为实现云相机控制系统的模块化设计、提高系统集成度,提出了基于FPGA(field programmable gate array)的在轨实时云判系统硬件平台的设计方法。在对系统各功能模块进行划分的基础上,确立了片内双核CPU软件架构,其中一个CPU独立处理云判算法以保证所提算法运行的实时性,另一个CPU用于与卫星进行数据交互和部分数据预处理,通过自定义IP (intellectual property)核实现CCD成像和外围接口的时序控制。以CCD图像采集与存储模块为例,重点讨论了CCD驱动电路、模拟前端电路、数据存储电路的设计和时序,并对该模块IP核的设计方法进行了说明。最后对云相机的成像性能参数进行测试,实验结果表明:设定积分时间内,随着积分时间的推移,相机暗电流噪声变化较小;中心视场处,输出非线性度为1.29%;在信号达到饱和光强的80%时,信噪比为128.1,满足云相机工作要求。

In order to realize the modular design of a cloud camera control system and improve the system integration, a design method for the hardware platform of an on-bit real-time cloud judgement system based on field programmable gate array (FPGA) is proposed. On the basis of the system functional module division, an on-chip dual-core CPU software architecture is established. One CPU independently processes the cloud judgement algorithm to ensure the real-time performance of the proposed algorithm, and the other CPU is used for data interaction with satellites and partial data pre-processing. The timing control of the CCD imaging and peripheral interface is realized by the self-defined intellectual property (IP) core. Taking the CCD image acquisition and storage module as an example, the designs and time sequences of the CCD driving circuit, analog front-end circuit, and data storage circuit are mainly discussed, and the design method of the IP core of the module is analyzed. Finally, the imaging performance parameters of the cloud camera are tested. Experimental results show that the change of the dark current noise of the camera with the integration time in the set integration time is small. The output nonlinearity is 1.29% at the central field of view, and the signal-to-noise ratio is 128.1 when the signal intensity is 80% of the saturated light intensity, which meets the requirements of the cloud camera.