论文部分内容阅读
采用传统单核实现组合导航,很难保证导航数据的更新频率,造成间断性丢失导航信息;针对飞行器对导航系统可靠性和实时性的要求,利用可编程片上系统技术,采用Nios Ⅱ软核构建了非对称双核处理器系统,分别负责捷联解算与卡尔曼滤波。在此基础上对双核导航系统配置流水线桥提升系统最高频率,并集成硬件浮点定制指令加速卡尔曼滤波和捷联解算,较之采用单核的处理导航算法,速度提高了约47倍;同时,为了保证代码执行耗时的确定性,加入紧耦合存储器确保系统的可靠性。
It is difficult to guarantee the renewal frequency of navigation data and result in intermittent loss of navigation information. To meet the requirements of aircraft reliability and real-time performance of navigation system, this paper uses programmable system-on-chip technology and Nios Ⅱ soft core Asymmetric dual-core processor system, respectively, responsible for the strapdown solver and Kalman filter. On the basis of this, we design the dual-core navigation system to increase the maximum frequency of the system by using the pipeline bridge, and integrate the hardware floating-point custom instruction to accelerate Kalman filter and strapdown solution. Compared with the single-core processing navigation algorithm, the speed is increased about 47 times; At the same time, in order to ensure the code execution time-consuming certainty, adding tightly coupled memory to ensure system reliability.