阵列天线SAR等新型超高分辨率航空遥感系统是高分重大专项的重要发展方向，迫切需要建立分布式位置姿态（Position and Orientation System,POS）的时空基准系统，以实现多节点运动参数的高精度同步测量。由于机体挠曲变形下各子阵天线间的空间距离（基线）呈复杂随机时变，此时如何实现柔性基线形变条件下的全局最优估计是分布式POS测量研究面临的新挑战。本项目拟研究基于柔性基线形变感知的分布式POS全局估计方法。研究基于光纤光栅的主/子位置姿态形变测量方法，在此基础上研究高精度自适应传递对准估计方法；采用时空深度神经网路学习建立基于数据驱动的柔性基线统一形变时空模型，提出高精度自适应全局估计方法，实现柔性基线多节点时空基准和相对空间信息精确测量。该项目将突破制约新型超高分辨率航空遥感系统成像精度和效率的关键技术瓶颈。

New ultra-high resolution aerial remote sensing systems like array antenna SAR, are an important development direction of high-resolution special projects.It is urgent to establish a distributed space-time reference based on position and orientation system(POS) in order to achieve high-precision and synchronous measurement of multi-node motion parameters.Because the spatial distance(baseline) among sub-array antennas is complex and stochastic,how to realize global estimation in the context of flexible baseline deformation is a new challenge for distributed POS.The project intends to study a global estimation method based on flexible baseline deformation sensing.The method about position and attitude deformation measurement based on fiber bragg grating is first studied.On this basis,the high-precision and adaptive transfer alignment estimation method is studied.A data-driven unified deformation space-time model of flexible baseline is established using temporal-spatial depth neural network learning,then a high-precision and adaptive global estimation method is proposed to realize accurate measurement of multi-node space-time reference and flexible baseline.The project will break through the technical bottleneck that restricts the imaging accuracy and efficiency of the new ultra-high resolution aerial remote sensing system.