机器人反向动力学方程的并行计算
Parallel Computation of Robot Inverse Dynamic Equations
-
摘要: 以PUMA560机器人的分解牛顿-欧拉反向动力学方程为模型,提出了方程分解的原则, 由此得到AOE(Activity On Edge)有向图.以此为基础,按照深度和时差的概念建立了 L-W优先表,并导出了一种启发式的调度算法.该算法在微处理机个数一定的情况下,可得 到最小调度时间.最后,以Stanford机器人的递推牛顿-欧拉反向动力学方程为例,说明了 该算法的有效性.Abstract: With the resolved Newton-Euler inverse dynamic equations for the PUMA 560 as the model, principles for decomposing the equations are proposed and the AOE (Activity On Edge) directed graph is obtained. Based on the AOE directed graph, an L-W precedence list is constructed according to the concepts of "level" and "time difference", and a heuristic scheduling algorithm is also presented. This algorihm can reach the minimal scheduling time on a fixed number of processors. Compared with other ones the model of Stanford manipulator's recursive Newton-Eular inverse dynamic equations, the algorithm presented in this paper is effective.
-
Key words:
- Robot dynamic equations /
- parallel computation /
- multiprocessor systems
-
计量
- 文章访问数: 3468
- HTML全文浏览量: 108
- PDF下载量: 1074
- 被引次数: 0
下载:
