Research on collision warning strategy considering characteristics of drivers and road conditions

Abstract

Collision warning strategy considering different characteristics of drivers and complex road conditions is of great importance for active safety control of autonomous vehicles. However,existing estimation methods for Tire-road friction coefficient (TRFC) cannot deal with the problem of estimation accuracy degradation caused by state mutation. Therefore, an adaptive cascaded strong tracking unscented Kalman filter (ACSTUKF) is proposed to estimate TRFC. The proposed algorithm consists of two strong tracking filters, which are connected in series with each other. Fuzzy control is introduced to estimate the reaction time of different drivers. Simulation experiments and real vehicle tests are carried out to demonstrate the effectiveness of the proposed strategy. Results show that the proposed ACSTUKF has a more excellent performance for TRFC than EKF and UKF, which greatly reduces the impact of state mutation. Moreover, ACSTUKF shows a strong robustness to driving conditions and road conditions. The established collision warning strategy is consistent with the actual situation, which realizes the adaptation of human and road.