基于蚁群优化FEKF算法的汽车状态估计
Vehicle state estimation based on ant colony optimization FEKF algorithm

针对汽车状态估计过程中观测噪声时变问题，提出一种双重迭代自适应滤波算法—蚁群优化模糊逻辑扩展卡尔曼滤波（FEKF）算法. 建立考虑Fiala轮胎模型的汽车二自由度非线性动力学模型，利用模糊逻辑对扩展卡尔曼滤波（EKF）算法估计过程中的观测噪声水平进行在线修正，同时引入蚁群优化算法对模糊逻辑中的输入输出隶属度函数进行优化，得到的双重迭代算法对处理强时变观测噪声水平下滤波估计过程具有很强的自适应性. 最后通过建立虚拟仿真试验来验证该蚁群优化FEKF新算法的估计精度，结果显示，蚁群优化FEKF算法相比较于FEKF算法估计精度更高，鲁棒性更强.
For time-varying problem of observation noise problem in vehicle state estimation，a new dual iterative adaptive filtering algorithm named the FEKF algorithm is put forward. The vehicle two freedom degrees of dynamics model based on the nonlinear Fiala tire model was established. The fuzzy logic was used on the online correction for the estimation process of the EKF algorithm. Ant colony optimization algorithm was introduced to optimize the input and output membership function in the fuzzy logic operations. So the dual iterative algorithm was obtained. The dual iterative algorithm for dealing with strong time-varying noise levels under the filtering estimation has a strong adaptability. Robustness and accuracy of the ant colony optimization FEKF algorithm compared to the FEKF algorithm is verified through the virtual experiment