Multi-Objective Optimization Algorithm and Preference Multi-Objective Decision-Making Based on Artificial Intelligence Biological Immune System

The operating mechanism of the biological immune system is often used for the development of intelligent technology. This research introduces the multi-functional optimization algorithm of the biological immune system into the VR image segmentation, and proposes a multi-purpose VR image segmentation method with more stable and better segmentation performance. In order to combine it with the characteristics of the VR image itself, a complementary feature extraction method combining filters and gray-level symbiosis probability is used. In addition, in order to enable the algorithm to solve the segmentation problem of huge pixel images, the best solution is found through communication and exchange between subgroups. Use the excellent genes of the memory genes as the imported genes and introduce the inferior individuals to strengthen the mining of the best solution of Pareto at the boundary of the impossible field. In order to verify the performance of the algorithm, 3 synthetic texture images and 2 actual VR images are used, 8 constrained targets and 4 unconstrained target benchmark functions are selected to test the optimization function of PCMIOA. Multipoint parallel search uses two different search schemes, local and global. In this way, the domain value of the highest value can be searched globally, and the local best solution can be searched at the same time, realizing the global search mechanism. The relatively satisfactory target value is 98.25, and the deviation between the corresponding solution and the ideal solution is 0.093. The results of the research show that Multi - objective optimization algorithm is an excellent demonstration of the diversity of Pareto-oriented methods and solutions. Compared with the previous prediction methods, this method has higher prediction accuracy and robustness. The choice of decision makers can be taken into consideration, and subjective willfulness can be reduced to make decision results more realistic and reliable.