An automatic control method for semi-active suspension of driverless vehicle based on multi-sensor information fusion in complex environment

Abstract

A semi-active suspension control method based on multi-sensor information fusion was proposed to improve the safety and smoothness of driverless A quarter suspension vibration model considering multi-sensor information fusion was established, revealing the relationship between road roughness information and vehicle vibration. A quarter suspension vibration model considering multi-sensor information fusion was established, revealing the relationship between road roughness information and vehicle vibration. The camera and radar wave were used to scan and identify the uneven road conditions, and a mathematical model of road roughness was created. The information fusion and matching of uneven road surface were carried out by The information fusion and matching of uneven road surface were carried out by detecting the edge intersection ratio and Graph Neural Network(GNN) algorithm, obtaining a more reliable mathematical model of uneven road surface in complex environment. It is proposed to calculate the optimal damping ratio of semi-active suspension using the information of vehicle speed and road roughness, and to adjust the suspension to the optimal damping ratio of semi-active suspension. It is proposed to calculate the optimal damping ratio of semi-active suspension using the information of vehicle speed and road roughness, and to adjust the suspension to this damping ratio to adapt to different road conditions in real time. The vehicle ride comfort test under typical road input conditions was carried out. The vehicle ride comfort test under typical road input conditions was carried out, and the vibration acceleration time domain response signals of different suspensions were compared and analysed. The results show that the maximum peak vibration acceleration of the unmanned suspension controlled by multiple information fusion is reduced by 43 % comparing with that of the unmanned suspension. reduced by 43 % comparing with that in the passive suspension under the same conditions, which verifies the superiority of the proposed method.