This paper presents the control signal rectification and actuation mapping (CSRAM) framework, developed to improve the reliability and precision of real-time robot joint control. The framework integrates three modules, namely the drive signal rectifier (DSR), the signal pole detector (SPD), and the rising/downstream detector (RDD), which ensure signal compatibility, dynamic mapping consistency, and directional stability during actuation. Unlike conventional control converters, CSRAM effectively compensates for nonlinearities, latency, and synchronization issues in closed-loop systems. Experimental validation using a hexapod-to-quadruped (Hexaquad) robot showed that the proposed method, when combined with an anti-windup PI controller, reduced steady-state error from 14% to below 1%, improved transient and settling times by 0.3 to 0.4 seconds, and decreased three-dimensional trajectory RMSE by 63.7%. These results confirm that CSRAM provides a low-complexity and computationally efficient preprocessing layer for improving real-time performance in multi-joint and legged robotic systems, with strong potential for adaptive and industrial robotic platforms.
Copyrights © 2026