摘要
针对传统递推式鲁棒H∞滤波收敛条件在工程中较难满足的问题,提出了一种自适应补偿的递推式鲁棒滤波算法。在分析滤波器收敛条件的基础上,对滤波器中迭代矩阵进行补偿程度自适应的修改,并结合实际系统要求给出了滤波误差定量度量指标,即响应时间、偏离度、波动程度,以验证算法的正确性。GPS/IN S等组合导航系统仿真结果表明,此方法滤波误差下降速度、无偏性及平稳性均优于改进前滤波算法,且计算量几乎不增加;此方法克服了收敛条件不满足导致滤波精度下降的缺陷,扩大了递推式鲁棒滤波的应用范围,具有较大的工程应用价值。
An adaptive compensation recursive robust filtering algorithm is given in this paper, aiming at the problem that the convergence condition of traditional recursive robust H∞ filtering is hard to meet in engineering applications. Based on analysis of convergence condition, the recursive matrix is modified with adaptive compensation. Moreover, to validate the accuracy of the algorithm, the evaluation criteria,including response time, deviation degree and fluctuation degree, are presented. The simulation results in GPS/INS integrated navigation system show that the decreasing rate, unbiasedness and steadiness of the developed algorithm is better than the algorithm unmodified,without more computational burdens. That indicates this modified algorithm can solve the accuracy decreasing problem resulting from the convergence condition unsatisfying and broaden the application of recursive robust filtering and has great value in engineering application.
出处
《火力与指挥控制》
CSCD
北大核心
2010年第10期15-18,23,共5页
Fire Control & Command Control
基金
国家自然科学基金(60634030
60702066)
航天科技创新基金(CASC0214)
教育部新世纪优秀人才支持计划项目(NCET-06-0878)
西北工业大学重点扶持基金项目
关键词
自适应补偿
递推式H∞滤波
组合导航
adaptive compensation
recursive H∞ filtering
integrated navigation
