针对GPS定位中的伪距观测值出现奇异及跳变而无法进行准确定位的问题,研究了剔除奇异及跳变值的方法,给出了具体的公式推导和判定准则;为提高定位精度,建立了定点模型条件下的衰减记忆法滤波,并给出了具体的实现方法;仿真结果表明:该方法能有效解决跳变条件下准确定位问题,衰减记忆法滤波的定位精度优于常规卡尔曼滤波和迭代最小二乘法的定位精度。

As a result of observation pseudo-range’s oddity or jumping that effecting accuracy in the positioning of GPS, the correct positioning method was studied, the material formula and judgment rule were also given; The attenuated memory filtering of fixed pointing model was build for enhancement pointing accuracy, also the realize method was proposed; The simulation results indicate that, the method can give the correct position with solving observation pseudo-range’s oddity or jumping problem effectively, the attenuated memory Kalman filtering can achieve a higher positioning precision, and it’s also higher than that of general Kalman filtering and iterative least square.