人口系统迁移过程的参数辨识及其仿真

On the Time-Varying Parameters Identificationn of Migration Process in the Population Dynamics and Its Simulation

本文以系统辨识为工具，提出了迁移过程定量研究的一种方法，即运用随机牛顿法于时变系统中以辨识模型迁移表参量，同时结合卡尔曼滤波技术辨识迁移过程中另一参数──总和迁移率，将这两者融为一体，形成了辨识迁移模型参量的递推算法．本文又借助微分方程的稳定性理论讨论了净迁移情形下的辨识算法的收效性，给出算法收敛的一个充分条件．

In this paper the identification algorithm is given for the time-varying parameters of the migration process in the population dynamics.The parameters in model migration schedules are identified by the stochastic Newton method and total migration rate by Kalman filtering.We connected together the stochastic Newton method with the Kalman fitering to form a recursive algorithm.And it is studied the convergence problem of the algorithm in the case of net migration with the aid of the stability theory of differential equations.