2003 Volume 68 Issue 568 Pages 93-100
Japanese population is predicted to start decreasing early this century. This inexperienced phenomenon will most drastically appear in local municipalities. In order to establish planning for such municipalities facing this new phase of development, more accurate demand estimation, that is totally different from the conventional empirical method, will become indispensable. Therefore we propose a numerical procedure using a multi-layered cellular automata with the cells of the layers interacting each other. The cell is a population set in the region. One layer is an age composition of the population. The basis of our procedure guesses the population fluctuation by the state transfer relations (local rules) between each cell. In order to verify the validity of the numerical procedure with the multi-layered cellular automata, we analyze Kagoshima City as a sample. Firstly, a function of state transfer is defined from the characteristic of the population fluctuation using equations with various computational parameters. The state transfer relation is decided from.the measurement geography. The computation results are evaluated by their chronological conformity in comparison with the statistics data. Finally, the parameters that showed the best conformity are used for the future estimation of the population in urban and regional planning with a charm project. Moreover, the tendency for the population fluctuation of the local municipalities is investigated by analyzing the value of the computation parameter.
