Abstract
In this paper the authour constructed the computer system with the object of supporting negotiators for agreement by giving them some rational compromise in practice. Negotiators here mean constituents of a group, each of whom possesses a different opinion and view but has a common goal. Based on the negotiation model in which a compromise takes place in accordance with the change of the negotiators' utility values, this system suggests a means to search for the compromise point utilizing the genetic algorithms, which applied one type of chaotic fluctuations derived from the natural selection theory for generation of the individual. As a concrete example of the simuation, group car buying decision is dealt with. And some rules in the negotiation and the way to search for the compromise point by genetic operation are shown with discussing its utility. It is concluded that this method introduced here has the following two features. (1) It can process operations much faster and needs less memories because it does not have to generate individuals to form concessional trajectories nor to set up the fitness before searching for a compromise point. (2) It is possible to obtain a reasonable compromise point because the chaotic fluctuation is applied to the process of the generation of parent individuals. (3) It is modeled on the growing process of individuals of the natural selection theory and it does works as such practically and regularly: that is, the evolution originated in mutations and thereafter it proceeds as individuals are produced repeatedly. From the above, it can be deduced that this system suggests negotiators by giving a most efficient and suitable solution based on the natural selection theory.
