FNT-Based Road Profile Classification in Vehicle Semi-Active Suspension System

Combining the computational intelligence with dynamic responses of vehicle suspension for estimating the road profiles provides effective tool for designing various control strategies. In this paper, a FNT-based road profile classification method is proposed based on the dynamic responses of a quarter semi-active suspension under PID controller and road disturbances generated from power spectral density under the ISO 8608 standard. More specially, a data preprocessing method is designed to reduce the impact of vehicle velocity on dynamic response and determine the appropriate size of the spatial domain for data collection. After that, FNT is employed as the basic model to screen these extracted features for road profile classification with low computational consumption of road evaluation. From the numerical simulation results, the classification accuracy is 98.41% under the proposed road profile classification with six input variables.

[1]  Patrick Gruber,et al.  Regionless Explicit Model Predictive Control of Active Suspension Systems With Preview , 2020, IEEE Transactions on Industrial Electronics.

[2]  Saied Taheri,et al.  Road Profile Estimation for Active Suspension Applications , 2015 .

[3]  Tomonori Nagayama,et al.  Road profile estimation, and its numerical and experimental validation, by smartphone measurement of the dynamic responses of an ordinary vehicle , 2019, Journal of Sound and Vibration.

[4]  Shi-Yuan Han,et al.  Active Fault-Tolerant Control for Discrete Vehicle Active Suspension Via Reduced-Order Observer , 2021, IEEE Transactions on Systems, Man, and Cybernetics: Systems.

[5]  Nikolay Lushnikov,et al.  Methods of Assessment of Accuracy of Road Surface Roughness Measurement with Profilometer , 2017 .

[6]  Huihui Pan,et al.  Nonlinear Output Feedback Finite-Time Control for Vehicle Active Suspension Systems , 2019, IEEE Transactions on Industrial Informatics.

[7]  Shahram Azadi,et al.  Road profile estimation using neural network algorithm , 2010 .

[8]  Shi-Yuan Han,et al.  Fault diagnosis and fault-tolerant tracking control for discrete-time systems with faults and delays in actuator and measurement , 2017, J. Frankl. Inst..

[9]  Yanjun Huang,et al.  Speed independent road classification strategy based on vehicle response: Theory and experimental validation , 2019, Mechanical Systems and Signal Processing.

[10]  Olivier Sename,et al.  Adaptive control scheme for road profile estimation: application to vehicle dynamics , 2014 .

[11]  Wei Liu,et al.  On-line estimation of road profile in semi-active suspension based on unsprung mass acceleration , 2020 .

[12]  Y Delanne,et al.  Road profile input estimation in vehicle dynamics simulation , 2006 .

[13]  Yechen Qin,et al.  Adaptive Robust Nonlinear Active Suspension Control Using an Observer-Based Modified Sliding Mode Interval Type-2 Fuzzy Neural Network , 2020, IEEE Transactions on Intelligent Vehicles.

[14]  Shi-Yuan Han,et al.  Fuzzy Guaranteed Cost \(H\infty\) Control of Uncertain Nonlinear Fuzzy Vehicle Active Suspension with Random Actuator Delay , 2019, Int. J. Fuzzy Syst..

[15]  Bo Yang,et al.  Feature selection and classification using flexible neural tree , 2006, Neurocomputing.

[16]  Zhang Yonglin,et al.  Numerical simulation of stochastic road process using white noise filtration , 2006 .

[17]  Feng Zhao,et al.  Road profile classification for vehicle semi-active suspension system based on Adaptive Neuro-Fuzzy Inference System , 2015, 2015 54th IEEE Conference on Decision and Control (CDC).

[18]  Ye Chen Qin,et al.  The Research of Road Profile Estimation Based on Acceleration Measurement , 2012 .