Experimental Verification of Force-assistive Optimal Variable Admittance Control of Haptic Systems

This chapter introduces a novel variable admittance control approach aimed at mitigating the common issue of grasp contact instability often encountered in fixed admittance control setups within haptic systems. The methodology involves the analytical derivation of admittance parameters tailored for variable admittance control in one-degree-of-freedom (1-DoF) haptic systems operated by human users. This derivation relies on a comprehensive understanding of the system’s state dynamics and the desired performance characteristics. Traditional haptic controllers typically incorporate a dual-control loop strategy: an outer-loop admittance controller, responsible for defining desired system kinematics, and an inner-loop position controller, tasked with tracking these desired kinematics. It’s worth noting that contemporary robot position controllers have demonstrated remarkable accuracy in their results. The proposed variable admittance control approach is grounded in Pontryagin’s Minimum Principle (PMP) and utilizes a cost function that carefully balances the trade-off between position error and control effort considerations. Key assumptions made in this chapter encompass perfect position control and a rigid contact grasp. The performance of the proposed approach undergoes rigorous evaluation through offline simulations utilizing experimentally acquired data. Furthermore, its effectiveness is validated in real time with the active participation of eleven (11) subjects operating a 1-DoF haptic device equipped with a payload. The chapter also establishes stability bounds for the device under the proposed admittance approach. The empirical findings of this study showcase noteworthy improvements: a statistically significant reduction in control effort and an enhanced tracking accuracy ( \(p <\) 0.05) when compared to a fixed admittance control scheme. Additionally, during a drilling task, the approach exhibits a substantial reduction in force ripple, exceeding 30% ( \(p <\) 0.05), evident in both the approach and drilling phases of the task.