Automobile Technology

三点预瞄的智能控制补­偿驾驶员模型

- 杨浩 黄超群 魏显坤 税永波401520) (重庆工商职业学院,重庆

【摘要】为适应不同驾驶员的预­瞄风格和推理决策能力,提出了一种基于三点预­瞄的智能控制补偿驾驶­员模型,根据预瞄前方远、中、近3个点的坐标值对目­标路径的位置关系进行­判断,结合当前车速建立了预­瞄距离自适应的经验指­数模型,提出了以车速、航向角、中点侧向误差为输入,转向盘转角为输出的模­糊智能控制驾驶员模型,并对转向角进行补偿校­正。仿真结果表明:该模型能够合理地判断­目标路径的位置关系,适应预瞄距离的动态调­节机理;在不增加失稳风险的条­件下,有补偿模型较未补偿校­正模型的路径跟踪精度­更高。主题词:三点预瞄 智能控制 驾驶员模型 经验指数模型 补偿校正U469.7 A 10.19620/j.cnki.1000-3703.20200180中图­分类号: 文献标识码: DOI: Three-Point Preview Compensati­on Driver Model with Intelligen­t Control Yang Hao, Huang Chaoqun, Wei Xiankun, Shui Yongbo Chongqing Technology and Business Institute, Chongqing 401520) ( Abstract In order to accommodat­e different drivers’preview manner and their reasoning & decision- making【 】abilities, an intelligen­t control compensati­on driver model is proposed based on three- point preview. The positional relationsh­ip of target path is judged based on coordinate values of the three points ahead - far, middle and near, and an empirical index model is establishe­d that adapts the preview distance based on the current vehicle speed. A fuzzy intelligen­t control driver model is proposed with vehicle speed, heading angle and midpoint lateral error as inputs and steering wheel angle as output, and the steering angle is compensate­d and corrected. The simulation results show that the model can properly judge position relationsh­ip of the target path, adapt to the dynamic adjustment mechanism of the preview distance, and the path tracking accuracy of the compensate­d model is higher than the uncompensa­ted correction model without increasing the risk of instabilit­y. Key words: Three- point preview, Intelligen­t control, Driver model, Empirical index model, Compensati­on correction

前言

基于驾驶员的前视作用、推理决策能力和补偿跟­踪行为,不同学者分别提出了预­瞄驾驶员模型[1- 3]、智能控

[1]制驾驶员模型[4-6]、补偿校正模型[7-8]。文献 基于不同的

5

假设,通过控制横向误差建立­了 种单点预瞄模型。文

[2]

献 基于不同的评价指标,通过优化所设置的权重,建

[3]

立了预瞄时间自适应的­驾驶员模型。文献 假设驾驶员根据侧向偏­差决策出最优加速度变­化率,建立了三阶

[4]

预瞄模型。文献 利用神经网络对任意函­数的逼近能

BP [5]

力,建立了 神经网络驾驶员模型。文献 通过对一段路程进行预­瞄,结合模糊控制,提出了一种路程预瞄

[6]

驾驶员模型。文献 针对转向控制中对于驾­驶员参与和驾驶员状态­考虑较少的问题,提出一种基于驾驶员状­态预测的人机力矩协同­转向控制方法,进一步提出驾驶

[7]

员参与和状态模型预测­控制算法。文献 提出一种任

-

意车速下校正环节参数­的求解方法,用于人 车闭环仿

[8]

真,具有较高精度。文献 提出一种对汽车非线性­动力学具有自适应性的­复合校正方法,对汽车动力学的非线性­特性具有很好的自适应­能力。上述方法在路径跟踪

Newspapers in Chinese (Simplified)

Newspapers from China