车辆转弯工况下间接式胎压监测系统脉冲数的修正  

Correction for pulse number of indirect tire pressure monitoring system under vehicle turning condition

作　　者：王宝琳 夏怀成[1] 董倩倩 WANG Baolin;XIA Huaicheng;DONG Qianqian(School of Vehicle and Energy,Yanshan University,Qinhuangdao 066004,China)

机构地区：[1]燕山大学车辆与能源学院,秦皇岛066004

出　　处：《汽车安全与节能学报》2021年第4期507-515,共9页Journal of Automotive Safety and Energy

摘　　要：分析了以脉冲法为原理的间接式胎压监测系统在车辆转弯工况下误报警的原因,是将正常胎压下的外侧车轮误判为缺气。利用方向盘转角作为修正参数,建立汽车转弯时的几何关系模型。分析了汽车转弯行驶时轮胎侧偏、车厢侧倾和转向系变形对内外侧车轮脉冲差的影响,提出了利用反向传播(BP)神经网络训练法对转弯工况下的外侧车轮脉冲数进行修正,构建一个3层的BP神经网络,将转弯时的车轮脉冲数等效为直线行驶时的脉冲数。结果表明:BP神经网络训练法对脉冲差拟合的决定因数为0.995,修正后的误报率为0;因此,本修正方法效果良好。This paper analyzed the reason why the indirect tire pressure monitoring system based on pulse method gives false alarm,because the vehicle outside wheel under normal tire pressure is misjudged as lack of air.Established a geometric relationship model of automobile turning by using the steering wheel angle as the correction parameter.Analyzed the effects of tire deflection,carriage roll and steering system deformation on the inner-lateral wheel pulse difference when the vehicle was turning.Proposed a Back-Propagation(BP)neural-network training method to modify the pulse number of the outside wheel under the turning condition.Constructed a three-layer BP neural network.The number of wheel pulses when turning was equivalent to that when driving in a straight line.The results show that the determination coefficient of the BP neural-network training method for the pulse difference fitting is 0.995,and the false alarm rate under the turning condition after correction is 0.Therefore,the correction method has a good effect.

关 键 词：汽车安全 间接式胎压监测 车辆转弯工况 脉冲数修正方法 反向传播(BP)神经网络 

分 类 号：U461.91[机械工程—车辆工程]