The invention provides a driving behavior analysis method based on a weighted cost function. The method comprises the following steps: acquiring initial feature performance data of a plurality of driving tracks; preprocessing the initial feature performance data to obtain feature performance data; dividing the driving track into a plurality of decision period segments according to the duration of different driving events; according to the characteristic performance data, a cost function representing the degree of attention of a driver to comfort, safety and rapidness in the driving process is constructed; processing the characteristic performance data corresponding to a plurality of decision period segments based on the cost function and an entropy weight method to obtain a weight coefficient corresponding to the characteristic performance; clustering the different driving tracks based on a k-means method and weight coefficients corresponding to the characteristic performance of the different driving tracks; therefore, the drivers corresponding to different driving tracks are clustered.
本发明提供一种基于加权代价函数的驾驶行为分析方法,所述方法包括:采集若干行驶轨迹的初始特征性能数据;对初始特征性能数据进行预处理,得到特征性能数据;将所述行驶轨迹根据不同的驾驶事件的持续时间划分若干决策周期段;根据特征性能数据构建表征行车过程中驾驶人对舒适性、安全性和快捷性重视程度的代价函数;基于所述代价函数、熵权法对若干决策周期段对应的特征性能数据进行处理,得到所述特征性能对应的权重系数;基于k‑means方法、以及不同行驶轨迹的特征性能对应的权重系数对不同行驶轨迹进行聚类;从而实现对不同行驶轨迹对应的驾驶人员进行聚类。
Driving behavior analysis method based on weighted cost function
一种基于加权代价函数的驾驶行为分析方法
2023-05-09
Patent
Elektronische Ressource
Chinesisch
Driving behavior analysis method based on weighted cost function
Europäisches Patentamt | 2022
|Modeling and Analysis of Driving Behavior Based on Probability-Weighted ARX Model
Online Contents | 2013
|Orbit Transfer Optimization for Asteroid Missions Using Weighted Cost Function
Online Contents | 2015
|