Enhanced tire road friction coefficient estimation through interacting multiple model design based on tire force observation

Accurately understanding the tire-road friction coefficient (TRFC) interaction is crucial for enhancing overall vehicle control performance. However, existing estimation methods heavily rely on accurate tire modeling, which may fail to predict TRFC effectively under conditions of poor modeling accuracy or varying operational environments. This paper introduces an interacting multiple-model estimation method for TRFC, leveraging tire force observations. Firstly, taking advantage of the observation capabilities of distributed drive electric vehicles, a longitudinal tire force estimator with unknown inputs is designed. Simultaneously, a lateral tire force observer based on adaptive sliding mode observer is constructed, making full use of the vehicle's onboard sensor data. A TRFC estimator is proposed by employing a square-root cubature Kalman filter, incorporating square-root filtering to reduce algorithm complexity while maintaining accuracy. Finally, an interacting multiple-model mechanism is specifically designed to handle both pure longitudinal dynamics and combined operating conditions. The proposed algorithm is co-simulated by Carsim-Matlab and further tested in a real vehicle equipped with a 6-axis wheel transducer. The results show that the algorithm proposed in this paper can accurately estimate the tire force and TRFC across various maneuvering and road scenarios. Compared to traditional algorithms, the proposed approach demonstrates superior estimation accuracy and robustness.