Mobile phone signaling data-based user travel residence behavior identification method

Abstract

The invention discloses a mobile phone signaling data-based user travel residence behavior identification method. The method comprises the following steps of (1) performing cleaning, conversion and segmentation on mobile phone signaling data; (2) by utilizing an unsupervised classification method, setting a plurality of clustering numbers, performing clustering on spatial points in a signaling data unit, and evaluating each clustering result through a clustering scoring index, wherein the clustering with a high score is optimal clustering; (3) obtaining an information set of all candidate residence points and related time in the signaling data unit according to the optimal clustering; and (4) according to a time threshold and the information set of the candidate residence points, performing calculation and screening on residence durations of the candidate residence points, and outputting a spatial position, arrival time and a residence duration of each residence point of each user at each day. The method is simple and convenient in use, has good expandability, avoids interference caused by human objective judgment, and avoids an eclectic phenomenon of identification accuracy due to non-uniform distribution of base stations in urban and suburban areas.

Description

technical field [0001] The invention relates to the technical field of large data, in particular to a mobile phone signaling data-based method for identifying user travel and parking behaviors. Background technique [0002] Understanding the travel status of urban residents is an important consideration for urban planners and traffic planners when planning urban layouts and road networks, including the residence points of citizens traveling and the arrival time and duration of residence corresponding to the residence points. The traditional method of obtaining the travel status of citizens is mainly through the manual method of distributing questionnaires. The main disadvantages of traditional methods are high survey costs, small sample size, high accuracy rate subject to human interference, and low frequency of information updates, which make it impossible for planners to accurately and timely understand the travel needs of urban residents. [0003] With the popularization...

AI Technical Summary

Problems solved by technology

However, the existing methods of extracting user travel status through mobile phone signaling data have the following deficiencies: (1) Due to the non-uniform distribution of mobile phone communication base stations in geographical space, the existing methods need to be manually adjusted repeatedly and observed in experiments. To set the spatial threshold to identify the user’s residence point; (2) Considering that the distribution density of mobile phone communication base stations in the urban area and the suburbs is different, setting a unified spatial threshold leads to accurate identification of the residence point of suburban travel and urban travel There is a trade-off

The above deficiencies make it difficult for existing methods to be directly used by planners and other users, which increases the difficulty of learning and unnecessary workload for users to analyze citizens' travel conditions through mobile phone signaling data, and makes urban and suburban travel stay points Recognition accuracy cannot be balanced

Images