智能四轮独立驱动轮毂电机电动汽车自适应转向研究

发布时间：2018-03-02 17:04

本文选题：轮毂电机电动汽车　切入点：驾驶员转向特性　出处：《辽宁工业大学》2016年硕士论文　论文类型：学位论文

【摘要】：现如今,汽车正日益向电动化、智能化方向发展,四轮独立驱动轮毅电机电动汽车作为汽车发展的重要方向,其控制已成为研究的热点,但目前研究工作大多只针对汽车本身的底盘系统,没有考虑驾驶员特性。论文基于“智能四轮独立驱动轮毂电机电动汽车自适应转向研究”这一国家自然科学基金青年基金项目(51305190),对考虑驾驶员转向特性的四轮独立驱动轮毂电动汽车控制进行研究。首先对驾驶员转向特性分类方法进行研究,基于驾驶模拟器实验,设计了具有多个直角转弯实验路况,选取具有一定驾驶经验的驾驶员,采集能够反映驾驶员转向特性的数据,采用模糊C均值聚类方法将驾驶员转向特征数据分为谨慎型、一般型和激进型3类。其次,在对驾驶员转向特性合理分类的基础上,利用BP神经网络建立驾驶员转向特性辨识模型,选取驾驶员在转弯时的特征值作为模型的输入量,驾驶员转向特性类型作为模型的输出量,设计网络结构,训练特性辨识模型,并对模型辨识准确度进行验证。在此基础上,通过驾驶模拟器实验采集三类人的实验数据,采用RBF神经网络方法建立不同转向特性的驾驶员转向特性参考模型,设计网络结构,训练参考模型,并对参考模型的预测精度和可行性进行验证。然后研究通过四轮驱动力控制产生横摆力矩满足驾驶员喜好转向特性的四轮驱动力控制方法。基于自适应模糊控制理论,设计了横摆力矩参数自调整模糊控制器,控制得出的附加横摆力矩采用四轮驱动力规则分配方法。最后基于驾驶模拟器,将离线训练出的驾驶员转向特性辨识模型和参考模型嵌入到驾驶模拟器整车模型系统中,选取不同类型的驾驶员在驾驶模拟器上进行实验,对驾驶员转向特性辨识的准确性、匹配参考模型后的四轮驱动力控制效果进行在线验证。验证结果表明：所研究的方法能够实现对驾驶员转向特性的准确辨识、自动匹配转向特性参考模型,通过四轮驱动力矩控制实现了满足驾驶员喜好转向特性的电动汽车自适应转向控制。研究成果可为电动汽车智能控制提供理论基础和技术支持。
[Abstract]:Nowadays, the automobile is becoming more and more electric and intelligent. As an important direction of automobile development, the control of four-wheel independent drive wheel motor electric vehicle has become a hot spot. But at present, most of the research work only focuses on the chassis system of the car itself. Based on the research of adaptive steering of intelligent four-wheel independent drive hub motor electric vehicle, a project of National Natural Science Foundation Youth Foundation No. 51305190, this paper is concerned with the four-wheel one-wheel system considering the steering characteristics of the driver. The control of vertical drive hub electric vehicle is studied. Firstly, the classification method of driver steering characteristics is studied. Based on the driving simulator experiment, the road conditions with several right-angle turning experiments are designed, and the drivers with certain driving experience are selected to collect the data which can reflect the driver's steering characteristics. Fuzzy C-means clustering method is used to classify driver steering characteristic data into three categories: cautious type, general type and radical type. Secondly, on the basis of reasonable classification of driver's steering characteristics, BP neural network is used to establish the driver steering characteristic identification model. The driver's characteristic value is selected as the input of the model, the driver's steering characteristic type is taken as the output quantity of the model, the network structure is designed, and the training characteristic identification model is designed. The accuracy of model identification is verified. On the basis of this, the experimental data of three kinds of people are collected by driving simulator experiment, and the reference model of driver steering characteristic with different steering characteristics is established by using RBF neural network method, and the network structure is designed. Training reference model, The prediction accuracy and feasibility of the reference model are verified. Then the four-wheel driving force control method, which can produce yaw torque to satisfy the driver's preference steering characteristic, is studied based on the adaptive fuzzy control theory. The self-adjusting fuzzy controller of yaw torque parameters is designed. The additional yaw torque obtained by the control is assigned by four-wheel driving force rule. Finally, based on the driving simulator, The off-line driver steering characteristic identification model and reference model are embedded in the whole vehicle model system of driving simulator. Different types of drivers are selected to carry out experiments on the driving simulator, and the accuracy of driver steering characteristic identification is obtained. The control effect of four-wheel driving force after matching reference model is verified online. The results show that the proposed method can accurately identify the steering characteristics of the driver and automatically match the steering characteristic reference model. The adaptive steering control of electric vehicle can be realized by four-wheel drive torque control, which meets the steering characteristics of driver preference. The research results can provide theoretical basis and technical support for intelligent control of electric vehicle.
【学位授予单位】：辽宁工业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：U469.72

【相似文献】