基于轮胎加速度信号的胎路作用力估算模型  

Tire-Road Force Estimation Model Based on a Tire Acceleration Signal

作　　者：刘庆纲[1] 郭昊 魏旭明 郎垚璞 王奇 周兴林[2] Liu Qinggang;Guo Hao;Wei Xuming;Lang Yaopu;Wang Qi;Zhou Xinglin(State Key Laboratory of Precision Measurement Technology and Instruments,Tianjin University,Tianjin 300072,China;School of Automobile and Traffic Engineering,Wuhan University of Science and Technology,Wuhan 430081,China)

机构地区：[1]天津大学精密测试技术及仪器国家重点实验室,天津300072 [2]武汉科技大学汽车与交通工程学院,武汉430081

出　　处：《天津大学学报（自然科学与工程技术版）》2023年第2期177-183,共7页Journal of Tianjin University：Science and Technology

基　　金：国家自然科学基金-国家重大科研仪器研制资助项目(51827182).

摘　　要：研究重载卡车对道路的破坏机制能够指导道路的设计与养护,有效延长道路的使用寿命.其中,针对复杂重载荷作用下道路病害研究对轮胎路面作用力测量的需求,本文提出了一种复合工况下胎路三向力的建模估算方法.以11.00R20型全钢载重子午线轮胎为研究对象,经过对结构的合理简化,利用ABAQUS有限元仿真软件以参数化建模方法建立了轮胎有限元模型,并验证了模型的有效性;以不同的垂向载荷、侧偏角、纵滑率、轮速和胎压设置了多种复合工况,并对轮胎模型进行了稳态滚动分析,获得了轮胎同一断面内衬层上多个观测点的加速度响应信号;对采集的加速度信号进行了滤波处理,并对比了不同信号的响应程度,选取其中特征较为明显的信号作为观测加速度信号;提取了观测加速度信号的特征,并结合胎压和轮速作为输入,建立了基于卷积神经网络(CNN)的复合工况下胎路三向力的回归解析模型;以模型对15个测试工况的估算结果,近似地描述了该模型的泛化能力,该模型下胎路三向力估算的平均绝对百分比误差均在7%以内,均方根误差不超过1 kN.结果表明,轮胎内衬层上的加速度信号与胎路作用力具有较高的相关性,基于加速度信号与卷积神经网络建立的模型能够有效地实现胎路三向力的同时估算,为道路病害研究提供了有价值的载荷信息.The research on the damage principle of a heavy-duty truck to a road can guide the design and maintenance of the road and effectively prolong its service life.To meet the demand for the measurement of the tire-road interaction force in the research of road diseases under a complex and heavy load,an estimation method for the triaxial tire-road forces under composite working conditions is proposed in this paper.Taking the 11.00R20 all-steel radial truck tire as the research object and simplifying its structure,a tire finite element model was established using ABAQUS finite element simulation software and a parametric modeling method,and the effectiveness of the model was verified.The tire model was analyzed using steady-state rolling under composite working conditions that combined with different vertical loads,side slip angles,longitudinal slip rates,wheel speeds,and tire pressures,and the acceleration response signals at several observation points on the same section of the inner liner of the tire were obtained.The acceleration signals were filtered,and the observed acceleration signals were selected by the responsiveness.The characteristics of the observed acceleration signals were extracted,and combined with the tire pressure and wheel speed as inputs,a regression analytical model of triaxial tire-road force under composite working conditions based on a convolutional neural network(CNN)was established.Based on the estimation results of the model for 15 test conditions,the generalization ability of the model was approximately described.The mean absolute percentage error of the triaxial tire-road force estimation under the model is less than 7%,and the root mean square error is no more than 1 kN.The results show that the acceleration signal on the tire liner has a high correlation with the tire-road force.The model based on the acceleration signal and CNN can effectively estimate the triaxial tire-road force in real time,which provides valuable load information for the study of road diseases.

关 键 词：胎路三向力 加速度信号 卷积神经网络 有限元仿真 

分 类 号：TQ336.1[化学工程—橡胶工业]