Research On Immune System Model Based On Agent

In the 1990s, with the progress of biology, people’s understanding of biological systems is deepening. More and more computer scientists and engineers realized that biological systems can be used as an important method to solve many complicated problems. So people’s passion for studying biological systems is also rising. As one of the four large information systems of human body, the immune system is getting more and more attentions from computer scientists. Through the joint efforts and cooperative research of many experts, people’s understanding of the immune system has reached a certain level.Many mechanisms of the immune system, such as self-nonself, identification fault tolerating and robustness to interference, have been applying in engineering field. But, as a whole, people’s understanding of the immune system is not enough. Even many current theories of immune system are not absolutely right. Therefore, when studying the immune system, On the one hand, we try to find new immune phenomenon and the mechanism, on the other hand, we should verify the existing theory of immunology which is controversial. Based on the existing theory of immunology, we put forward a simulation model of the immune system, then try to analyze and verify the controversial theory of immunology. It has certain feasibility and applying value.We all know, with the help of the cooperation between the immune cells, the immune system can realize its function. And multi-Agent system can realize its function through the cooperation between the each Agent. There is a lot of similarities between them. Therefore, in this paper, according to the ideas of Agent, by defining corresponding Agents for each cell, we can establish a multi-Agent system to simulation the immune system,. Finally, we also research and analysis the simulation model and the results of it.The immune system is complicated. In order to better understand the immune system and more accurately define the Agent of the simulation, in this paper, we analyze the immune system and understand its requirement. We use the UML to draw structure diagram, class diagram, sequence diagram, etc.. Then we use those diagrams to describe the static structure and dynamic behavior model of the immune system. At the same time, to ensure the correctness and the unambiguity of the model, We converted the established UML system into B machine. And with the help of B machine, we can ensure the accuracy of the model.The key to building a multi-Agent system is defining the Agent and rules between differentAgents. The immune system contains many kinds of cells and moleculars, and the interactions between them is diverse. It is impossible and unnecessary to model and simulate all the details of the immune system. Thus, we chose several typical cells, then we use some Agents which have been predefined rules and behavior to simulate these cells. Finally, through the binary string pattern matching method to simulate the process of immune recognition.Finally, we use the Java programming language to turn the multi-Agent system into a computer program which is executable. In addition, we apply some related parameters of the influenza virus to this system. Through the analysis of simulation results we can prove that the dynamic behavior of model is similar to the basic laws of biology, and the simulation model is effective. At the same time, with the help of undetermined coefficient method, we discuss some key parameters.On the whole, the immune system is complicated. It is impossible to describe it unifiedly or establish an accepted theory model. In this paper, we use the Agent computer model and the computer simulation method to study the macroscopic properties of the immune system and some important details. All the work provide powerful basis to establish accurate perfect model of the immune system. Meanwhile, a new method to research the immune system is obtained.