转向传动比是影响车辆主动安全性及操纵稳定性的重要因素.为改善线控转向车辆在低附着系数路面下的转向特性,设计了随路面附着系数和车速变化而自适应调整的变增益传动比.建立汽车二自由度模型,分析影响横摆角速度增益的因素,并通过仿真得到影响因素与增益之间的数据关系.采用 Min-Max 归一化方法,将影响因素与横摆角速度增益之间的数据进行预处理,构建神经网络数据集;设计蛇算法优化的 BP 神经网络(SO-BP),利用预处理后的数据集对 SO-BP 神经网络进行训练,由此动态获取变横摆角速度增益.采用变横摆角速度增益与侧向加速度增益按比例综合的策略,设计线控转向变增益传动比.利用 Simulink-CarSim 搭建线控转向整车模型,分别在高附着系数路面双移线工况和低附着系数路面角阶跃工况下,将所设计的变增益传动比与传统定增益传动比进行对比分析.结果表明:高附着系数路面条件下,两种传动比车辆的行驶轨迹误差保持在 3%以内,而变增益传动比车辆的转向盘转角峰值降低了 9.1%;低附着系数路面条件下,变增益传动比车辆在中低车速下的横摆角速度稳态值降低了 22.3%、峰值降低了 24.6%,中高车速下的横摆角速度稳态值降低了 6.6%、峰值降低了10.8%.变增益传动比不仅可以提高车辆在高附着系数路面上的转向灵敏度,也改善了在低附着系数路面上行驶的安全性与操纵性.
The variable steering transmission ratio was a crucial factor affecting the active safety and handling sta-bility of vehicles.In order to enhance the steering characteristics of steering-by-wire vehicles on low-adhesion coef-ficient road surfaces,a variable gain transmission ratio that adapts to changes in road adhesion coefficient and vehi-cle speed was designed.A 2 DOF model was established for the vehicle,to analyze the factors influencing the yaw rate gain,and to obtain the data relationship between the influencing factors and the gain through simulation.The Min-Max normalization method was utilized to preprocess the data between the influencing factors and the yaw rate gain,constructing a neural network dataset.Design a Snake Optimizer Backpropagation Neural Network(SO-BP)was desighed and train to use the preprocessed dataset to dynamically acquire the variable yaw rate gain.A strategy was employed to combines the variable yaw rate gain with the lateral acceleration gain in proportion to design the variable gain transmission ratio for electronic control steering.Simulink-CarSim was used to build a steer-by-wire steering whole vehicle model.Compare and analyze the designed variable gain transmission ratio was analized and compared with a traditional fixed gain transmission ratio under conditions of both high-adhesion coefficient road sur-faces with a double lane change scenario and low-adhesion coefficient road surfaces with a step input scenario.Re-sults indicated that with high-adhesion coefficient road conditions,the trajectory error of the two transmission ratio vehicles remained within 3%,while the variable gain transmission ratio vehicle reduced the peak steering wheel an-gle by 9.1%.With low-adhesion coefficient road conditions,the variable gain transmission ratio vehicle showed a 22.3%reduction in steady-state yaw rate at low to moderate speeds and a 24.6%reduction in peak yaw rate.At moderate to high speeds,the steady-state yaw rate decreased by 6.6%,and the peak yaw rate decreased by 10.8%.The variable gain transmission ratio not only enhanced steering sensitivity on high-adhesion coefficient road surfaces,but also improved safety and maneuverability when driving on low-adhesion coefficient road surfaces.
寇发荣;方博;张新乾;常航涛
西安科技大学 机械工程学院,陕西 西安 710054
交通运输
线控转向;横摆角速度增益;变传动比;SO-BP神经网络;路面附着系数
steer-by-wire;yaw rate gain;variable transmission ratio;SO-BP neural network;friction coefficient
《郑州大学学报（工学版）》 2024 (005)
8-15,29 / 9
国家自然科学基金资助项目(51775426)
10.13705/j.issn.1671-6833.2024.05.001
评论