基于K-means聚类的汽车电动尾门防夹算法研究

打开文本图片集

中图分类号：U463.832 文献标识码：A 文章编号：1003-8639（2026）02-0090-03

【Abstract】Aiming at the problem that the anti-pinch function of the automotive power liftgate system is prone to failure due to complex working conditions，this paper proposesananti-pinch algorithm forthepowerliftgate basedonthe K-meansclustering method.Firstly，theHallfeedbacksignalsduringthenormal movementof thepower liftgateunder different workingconditionsarecolected，andananti-pinch feature modelbasedonK-meansclusteringisestablished.Then， inthe anti-pinch detection stage，the pinch force is estimated based onthe distance between the measureddata and the K clusters，and this estimated pinch force iscompared withathresholdvalue todetermine whetherapinch event has occured. The results ofthebench testsshow that： teanti-pinch functionis efectiveandreliableunderdifferentsuplyvolages and parking angles.The pinch forces are between 5ONand 8ON，meeting the regulationsof theautomotive industrystandard.

【Key words】power liftgate; anti-pinch control;K-means; machine learning

0 引言

电动尾门作为广泛应用的汽车智能化配置，极大地提升了用车便利性。（剩余2485字）