entropy immune algorithm as a framework, information entropy make antibodies at high speed convergence, whilst guarantee the diversity of antibodies, to avoid falling into the trap of local optimizing.(2) The immune clonel theory is an important basis theory of algorithm. Organic immune cells achieve a certain degree of diversity, each antigen invading body, organism choose immune cell that can identify and eliminate the antigen to cloning, activation, differentiation and proliferation, make use of immune response to removal antigen, this is the clonal selection theory. Use of clonal selection theory, according to a certain proportion cloned antibodies group, quickly expand the number of antibodies groups. Then, use the higher crossover rate and mutation rates to increasing the diversity of antibodies groups.(3) The third important theoretical of algorithm is the Boltzmann annealing selection operator of simulated annealing. Via setting the annealing selection operator to update memory library. The operator characteristic is accepting optimal solution. At the same time, still accept deterioration solutions in limit scope. The constant deepening of the process optimization, limite scope decreases gradually. The limited scope of progress through cooling process table (annealing temperature is t) to achieve. Through this mechanism, algorithm can get rid of the local optimal problem, more likely to achieve the overall optimal solution.The paper analyzed the merits of the algorithm in theory, explained in detail the steps of the algorithm, and the mapping of the algorithm flowchart. Advantage of algorithm is: Not a great number of initial population structures, through cloning select operation to maintain the diversity of antibodies, eliminate worse intimity degree of antibodies. This could greatly reduce the computational size, as much as possible to maintain the diversity of antibody population. Update antibody in the process memory library and the next generation antibody group, algorithm use two different updating mechanisms. Boltzmann annealing selection operator is used to update memory library, information entropy and antibody concentration select method is used to update the next generation antibody group.The paper makes use of Visual C++ complete the algorithm function; results of the computation make use of Matlab to image output. In order to test the performance of algorithm, selected four typical more peak functions (Camelback function, Rosenbrock function, Goldstein-Price function and Branin function) to testing. Selected based on information entropy algorithm and the adaptive immune algorithm comparison with algorithm. Through the actual calculations concluded, based on immune colonel theory adaptive mix algorithm than any other optimization algorithm is more efficient, and on the accuracy of optimization has more obvious advantages.In the paper, algorithm application in the two areas: first, establishment optical fiber weighting sensor's output data characteristic equation; second, optimize TFFF system.(1) Use algorithm establish microbend fiber weighting sensor characteristic equationFirst, introduce the theoretical foundation of the optical fiber sensor and new type products of microbend fiber weighting sensor experiment process, build the static and dynamic load experiment structures, detailed description steps of the experiments, the experiment results are discussed. Use based on immune colonel theory adaptive mix algorithm; establish characteristic equations in static and dynamic load experiment structures. In view of the special requirements of practical problems, algorithm was used two coding schemes, trigonometric coding and polynomial coding. The test data were calculated using two encoding schemes. Finally, compare the calculation results of the two coding schemes, trigonometry coding scheme is found more suitable to solve the problem. Compare characteristic equations of two type load experiment structures, static test calculated the characteristic equation has superior quality than the dynamic load test characteristic equation.(2) Algorithm optimtize TFFF systemCombination Jilin province and technological development projects-- Separation of polymer flow field of biological chips research, research algorithm the new application, optimize the TFFF system. Introduce the theoretical foundation of TFFF, analysis optimization problem of TFFF system. Different values of the parametersλof the optimization problem is solved, optimization results of the analysis and comparison, and analysis of the overall impact of the TFFF system. The results of optimize is a guide of TFFF experiment research.Based on immune colonel theory adaptive mix algorithm in the application of the above two areas, it shows that the algorithm has a strong practical value.