基于一种改进CORDIC算法的MATH处理器设计与优化

Design and Optimization of a MATH Processor Based on an Improved CORDIC Algorithm

在精密的电机数控领域中,对三角函数运算的硬件架构性能指标日渐严格。针对传统CORDIR算法求解三角运算时存在迭代次数多、迭代周期长、输入角度范围小等局限性,提出了对其进行角度预处理、镜像迭代、角度补偿、区间换算以及合并迭代结构等优化,并最终完成高精度计算三角函数的MATH处理器设计。在硬件实现上,本处理器在输入角度及坐标范围得到明显优化,计算速率显著倍增,且精度完全满足设计标准,适配于高精度电机驱动等应用领域。

In the field of precision motor numerical control,the performance indicators of the hardware architecture for trigonometric calculations are becoming increasingly strict.In view of the limitations of the traditional CORDIC algorithm for solving trigonometric operations,such as multiple iterations,long iterative cycles,and a small range of input angles,this paper proposes the optimization of angle preprocessing,mirror iteration,angle compensation,interval conversion,and merging iteration structures,thereby completing the design of a MATH processor for high-precision calculation of trigonometric functions.In terms of hardware implementation,this processor has significantly optimized the range of input angles and coordinates,markedly increased calculation rate,fully satisfies the design standard in terms of precision,and is suitable for applications in the field of high-precision motor drives.