Research On Parallel Immune Algorithm Based On Fine-grained Model With GPU-Acceleration

At present immune system is the most subtlety complex system that body withstands damages from external environment as we all know, the immune system is like a well-trained elite troops and protect body health. Immune cells would generate memory to intruder, when the same antigens invade body next time, the antibodies will perish the antigens according to memory. This is the fundamental principle of immune vaccine inoculation. Artificial immune system as a new research of artificial intelligence can be a means which is further enrich the study of artificial intelligence according to the immune system of biological information processing mechanism, in other words, biological immunology promotes the development of artificial intelligence research.The characteristics and function mechanism of immune system have an important inspiration and reference for solving many complex problems in engineering applications, immune algorithm is put forward an intelligent optimization method according to biology immune mechanism, it has proven to be a powerful global optimization method and shows great potential in practice. However, it still needs plenty of computing time when it processes much data and when large-scale complicated work is involved in which math modeling and optimization are highly demanded, whereas parallel IA comes into being and becomes a hot spot since it can reduce working-out time dramatically.After understanding the development trend of the current IA at home and abroad, aiming at the problem that IA can consume plenty time when settlement of large-scale optimization problems, we raised a fine-grained IA algorism based on GPU acceleration, we make use of CUDA to the calculating process of working-out into parallel process of thread block As achieving a good optimization effect, it also increases the IA in the fine-grained parallelism, speeds up its running and provides ordinary user with a feasible IA solution.Firstly, we recall the immune algorithm research background, the generating process, current state and trend of IA, and we explain the significance that we make use of GPU to research IA and IA modeling.Secondly, introducing the development of GPU and illustrating general purpose computation and its application, we have given a brief overview about NVIDIA CUDA execution model, storage model and communication model. Thirdly, describing the biological mechanism of immune algorithm in order to better understanding IA model.Finally, aiming at the existing problems such as heavy calculation and low speed in optimizing problem this paper raise a parallel immune algorithm based on fine-grained model with GPU-Acceleration., taking advantage of GPU parallelism, converting the recursive into concurrent computation, then the algorithm increase speed evidently. Meanwhile, the comparison of CPU and GPU time consuming is showed in this paper, the result shows that the new algorithm is feasibility and practicality.