This study proposes a variable weight model prediction control strategy based on road surface recognition (VW-MPC-RSR), aiming to address the problems of road recognition timeliness and model prediction control (MPC) weight adjustment under multi-working conditions. The road surface recognition is achieved by utilizing the VIT-B/16 network model. A test set and a continuous frame video are used to verify the effectiveness of the road recognition strategy. At the same time, a 2-DOF suspension dynamics model is established, and the variable weight MPC control strategy based on road surface recognition is designed. The optimal weights of six road types are determined using a cross-comparison method, and the corresponding weight-switching strategy is formulated. The performance of VW-MPC-RSR active suspension, MPC active suspension, and passive suspension is compared by simulation. The results show that the road surface recognition strategy based on the VIT network can recognize various road states efficiently and accurately. In addition, the suspension control strategy designed in this paper can adjust the control parameters in real time according to road conditions and significantly improve the overall performance of suspensions.
XuJKouFZhangX, et al. Multi-mode switching control of electromagnetic hybrid suspension based on human subjective sensation. Proc IMechE, Part D: J Automobile Engineering2023; 238: 3092–3108.
2.
QinYWeiCTangX, et al. A novel nonlinear road profile classification approach for controllable suspension system: simulation and experimental validation. Mech Syst Signal Process2019; 125: 79–98.
3.
BidgoliMAGolrooANadjarHS, et al. Road roughness measurement using a cost-effective sensor-based monitoring system. Autom Constr2019; 104: 140–152.
4.
EN 13036-7:2019. Road and airfield surface characteristics-test methods-part 7: irregularity measurement of pavement course: the Straightedge Test.
5.
LiuWWangRDingR, et al. On-line estimation of road profile in semi-active suspension based on unsprung mass acceleration. Mech Syst Signal Process2020; 135: 106370.
6.
ZhangQHouJHuX, et al. Vehicle parameter identification and road roughness estimation using vehicle responses measured in field tests. Measurement2022; 199: 111348.
7.
LiangGZhaoTShangguanZ, et al. Experimental study of road identification by LSTM with application to adaptive suspension damping control. Mech Syst Signal Process2022; 177: 109197.
8.
ZhangYMaZSongX, et al. Road surface defects detection based on IMU sensor. IEEE Sens J2021; 22: 2711–2721.
9.
OkamotoKIttiLTsiotrasP.Vision-based autonomous path following using a human driver control model with reliable input-feature value estimation. IEEE Trans Intell Veh2019; 4: 497–506.
10.
DrewsPWilliamsGGoldfainB, et al. Vision-based high-speed driving with a deep dynamic observer. IEEE Robot Autom Lett2019; 4: 1564–1571.
11.
YadavM.A multi-constraint combined method for road extraction from airborne laser scanning data. Measurement2021; 186: 110077.
12.
FakhriMKarimiSMBarzegaranJ.Predicting international roughness index based on surface distresses in various climate and traffic conditions using laser crack measurement system. Transp Res Rec2021; 2675: 397–412.
13.
ZhaoTHeJLvJ, et al. A comprehensive implementation of road surface classification for vehicle driving assistance: dataset, models, and deployment. IEEE Trans Intell Transp Syst2023; 24: 8361–8370.
14.
ZhaoTWeiY.A road surface image dataset with detailed annotations for driving assistance applications. Data Brief2022; 43: 108483.
ChenC-FRFanQPandaR.CrossViT: cross-attention multi-scale vision transformer for image classification. In: Proceedings of the IEEE/CVF international conference on computer vision, Montreal, QC, Canada, 10–17 October 2021, pp.357–366. New York: IEEE.
21.
DingRWangRMengX, et al. Intelligent switching control of hybrid electromagnetic active suspension based on road identification. Mech Syst Signal Process2021; 152: 107355.
22.
WangRLiuWDingR, et al. Switching control of semi-active suspension based on road profile estimation. Veh Syst Dyn2022; 60: 1972–1992.
23.
QinYRathJJHuC, et al. Adaptive nonlinear active suspension control based on a robust road classifier with a modified super-twisting algorithm. Nonlinear Dyn2019; 97: 2425–2442.
24.
PapadimitrakisMAlexandridisA.Active vehicle suspension control using road preview model predictive control and radial basis function networks. Appl Soft Comput2022; 120: 108646.
25.
YuSGuoJXuM, et al. Road preview MPC of semi-active suspension with magneto-rheological damper. Int J Robust Nonlinear Control. Epub ahead of print 7February2023. DOI: 10.1002/rnc.6625.
26.
TheunissenJSorniottiAGruberP, et al. Regionless explicit model predictive control of active suspension systems with preview. IEEE Trans Ind Electron2019; 67: 4877–4888.
27.
PedroJONhlapoSMMpanzaLJ.Model predictive control of half-car active suspension systems using particle swarm optimisation. IFAC Pap OnLine2020; 53: 14438–14443.
28.
DosovitskiyABeyerLKolesnikovA, et al. An image is worth 16x16 words: transformers for image recognition at scale. arXiv preprint arXiv:201011929, 2020.
29.
ChenXLiangCHuangD, et al. Symbolic discovery of optimization algorithms. Adv Neural Inf Process Syst2024; 36: 1–30.
30.
QinYDongMLangariR, et al. Adaptive hybrid control of vehicle semiactive suspension based on road profile estimation. Shock Vib2015; 2015: 636739.
31.
QuZLiuJLiY, et al. Study on multi-mode switching control of intelligent suspension under full road section. Processes2023; 11: 1776.
32.
LiYZhuZZhengL, et al. Multi-objective control and optimization of active energy-fed suspension based on road surface recognition. J Traffic Transp Eng2021; 21: 129–137. (in Chinese)
33.
LiuQSunJZhangH, et al. Road recognition and semi-active suspension control based on Convolutional neural networks. J Ordnance Eng2020; 41: 1483–1493. (in Chinese)
34.
LakMADegrandeGLombaertG.The effect of road unevenness on the dynamic vehicle response and ground-borne vibrations due to road traffic. Soil Dyn Earthq Eng2011; 31: 1357–1377.
