基于粒子群优化算法的AR模型参数估计

AR model parameter estimation based on particle swarm optimization algorithm

利用粒子群优化算法对模型AR(P)参数进行优化,提高了模型的预测精度。采用AIC准则判断出模型的最佳阶数。该估算方法优于传统的基于最小二乘估计和基于灰色理论估计,为此类模型的参数估计提供了一种新思路。

Using the particle swarm optimization algorithm to optimize the AR （P） model parameters, improve the prediction accuracy of the model .The AIC criterion can determine the best order number of the model.This estimation method is superior to the traditional estimation based on least squares estima-tion and gray theory estimation .The estimated parameters provides a new way of thinking for the parame-ter estimation of this kind of model .