Real-Time Minimum-Energy Operating-Point Tracking for Battery-Powered Micro DC Motors Under Dynamically Variable Loading

TL;DR

This paper introduces a real-time method for micro DC motors to optimize energy efficiency by dynamically tracking their minimum-energy operating point under variable loads, improving over traditional conservative strategies.

Contribution

It reveals the load-dependent non-monotonic energy behavior and proposes an adaptive voltage control method with a lightweight load metric for online energy optimization.

Findings

The method successfully tracks the minimum-energy point during load transitions.

Mean response times are approximately 11.5 seconds for load changes.

Convergence voltages are 2.73V for low-to-high and 2.0V for high-to-low load transitions.

Abstract

Micro DC brushed motors are widely deployed in battery-powered biomedical systems, where limited energy budgets and variable physiological loading impose stringent efficiency and safety constraints. However, conventional actuation strategies rely on conservative voltage margins to avoid stalling, leading to systematic energy inefficiency. Furthermore, existing methods primarily optimize steady-state performance, neglecting the energy required to complete individual actuation cycles under dynamic conditions. This paper reveals that the energy consumption per mechanical cycle of a DC motor exhibits a non-monotonic dependence on driving voltage, with a load-dependent minimum that shifts with external loading. Based on this insight, we propose a real-time operating-point tracking method that enables the motor to autonomously converge to its minimum-energy condition. A lightweight load metric derived from current waveform features is introduced to detect load variation, and a two-phase adaptive voltage strategy is developed to track the optimal operating point online. Experimental results demonstrate that the proposed method can track the new minimum-energy operating region under both low-to-high and high-to-low loading transitions. With 3-cycle averaging, the mean response time is 11.55s for the low-to-high case and 11.16s for the high-to-low case, while the mean convergence voltage is 2.73V and 2.0V, respectively.