Method and device for predicting entry of vehicle into logistics park, storage medium and terminal

Abstract

The invention discloses a method and device for predicting the entry of a vehicle into a logistics park, a storage medium and a terminal. The method comprises the steps of obtaining a driving track point set of a to-be-predicted vehicle, and loading preset road network data; matching road data corresponding to tracks in the driving track point set from the road network data; the road data comprising the road grade of the road where each track is located and the distance between the located roads; sequentially inputting the road grades of the roads where the track points are located and the distances between the roads where the track points are located into a preset sliding window algorithm, and extracting track sequence features of the track points; and inputting the track sequence features of the track points into a pre-trained prediction model, outputting a probability value of entering the logistics park, and determining whether the vehicle enters the logistics park or not accordingto the probability value. Therefore, by adopting the embodiment of the invention, the behavior that the vehicle enters the park at a certain place can be effectively predicted according to the massive vehicle track data and the road network information, the prediction accuracy is improved, and the management efficiency of vehicles in the park is further improved.

Description

technical field [0001] The invention relates to the technical field of data mining, in particular to a method, device, storage medium and terminal for predicting vehicles entering a logistics park. Background technique [0002] At present, the development of the logistics industry is in full swing, resulting in the birth of a large number of trucks. As the main force of land transportation, trucks undertake a large part of the cargo transportation volume. Therefore, it is very necessary to achieve efficient cargo transportation. How to predict when realizing efficient cargo transportation Whether the outbound vehicle enters the logistics park, arranges goods for the vehicle in advance according to the prediction results, and manages the route parking and parking position management of the vehicles in the park are necessary conditions for efficient cargo transportation. [0003] In the existing technical solution, when predicting whether the vehicle will enter the logistics p...

AI Technical Summary

Problems solved by technology

[0003] In the existing technical solution, when predicting whether the vehicle enters the logistics park, it is predicted by creating an electronic fence, according to the vertices of the convex polygon electronic fence area Construct a reference line, and further judge whether the vehicle enters the polygonal electronic fence area by judging whether the vehicle is located on the same side of all reference lines. Since the detection of park entry in the existing scheme is mainly based on manually demarcated electronic fences, it is created by manual labeling. Electronic fence, high labeling cost, incomplete information and information lag, thus reducing the prediction accuracy

Images