A vehicle trajectory and mileage prediction method based on historical data

Abstract

The invention discloses a method for predicting automobile driving trajectory and mileage based on historical data, using GPS to obtain historical driving trajectory data of automobiles and forming a historical trajectory database; performing starting point clustering and key point mining on historical driving trajectory; constructing a prediction model ;At the beginning of the car trip, the current starting point location information is obtained through the GPS device, and the travel trajectory and mileage of this trip are predicted according to the historical trajectory database; the current driving trajectory is identified based on the key points during the driving process, and the candidate predictions are updated in real time every time the key points pass Trajectory, until all the key points of the driving trajectory are passed and the end point is reached, the route prediction and mileage prediction of this driving are completed, and the driving trajectory is recorded in the historical driving trajectory database; this method is the rational planning and energy management of the power battery of the car The optimization of the strategy lays the foundation, and at the same time helps to optimize the opening moment of the engine, and contributes to the improvement of power and economy.

Description

technical field [0001] The invention belongs to the field of automobiles, and mainly relates to the fields of automobile trajectory data mining and trajectory prediction, in particular to a method for predicting automobile driving trajectory and mileage based on historical data. Background technique [0002] Trajectory prediction is to use GPS technology to dynamically predict the future location of mobile users based on the user's current location information, historical trajectory information and other auxiliary information. Most of the research conducted is based on predicting the user's trajectory based on the user's historical trajectory information. Trajectory prediction methods can be roughly divided into single-user and multi-user categories. At present, some scholars propose destination prediction methods, mostly based on large-scale historical travel data of multiple users, and need to use the road network, which is inconvenient for single users. It is applied to ...

AI Technical Summary

Problems solved by technology

Trajectory prediction methods can be roughly divided into single-user and multi-user categories. At present, some scholars propose destination prediction methods, mostly based on large-scale historical travel data of multiple users, and need to use the road network, which is inconvenient for single users. Applied to vehicle navigation equipment for real-time driving trajectory prediction, and the prediction accuracy is not high

In addition, the prediction of the vehicle's driving trajectory and mileage mostly depends on the driver's frequent input of destinations on the terminal, combined with GPS and electronic maps to give the prediction of the driving mileage. Although this method can also identify the driving trajectory and estimate the driving mileage, However, relying on the driver's complex operations and third-party electronic maps, it is not intelligent enough to dynamically predict the driving trajectory

Images