混沌野草扩展Kalman滤波器辨识六维力传感器下E膜模型参数

为解决动载环境下噪声污染导致六维力传感器测量精度急剧下降，以及扩展卡尔曼滤波器难以获得最优系统干扰矩阵的问题，提出了一种基于混沌野草算法优化的扩展卡尔曼滤波（CIWO-EKF）算法。根据挠度与应变之间的关系，构建了六维力传感器下E膜非线性模型。基于野草繁殖算法，以前6阶主振型信息构成的系统干扰阵为均值进行高斯采样，产生初始化的可行解。将混沌搜索技术与野草算法相融合，利用野草算法进行全局搜索，通过混沌序列对群体中适应度高于平均值的个体执行给定步数的局部搜索，指导种群向最优解方向逼近，避免搜索过程陷入局部最优。采用改进的野草算法对扩展Kalman滤波中的系统干扰矩阵进行优化处理。仿真实例表明，改进扩展卡尔曼滤波器在提高六维力传感器测量精度的同时，可以保持较好的稳定性和鲁棒性。

Model Parameter Identification for a Lower E-type Membrane of Six-axis Force Sensor Based on CIWO-EKF

The measurement accuracy of a sensor which worked on the environment of the dynamic load could be seriously affected by the pollution of noise signals and the EKF could not gain the optimal system noise matrix. A new EKF based on CIWO was proposed. The nonlinear state-space model was established based on the relationship between the deflection and the strain. The idea of the grass breeding was introduced to achieve the Gauss sampling of system interference matrix consisted of first six-order vibration mode informations and to produce the initial feasible solutions. After combining chaotic search technology with invasive weed optimization (IWO), the global search of the new algorithm was executed by the IWO. Then the chaotic sequences executed the local search to the individual which had the higher fitness value than average and guided the population approaching to the optimal solution. The search process can be avoided to fall into local optimum. Finally, the improved invasive weed optimization algorithm was introduced to optimize the system's noise matrix in EKF. The simulation results indicate that the new algorithm can enhance the measurement accuracy of six-axis force sensor effectively and maintain better robustness and real-time performance.