基于时间序列的GPS定位误差分析

摘要

在现代海战场的环境监测中,针对全球定位系统（GPS）的缺陷,采用时间序列分析的方法建立定位误差模型。首先将获得的数据进行平稳化处理,通过依据样本数据的自相关函数和偏相关函数的统计特性确定采用自回归滑动平均（ARMA）模型,然后根据参数的最小二乘估计和AIC准则建立ARMA（4,2）模型。通过对模型的残差分析,得出残差符合白噪声要求,与实际模型拟合程度较高,最后采用预报器对模型进一步预测,根据预报结果修正误差,明显提高了定位的精度。仿真结果表明了时间序列方法可以有效提高GPS的定位精度的有效性。%In modern naval battle environmental monitoring,the model of GPS dynamic position error is presented by using time series analysis for the defect of GPS.The error series was disposed to be stable firstly.Then the ARMA model was set up based on the statistical quality of auto correlation function and the partial correlation.After that,the time series method was used to make ARMA（4,2） model by parameter estimated and AIC criterion.After analyzed the residual errors,we can judge that the white noise is conform to the requirement and the model is conform to the real system.At last the model was predicted by one step predictor.The position precision can be improved by the predicted result used to revise error.The result of emulation showed that the position error data can be improved validity.