Improved immune algorithm based on a global strategy

Imitating the antibody diversity maintaining mechanism of immune system to realize the global optimization is a target that the immune algorithm try to achieve. Based on the in-depth study of inhibition concentration mechanism, the global optimization characteristic of existing immune algorithm is analyzed, then a global conservation strategy for colony is proposed. Based on the strategy, the improved algorithm is of more outstanding global and fast convergence ability. Simulation is implemented based on Matlab, the algorithm is applied to train a neural network prediction model and it is compared with an existing typical immune algorithm. Simulation results show that the immune algorithm improved by the strategy in this paper is better than the previous algorithms in global evolution, fast convergence and other key indicators.