基于噪声处理的网络搜索和QCR-HHT模型的九寨沟客流量预测

Tourist Flow Forecasting in Jiuzhaigou Valley Based on the Search Engine with Denoising and QCR-HHT Model

客流量预测可以弥补强周期性和波动性客流冲击给景区和游客造成的影响,使有限的旅游资源提前得到合理调度和配置.在考虑网络搜索噪声的基础上,建立QCR(Query Chain Retrieve)搜索词链和HHT的网络搜索数据预测模型,对九寨沟旅游日客流量进行预测.通过对比时间序列模型、未经噪声处理的网络搜索预测模型和BP神经网络发现,QCR-HHT拟合效果最佳,对九寨沟客流量的预测精度显著提高.使用考虑噪声的QCR-HHT网络搜索预测模型,能够更准确地对旅游客流量进行预测,便于景区和管理部门制定更加高效准确的决策.

The forecast of tourist flow can make up the impact from the strong periodic and volatile passenger flows on the scenic spots and tourists,so the limited tourism resources can be reasonably scheduled and allocated in advance.With the noise of the search engines data taken into consideration,this paper proposes to build a novel prediction model with QCR(Query Chain Retrieve)and HHT to forecast the daily tourist flow of Jiuzhaigou.It is found that QCR-HHT prediction model is the best and the prediction accuracy is remarkably improved,compared with the traditional regression model,ARMAX model with search engines data and BP neural network model.The use of the QCR-HHT prediction model with denoising is of great help to forecast the tourist flow more accurately,enabling the management departments of scenic spots to make more efficient and accurate decisions.