Application of machine learning approach on improving quality of semi-trailer truck air suspensions. (4th January 2022)

• Record Type: Journal Article
• Authors: Liem, Nguyen Van; Huan, Yuan
• Is Part Of: International journal of vehicle noise and vibration. Volume 17:Number 1/2(2021)
• Page Start: 101
• Page End: 120
• Publication Date: 2022-01-04
• Subjects: semi-trailer trucks -- ride quality -- road-friendly -- air suspensions -- MLA -- machine learning approach -- genetic algorithm; Automobiles -- Noise -- Periodicals; Automobiles -- Vibration -- Periodicals; 629.231
• Languages: English
• ISSNs: 1479-1471

Abstract

A combination of the machine-learning-approach (MLA) and optimal-fuzzy-logic-control (OFLC) is proposed for semi-trailer truck air suspensions to enhance the ride-quality and road-friendliness. A dynamics model of semi-trailer truck is built for the simulation. The root-mean-square accelerations of the vertical vibration (awzd ), pitching (awφd ), and rolling angle (awθd ) of the tractor driver, and dynamic load-stress factor (κ) of wheel-axles are selected as indexes. Based on the data map of road surfaces and optimal rules of OFLC, the MLA is developed to control semi-trailer truck air suspensions. Results show that vehicle air suspensions optimised by OFLC and MLA are overall superior to passive air suspensions under different simulations. Especially, the awzd, awφd, awθd, and κ with MLA are markedly reduced by 13.29%, 11.05%, 17.78%, and 11.41% compared to OFLC under combined roads of level-E, level-C, and level-D. Consequently, the vehicle's ride-quality and road-friendly are further enhanced by applying MLA.