一种基于长基线交汇的超短基线定位系统精度评价方法

A precision evaluation method of USBL positioning systems based on LBL triangulation

超短基线定位系统的定位精度和准确度是评价系统性能的重要指标,通常采用固定点的定位重复度来评价定位精度,采用其他解算方法作为真值参考评价定位准确度.本文首先分析了以误差椭圆理论为基础的超短基线精度评价方法,给出了理论推导公式,证明了观测数据和理论误差椭圆的关系.本文提出了一种基于长基线交汇的超短基线定位系统精度评价方法,通过长基线交汇模型求解目标的真实位置评价系统的准确度.根据该方法解算得到的待定目标位置作为真值参考,能够反应系统误差的修正情况.最后采用该方法进行海试数据处理,处理结果表明该方法能够较好的反应定位精度,进一步修正了系统偏差,修正系统偏差后和修正前相比定位精度提高了0.2%,具有良好的工程应用价值.

Ultra short baseline （USBL） positioning system is widely used in underwater geophysical field survey, acoustic tow fish positioning, and sea oil engineering The precision and accuracy are important technical indexes. Normally, people often care about how to improve the precision of position, but there is no unified method about how to evaluate the precision and accuracy. In most experiments, a beacon as target is moored on the seafloor using buoyancy, and survey positions of the beacon repeatedly in a circle track. The waviness of positioning results is used to evaluate the precision, which is analyzed by a statistical method. This paper analyzes the precision evaluation method based on error ellipse, gives the theoretical formulations, and proves the relationship between observation data and error ellipse. This paper also proposes a precision evaluation method of USBL positioning systems based on long baseline （LBL） triangulation, using the obtained result as the true position to evaluate the accuracy of USBL which can offer suggestions to find system error. Using multiple observations to increase redundancy, the precision is far greater than USBL positioning method. Estimated positions can be used as the true ones to serve as a reference in evaluating the accuracy. If the deviation between estimated positions using the USBL and LBL methods is larger than expected, the system needs to be recalibrated. Finally, this paper processes the data from sea experiment. The actual sea trial is processed using the LBL method proposed in this paper. Result shows that the precision of a fixed target is well reflected and the system error is modified further, and thus improves the positioning precision of 0.2%. Result also shows that this method may be of a great application value.