Internet rental bicycle demand characteristic evaluation method and storage medium

Abstract

The invention provides an internet bicycle renting demand feature evaluation method and a storage medium, and the method comprises the steps: calculating bicycle borrowing and returning normalization indexes of a station or a region according to user use data; constructing a clustering analysis model according to the bicycle borrowing and returning normalization index, and carrying out clustering analysis; and according to a clustering analysis result, classifying the stations or the areas according to a use mode. According to the invention, clustering analysis is carried out based on the normalized index of the vehicle borrowing and returning data of the user, the use characteristics of the vehicle are represented, and discreteness errors caused by the use of NAB data in the conventional technology are avoided. Preferably, the user use data is based on continuous time and comprises real-time bicycle borrowing and bicycle returning data, so that errors caused by mutual offset of bicycle borrowing and bicycle returning in a conventional method can be avoided.

Description

Technical field [0001] The present invention relates to the use of demand feature evaluation techniques for Internet rental bikes, and more particularly to an Internet rental bicycle demand feature evaluation method and a storage medium. Background technique [0002] At present, the Internet rental bicycle (including public bicycles, sharing bicycles, electric single cars, etc.) has no extensive approval, evaluation methods. For a network rental (electric) bicycle - public bicycle, researchers and operators can borrow the number of vehicles by taking the NAB data of the rental pile (Number of Available Bike, if the number of vehicles is 20 cars At some time, 6 vehicles can be borrowed, and NAB is 6 / 20) observes the borrowing of the vehicle station, thus estimates the use of demand. This method is usually analyzed in units of a site. [0003] The existing analysis method has the following issues: [0004] First, NAB data is mostly static data, based on discrete timestamp, may brin...

AI Technical Summary

Problems solved by technology

[0005] 2. The feature evaluation method based on NAB data is difficult to eliminate the error of operation scheduling: for network rental (electric) bicycles, operators usually carry out manual scheduling, and transport the vehicles to areas with more demand but less vehicle supply (such as , transport the public bicycles to the public bicycle stations around the rail station in the morning); or conversely, transport too many vehicles away to make room for users to return the vehicles

However, synchronously prompting NAB data when dispatching vehicles requires more advanced data coordination capabilities, and there are still major difficulties

[0006] 3. Lack of effective mining of user usage data: Conventional demand assessment technology has little mining of user usage data (borrowing and returning car data), and it is not suitable to discover user usage characteristics

[0007] 4. Lack of evaluation methods for demand characteristics of pile-free travel methods such as shared bicycles and motorcycles: conventional evaluation methods are mainly aimed at public bicycles with piles, and the analysis is performed on a site basis

For the current fast-growing dockless shared bicycles and motorcycles, because there is no NAB data, it is impossible to evaluate and analyze demand characteristics

Images