【摘要】 为分析H∞滤波算法在惯性导航系统INS/GPS组合导航应用中对动态环境的自适应能力,选用卡尔曼滤波算法和Sage-Husa算法作为对照,并改进了Sage-Husa算法,构造了较为全面的惯导系统误差模型、扰动数据情形和飞行轨迹,比较分析了三种算法的自适应能力.仿真结果表明:在这种验证环境中,H∞滤波算法可调参数受到的约束较多,与改进的Sage-Husa算法相比,卡尔曼增益和估计协方误差与量测值变化的相关性较弱,导致自适应能力较弱.在这类模拟动态环境中H∞滤波算法的自适应能力要低于改进的Sage-Husa算法.方法和结果对于鲁棒的INS/GPS组合导航算法的工程化应用有较高的实用价值.更多还原

【Abstract】 In order to compare and analyze the adaptive ability of H-infinity filtering algorithm to dynamic circumstances in INS/GPS integrated navigation application,Kalman filtering algorithm and Sage-Husa algorithm were chosen as the comparisons.Sage-Husa algorithm was improved.The more comprehensive INS error models,data disturbance circumstances and flight track were constructed.Adaptive abilities of the three algorithms were compared.The simulation results show that in the validation environment,the adjustable parameters of H-infinity filtering algorithm are subjected to more restrictions and that,compared with the improved Sage-Husa algorithm,the correlation of the Kalman gain and the estimation error covariance with the change of measurement is low,which results in the weaker adaptive ability.In the simulated dynamic environment,the adaptive ability of H-infinity filtering algorithm is weaker than the ameliorated Sage-Husa algorithm.The methods and results have higher practical value for a further study of the engineering application of the robust INS/GPS integrated navigation algorithm.更多还原