Using of fuzzy logic for intelligent agents reasoning for simulating of hepatitis в epidemic process A. V. Tovstik, postgraduate National aerospace university named after N. Ye. Zhukovsky ".KhAI”
In present time hepatitis В is one of most widespread and dangerous viral infections and the problem of its spreading remains actual. This infection features extremely high degree of contagiousness, complicated course of disease and frequent occurrence of severe consequences. The problem is also compounded by the fact that 65-80% of infected by virus of hepatitis В disease have no external clinical manifestations.
In work the multi-agent environment for simulation the hepatitis В epidemic process is described. Experiments with simulation model enable to answer the questions about the influence of certain anti-epidemic measures against incidence level. Nevertheless, epidemic process as a whole depends on each population member separately. The risk of infection for a person is largely dependent on his lifestyle and behavior, which is not constant, but may change with the accumulation of life experience. Therefore, there is need to consider the individual characteristics of each member of population from standpoint of participation in epidemic process. In order to avoid of excessive complication of simulation model proposed use of intelligent agents that make decisions based on fuzzy logic. To solve this problem has been developed the intelligent agent structure that provides presence of certain elements of individuality. Agents have such mental properties as beliefs, desires and intentions (BDI architecture). In order to simulate agents' individuality in decision making process they contain fuzzy logic conclusion mechanism and individual set of membership functions. The agents' subsystem of fuzzy inference has a standard structure and involves two input variables: agent's knowledge about the risk of infection because of certain action and his degree of desire to commit that action, and one output variable: duration of stay in certain area. Thus agents can independently make decisions about their behavior that may expose themselves or others the risk of infection. In addition such structure of agents allows them to reconsider their attitude to certain actions as a result of life experience and interaction with other agents.