Online hailing traffic flow multi-scale feature analysis method based on quantum walk

The invention discloses an online hailing traffic flow multi-scale feature analysis method based on quantum walk, which comprises the following steps of: firstly, simulating a dynamic evolution process of online hailing cars in an urban space by using quantum walk, and generating a multi-scale probability mode evolving along with time when the online hailing cars appear at different positions according to the dynamic evolution process; secondly, screening the multi-scale probability mode according to the spatial-temporal heterogeneity of the online booked traffic flow; with the observed online hailing traffic flow as a boundary condition, screening out multi-scale probability modes of online hailing cars at different positions based on stepwise regression; and finally, constructing a mapping conversion mechanism between the multi-scale probability mode and the network-booked traffic flow, thereby realizing multi-scale feature analysis of the network-booked traffic flow. According to the method, the multi-scal