基于蚁群优化支持向量机模型的公路客运量预测

针对公路客运量预测难以建立精确预测模型的问题,文章引入基于蚁群优化的支持向量机算法对公路客运量进行预测。由于支持向量机的预测精度很大程度上取决于参数的选取,因此利用蚁群算法来优化其训练参数的选择,以得到优化的支持向量机预测模型,利用其对小样本及非线性数据优越的预测性能进行公路客运量的预测。以北京市的数据作为应用算例,并与BP神经网络及传统SVM的预测结果进行对比分析。实验结果表明,基于蚁群的支持向量机模型的预测精度更高,误差更小,可以更有效地对公路客运量进行预测;也说明利用蚁群算法进行支持向量机参数优选的方法是可行有效的。

Passenger traffic volume forecasting based on support vector machine model optimized by ant colony algorithm

Aiming at the difficulty in setting up an accurate forecasting model for the passenger traffic volume,a new prediction method is proposed by integrating support vector machine(SVM) and ant colony algorithm.Since the accuracy of the SVM depends on the parameter choosing to a great extent,the ant colony algorithm is used to determine the training parameters of this model so as to obtain the optimized SVM forecasting model.Then the passenger traffic volume is forecasted by utilizing the excellent forecasting performance of the model in small sample and nonlinear data of the SVM.Taking the data of the city of Beijing as the experimental data,the prediction results are compared with those of BP neural networks and traditional SVM.The experimental results indicate that the SVM forecasting model based on ant colony algorithm can forecast the passenger traffic volume effectively with greater accuracy and fewer errors.The ant colony algorithm is a feasible and effective method for optimizing parameters of the SVM.