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。
Traditional Kalman filtering algorithm easily leads filter to diverge and reduces the filtering accuracy and system real-time performance when multi-sensor is mixed with data. This paper puts forward an improved adaptive filtering algorithm for multi-sensor mixing data, which gets more accurate information data. The algorithm restrains filtering divergence and improves filtering accuracy and stability with introducing filtering convergence criterion to the simplified Sage-Husa filter. And the algorithm adjusts gain matrix with strong tracking filter, thus making the filter have strong tracking performance and improving the filter＇s robustness for uncertain system model and processing capacity for mutation status. Simulation comparison between the improved algorithm and the traditional Kalman filtering algorithm shows that the improved adaptive filtering algorithm has better robustness and filtering effectiveness with system model parameters mismatched and the real variable noise unknown. And it also shows that the filtering accuracy and real-time performance have been improved obviously.
出处
《太原科技大学学报》
2016年第3期163-168,共6页
Journal of Taiyuan University of Science and Technology
基金
山西省自然科学基金(2013011035-2)
