Joint estimation algorithm and verification of vertical and longitudinal forces for heavy duty tires

DOI:	10.11809/bqzbgcxb2024.09.028
Keywords:	heavy duty tires; circumferential strain; support vector regression machine; tire force estimation
Abstract:	Aiming at the problems of error accumulation and force linear saturation zone in the existing indirect estimation methods of vertical force and longitudinal force, we carry out the research on the joint estimation method of vertical force and longitudinal force of heavy duty tires based on the strain signal inside the tire, establish the finite element model of 16.00R20 heavy duty tires with consideration of the material mechanics and geometrical structure, and validate the accuracy of the finite element model through the tire stiffness loading test and the modal vibration test; The mechanical deformation characteristics and circumferential and axial strain characteristics of the tire under the composite loading condition of vertical force and longitudinal force are investigated; through the asymmetric analysis of circumferential strain curves of the inner liner ply points of the tread, the front and rear grounding angles and grounding lengths are put forward as the identification characteristics of the vertical force estimation, and the difference of circumferential strain valleys is taken as the longitudinal force estimation identification characteristics, and the jointing of the vertical and longitudinal forces is explored and discussed based on which the jointing of vertical and longitudinal forces is investigated. On this basis, the joint estimation model of vertical force and longitudinal force is established by combining the support vector machine. The calculation results show that: when the loading range of vertical force is 5 000~800 000 N, the root mean square error of the training set of the estimation model is 30.34 N, and the average absolute percentage error is 0.14%; when the loading range of longitudinal force is 0~25 000 N, the maximum error of the training set of the estimation model is 128.5 N, and the average absolute percentage error is 0.28%, which makes the model realize accurate estimation of the vertical and longitudinal forces of tires. The model can realize the accurate estimation of vertical force and longitudinal force of tires.
Published:	2024-09-30
PDF