【摘要】 设计了一种“双CPU+FPGA”架构的冗余导航计算系统，旨在提高无人直升机在高振动环境下的导航系统适应性、可靠性与精度，并有效应对传感器故障及精度不足的问题。系统采用模块化、层次化的软件设计，硬件包含CPU解算模块、驱动模块、信号采集模块、通信模块和电源模块，构建了三重冗余结构。根据所设计的导航系统提出了基于软阈值的三冗余表决容错算法，并对冗余导航信息进行降维处理以减小计算量。动态车载测试结果表明，该多源冗余容错导航系统可以满足无人直升机对新型惯性导航的技术要求，为飞控系统的高性能导航提供了工程应用基础。更多还原

【Abstract】 This paper designs a redundant navigation computing system with a “dual CPU+FPGA” architecture, aiming to improve the adaptability, reliability, and accuracy of the navigation system of unmanned helicopters in high vibration environments, and effectively address the problems of sensor failures and insufficient accuracy.The system adopts a modular and hierarchical software design, with hardware including CPU calculation module, driver module, signal acquisition module, communication module, and power module, constructing a triple redundant structure.A three redundant voting fault-tolerant algorithm based on soft threshold was proposed for the designed navigation system, and the redundant navigation information was reduced in dimensionality to reduce computational complexity.The dynamic onboard test results indicate that the multi-source redundant fault-tolerant navigation system can meet the technical requirements of unmanned helicopters for new inertial navigation, providing an engineering application foundation for high-performance navigation of flight control systems.更多还原