Lane changing intention identification method based on LSTM under multi-source exponential weighting loss

Abstract

Aiming at the problems that a data source is single, a sequence model difficultly captures a lane changing intention in a long sequence range and long-term dependence exists in lane changing intentionidentification, the invention provides a long-term short-term memory network vehicle lane changing intention identification model under a time information weighting index loss function. The method comprises the steps: firstly, conducting a highway driving experiment through a driving simulation cabin and an eye tracker, and collecting vehicle operation data and driver eye movement data; constructing a vehicle lane changing intention identification model in a highway environment based on an LSTM structural unit, and optimizing the model weight through a proposed index loss function based on time information weighting; and finally, verifying the proposed model by using the vehicle operation data and the driver eye movement data and comparing the proposed model with other models, wherein the lane changing identification accuracy of the proposed model is 96.78%, the precision is 95.72%, the recall rate is 95.83%, and the F1 value is 95.73%. The LSTM network has good resolution capabilityfor a long-sequence lane changing intention identification process, and the proposed loss function has a good effect on model weight optimization.

Description

technical field [0001] The present invention relates to the field of driving safety of motor vehicles, in particular to a lane-changing intention recognition method based on LSTM under multi-source exponential weighted loss. Background technique [0002] As a common and common driving behavior, vehicle lane changing always affects the safety and efficiency of surrounding vehicles, and even causes loss of life and property. Among the accidents caused by lane changes, the number of accidents caused by driver's misjudgment accounts for 75% of the total. It can be seen that the study of driver's driving behavior and lane-changing intention has a positive impact on road safety, travel efficiency and the autonomous decision-making behavior of future unmanned vehicles. [0003] In the modeling research of driving intention, the model input is mostly based on a single data source such as the state and position information of the vehicle, surrounding vehicles, or driver's visual inf...

AI Technical Summary

Problems solved by technology

However, the distance between the intention recognition moment and the lane change execution point is short. When the model determines that the lane change intention is inappropriate, the driver’s reaction time is short, and there is a lag in operation, which has certain safety hazards.

Images