空间站科学实验载荷通用地面检测系统设计

目前空间站各科学实验载荷地面检测系统不通用,同时在轨开展实验可调参数众多,且天地数据交互存在延时,使得在轨实验耗时长。为解决上述问题,提出了一种基于Qt、SQLite3、Python的通用地面检测系统设计。采用了软件模块化思想,通过修改配置文件满足不同科学实验载荷的地面测试需求。同时该设计基于人工智能技术,通过遥科学手段与在轨载荷实时通信,将载荷的科学实验当作黑盒函数,利用遥科学回传的实验结果来构建目标函数。运用神经网络算法拟合科学实验、SciPy调用LBFGS-B预测参数的方法优化载荷的在轨实验。目前该设计已在空间站超冷原子实验和空间站高精度时频柜冷原子微波钟实验设备的地面调试中开展应用,预计载荷在轨开展实验后,能将在轨调试参数的时间从数个月缩短至几周内完成。

Universal Ground Testing System for Scientific Payload on Chinese Space Station

In order to solve the problems that the ground testing systems for scientific payloads of the space station are not universal,and there are too much time consuming in the experimental process due to a large number of adjustable parameters and time delay of data interaction in the experiments,a design of a universal ground testing system based on Qt,SQLite3 and Python is proposed.The idea of software modularization is adopted,and the configuration file is modified to meet the ground test requirements of different scientific experimental loads.Based on artificial intelligence technology,real time communication with the on-orbit payload is realized through telescience.The scientific experiment of the payload is regraded as a black box function,and the objective function is constructed by the experimental results of the telescience return.The neural network algorithm fitting scientific experiments and SciPy calling the L-BFGS-B prediction parameter methods are used to optimize the on-orbit experiments.The design has been applied to the ground debugging of two space station loads,namely the super-cold atomic experiment cabinet and the high-precision time-frequency cabinet cold atomic microwave clock.It is expected that after the load in orbit experiment,the on orbit debugging time can be shortened from several months to several weeks.