为提升智能网联汽车在道路行驶过程中的稳定性和安全性，基于传统 EKF 算法提出自适应优化EKF 估计算法。首先，针对智能网联汽车难以估计的行驶状态进行分析，确定需要观测的参数为纵向速度、横摆角速度与质心侧偏角；然后，基于传统 EKF 算法中噪声难以处理的问题，采用自适应优化算法对系统噪声和量测噪声同步进行估计和优化，使估计过程更能够贴合实际运行工况；最后，采用基于 MATLAB 与 Carsim 联合仿真平台，建立智能网联汽车模型，对高速工况下高附着和低附着路面行驶状况进行估计。试验结果表明，相较于对比算法，EKF 估计算法在响应速度、估计精度和曲线拟合方面效果更优，对车辆的稳定和安全行驶更具保障。

In order to improve the stability and safety of intelligent connected vehicles on the road, an adaptive optimization EKF estimation algorithm is proposed based on the traditional EKF algorithm. Firstly, based on the analysis of the driving state that is difficult to estimate, the parameters that need to be observed are determined as longitudinal speed, yaw Angle speed and side deflection Angle of the center of mass. Then, based on the problem that the noise is difficult to deal with in the traditional EKF algorithm, the adaptive optimization algorithm is used to estimate and optimize the system noise and the measurement noise synchronously, so that the estimation process can better fit the actual operating conditions. Finally, based on MATLAB and Carsim co-simulation platform, an intelligent connected vehicle model is established to estimate the driving conditions of high adhesion and low adhesion road under high-speed conditions. The experimental results show that compared with the comparison algorithm, the algorithm has better effect in response speed, estimation accuracy and curve fitting, and is more secure for the stability and safety of the vehicle.