Research on passenger flow prediction of tourist attractions by integrating differential evolution and GWO optimization

Abstract

With the rapid development of tourism, the rapid growth in the number of tourists caused by the imbalance of the passenger flow carrying capacity of tourist attractions, tourist crowding, overloading, and other problems caused by the frequent occurrence of safety accidents, to the tourist attractions has caused a huge negative impact. For this reason, this paper constructs a model for predicting passenger flows at tourist attractions based on the GWO algorithm. Optimizing the model involves using a differential evolutionary algorithm following a feature study with the Gray Wolf Optimization (GWO) algorithm. Then, for the problems that are prone to occur in the fusion of the GWO algorithm and DE algorithm for solving nonlinear systems of equations, a combined GWO-DE-SVM model is proposed to realize the accurate prediction of tourist attractions’ passenger flow. On this basis, the prediction effect of the GWO-DE-SVM combination model is examined. The loss values of the model in the training set and test set of this paper are around 0.034 and 0.029, respectively, with the lowest average error of 2.524% among all the models. The passenger flow and the total tourism revenue of W tourist attraction in the coming year are successfully predicted in the practical application, which is estimated to be 681 million and 106.88 million yuan, respectively. Million and 106.88 million yuan are the respective amounts. And two peaks of tourist attractions in W were predicted in May and October. This study provides a scientific basis for the management of scenic spots to prevent tourist crowding.

Keywords

• gwo algorithm
• differential evolutionary algorithm
• svm
• prediction model
• tourist traffic flow
• 97p10