The invention discloses a driver overtaking intention recognition method based on GM-HMM, and the method comprises the following steps: carrying out a simulated driving test, carrying out the whole-course video recording in a test process, the collected test data comprising the speed of a main vehicle, the relative speed of a front vehicle, the relative speed of a rear vehicle, the relative distance of the front vehicle, the relative distance of the rear vehicle, a driver face yaw angle and a steering wheel angle; processing test data; dividing the processed test data into left-turn overtaking test data, right-turn overtaking test data and non-overtaking test data, and marking intention labels of corresponding overtaking intentions in the data; training an overtaking intention recognition model based on the GM-HMM; and performing driver overtaking intention recognition according to the obtained GMHMM-based overtaking intention recognition model. The GM-HMM-based overtaking intention recognition model is established by simulating driving test acquisition data, the invention has the advantages of convenience in data acquisition, low cost and high prediction accuracy, and the overtaking intention of a driver can be recognized according to the model, so that the safety can be effectively improved.
本发明公开了一种基于GM‑HMM的驾驶员超车意图识别方法,其步骤包括:进行模拟驾驶试验,试验过程中全程录像,采集的试验数据包括主车速度、前车相对车速、后车相对速度、前车相对距离、后车相对距离、驾驶员面部横摆角、方向盘转角;处理试验数据;将处理后的试验数据划分为左转超车试验数据、右转超车试验数据和非超车试验数据,并在数据中标记相应超车意图的意图标签;训练基于GM‑HMM的超车意图识别模型;根据所获得的基于GM‑HMM的超车意图识别模型进行驾驶员超车意图识别。通过模拟驾驶试验采集数据,建立基于GM‑HMM的超车意图识别模型,具有数据采集便捷,成本低廉,预测准确度高的优点,根据该模型识别驾驶员超车意图,可有效提升安全性。
Driver overtaking intention recognition method based on GM-HMM
基于GM-HMM的驾驶员超车意图识别方法
2021-04-20
Patent
Elektronische Ressource
Chinesisch
Driver overtaking intention recognition method and system based on hybrid strategy game
Europäisches Patentamt | 2020
|