论文部分内容阅读
为了提高参数投影寻踪回归(parameter projection pursuit regression,PPPR)模型对城市客运量的预测精度,基于cat映射、高斯分布和精英局部搜索对加速遗传算法进行改进.提出了新的混沌加速遗传算法(new chaos accelerating genetic algorithm,NCAGA),用于对PPPR模型的最佳投影方向a的优选.建立了在外层优化岭函数个数M的同时,内层利用NCAGA优化最佳投影方向a的NCAGA-PPPR混合优化城市客运量预测模型,结合某市统计资料进行了仿真预测.结果表明该方法的预测精度优于BP神经网络模型、传统PPR模型和基于加速遗传优选的PPPR模型,平均绝对相对误差小于3.1%,提高了城市客运量的预测精度,可有效应用于城市客运量的预测.
In order to improve the prediction accuracy of PPPR model on urban passenger traffic volume, the accelerating genetic algorithm is improved based on cat mapping, Gaussian distribution and elitist local search, and a new chaos acceleration genetic algorithm ( NCAGA), which is used to optimize PPPR model’s optimal projection direction a.With the optimization of the number of outer ridge M function, the NCAGA-PPPR The results show that the prediction accuracy of this method is better than BP neural network model, traditional PPR model and PPPR model based on accelerated genetic optimization, the average absolute relative error is less than 3.1 %, Improve the prediction accuracy of urban passenger volume, which can be effectively applied to the prediction of urban passenger volume.