Research On Algorithm Of Road-tire Adhesion Coefficient Estimation Based On Multi-source Information Fusion

Comprehensive and accurate environmental perception is a necessary condition for vehicle intelligence.The road-tire adhesion coefficient describes the adhesion performance of the vehicle on the current road.It is an important part of road information.Accurate and rapid identification of the road-tire adhesion coefficient is of great significance for intelligent vehicle decision-making and control.The traditional sensor-based and dynamics-based identification methods analyze the physical properties of the road surface and the vehicle dynamic response respectively,and estimate the road-tire adhesion coefficient from different aspects.In order to improve the convergence speed and accuracy of road-tire adhesion coefficient estimation results,this paper proposes a road-tire adhesion coefficient estimation algorithm based on multi-source information fusion,which makes full use of road image information and vehicle dynamics response to realize the Combination of sensor-based and dynamics-based recognition methods.Focusing on the estimation algorithm of road-tire adhesion coefficient based on multi-source information fusion,the following contents are studied:Firstly,the road type recognition algorithm based on support vector machine is studied and the algorithm is verified.Through the analysis of the road image information,the article extracts the feature of the road image color and texture based on the HSV color space and the gray level co-occurrence matrix,and constructs the road image feature vector,and then realizes the road type recognition through the support vector machine algorithm.The algorithm is verified by collecting real road images,and the experimental results show that the algorithm can accurately identify a variety of road types.Secondly,road-tire adhesion coefficient estimation algorithm based on cubature Kalman filter is studied,and the algorithm is verified by simulation experiment.Based on the analysis of the classic vehicle dynamics model and tire model,this article uses the vehicle 7-degree-of-freedom model and the Dugoff tire model to model the vehicle and tire,and then realizes the accurate estimation of the road-tire adhesion coefficient under the framework of cubature Kalman filter.The algorithm is verified by Simulink simulation model,and the experimental results show that the algorithm can accurately estimate the road adhesion coefficient.Finally,a road-tire adhesion coefficient estimation algorithm based on multi-source information fusion is designed.The article studies the Burckhardt adhesion coefficient-slip rate model and revises it.On this basis,the internal connection between the road adhesion coefficient and the road surface type is established.By analyzing the advantages and disadvantages of the image estimation method and the dynamics estimation method,a dynamics-image fusion road-tire adhesion coefficient estimation algorithm of dynamic gain scheduling is proposed.The algorithm effectively utilizes image information and vehicle dynamics response,and achieves rapid convergence of vehicle adhesion coefficient estimation.In order to verify the effectiveness of the proposed algorithm,the paper carried out simulation experiments and real vehicle verification.The experiment compares the proposed algorithm,traditional dynamics method and image estimation method.The results show that the proposed algorithm can effectively compensate for the loss of potential adhesion coefficient in the dynamics estimation algorithm,and make full use of the vehicle driving state information,thereby improving the accuracy of road-tire adhesion coefficient estimation result;and the image estimation result responds quickly after identifying the change of road type,which makes the convergence speed of the fusion algorithm significantly better than that of the traditional dynamics estimation algorithm.