基于蚁群优化UKF算法的汽车状态估计

Vehicle State Estimation Based on Ant Colony Optimization Algorithm

针对汽车状态估计中过程噪声和观测噪声的时变特性,提出一种新的自适应滤波算法。该算法基于三自由度非线性汽车动力学模型,在利用UKF对汽车状态量进行估计的同时,引入蚁群优化算法,根据目标函数对过程噪声和观测噪声进行寻优,实现了过程噪声和观测噪声的自适应作用,估计精度的大幅提高。虚拟实验验证了蚁群优化UKF算法的鲁棒性和精度。研究结果对汽车主动控制系统的开发具有重大的理论指导意义。

For time-varying characteristics of process noises and observation noise problems in ve- hicle state estimation, a new adaptive filtering algorithm was put forward. The new adaptive filtering algorithm was based on the 3-DOF nonlinear vehicle dynamics model, when UKF algorithm was ap- plied to estimate vehicle state, the ant colony optimization algorithm was introduced at the same time. The adaptive functions of process noises and observation noises were achieved according to optimiza- tion principles based on objective function. The estimation precision increases greatly by using the new adaptive filtering algorithm. Robustness and accuracy of the ant colony optimization UKF algorithm were verified through the virtual experiments, the results will have important theoretical significance for the development of automobile active control system.