一种改进的自适应卡尔曼滤波算法

发布时间：2018-01-20 18:29

本文关键词： 卡尔曼滤波 Sage-Husa自适应滤波强跟踪滤波　出处：《太原科技大学学报》2016年03期 　论文类型：期刊论文

【摘要】：针对多传感器数据融合时传统卡尔曼滤波算法极易引起滤波发散,降低滤波精度和系统实时性的问题,研究一种改进的自适应滤波算法对多传感器数据进行融合,得到更为准确的信息数据。该算法在简化的Sage-Husa滤波基础上引入滤波收敛性判据,抑制滤波发散并提高滤波精度和稳定性。同时结合强跟踪滤波思想调整增益矩阵,使滤波器具有强跟踪滤波的特性,提高改进的滤波算法对不确定系统模型的鲁棒性以及对突变状态的滤波处理能力。将改进算法与传统卡尔曼滤波算法进行仿真比较。仿真结果表明,在系统模型参数失配或实变噪声未知情况下,改进的自适应滤波算法有更好的鲁棒性,并且在系统状态突变时仍有较好的滤波效果,明显提高了滤波精度和实时性。
[Abstract]:For the multisensor data fusion of traditional Calman filtering algorithm can easily cause filter divergence, reduce the problem of real time and accuracy of filtering system, an improved adaptive filtering algorithm for multi-sensor information fusion, get more accurate data. The calculation method is introduced into the filtering convergence criterion in Sage-Husa filter based on simplified, restrain the divergence and improve the accuracy and stability of filtering. Combined with the strong tracking filtering idea adjustment gain matrix, the filter has the characteristics of strong tracking filter, improved filtering algorithm to improve the robustness of uncertain system model and filtering ability of the mutation status. The improved algorithm and the traditional Calman filtering algorithm are simulated and compared. The simulation results show that the system model parameter mismatch or consolidation noise of unknown circumstances, an improved adaptive filtering algorithm is better. It is robust and still has better filtering effect when the system state is abrupt, which obviously improves the filtering precision and real time.

【作者单位】：太原科技大学电子信息工程学院;
【基金】：山西省自然科学基金(2013011035-2)
【分类号】：TN713
【正文快照】： 两轮自平衡车系统是高度不稳定的非线性系统,在保持车体平衡过程中,控制器通过传感器检测到的姿态角度信号控制电机产生控制力矩来控制两个车轮的速度和方向,以此控制车体平衡。因此,两轮自平衡车的稳定性很大程度上取决于姿态检测的准确性,然而实际应用中加速度计动态过程采

【参考文献】