Actuator Selection and Control of an Array of Electromagnetic Soft Actuators

Hussein Zolfaghari, Nafiseh Ebrahimi, Xaq Pitkow, Mohammadreza DavoodiElectronics 2025 · DOI

TL;DR

This paper introduces a novel control framework and an intelligent actuator selection strategy for arrays of electromagnetic soft actuators (ESAAs) to achieve precise trajectory tracking. The system uses a Kalman filter for state estimation and an LQG controller, but crucially, it dynamically switches between subsets of actuators to manage thermal buildup and actuator wear, revealing a surprising trade-off where switching can initially increase heat generation.

What Problem Does This Paper Address?

Soft robots, inspired by biological systems, offer unique advantages like safety and adaptability. A key component of these robots is soft actuators, and electromagnetic soft actuators (ESAs) are promising due to their speed, controllability, and scalability. When these ESAs are arranged in arrays (ESAAs), they can mimic biological muscles. However, controlling these arrays presents challenges:

  • State Estimation: Directly measuring the state of each individual soft actuator is often impractical due to their small size and flexible nature.
  • Precise Control: Achieving accurate trajectory tracking with these compliant systems requires sophisticated control strategies.
  • Thermal Management: Continuous operation of ESAs can lead to overheating, potentially damaging them and degrading performance. This is exacerbated when fewer actuators are used to achieve the same output.

This paper aims to address these issues by developing a control system that can accurately track desired trajectories while also managing actuator health and preventing overheating through intelligent switching.

Methods

The authors propose a multi-faceted approach:

  1. Kalman Filter for State Estimation: Since direct state measurement is difficult, a Kalman filter is employed to estimate the full system state (positions and velocities of all actuator masses) from limited, noisy sensor data. This is a standard technique for state estimation in the presence of noise and partial observability.
  2. Linear Quadratic Gaussian (LQG) Controller: Based on the estimated states, an LQG controller is designed to achieve accurate reference tracking. This controller minimizes a quadratic cost function that balances tracking error and control effort. The tracking problem is reformulated as a regulation problem by augmenting the state with the reference trajectory.
  3. Actuator Selection Strategy: This is the core novelty. Instead of using all actuators continuously, the system dynamically selects subsets of actuators to activate. This is a two-phase process:
  4. Phase 1: Optimal Number Determination: An optimization procedure determines the minimum number of actuators (K*) required to achieve acceptable tracking performance while satisfying a constraint on the number of simultaneously active actuators (at most half the total). This balances performance with reduced actuator usage.
  5. Phase 2: Switching Strategy with Smooth Transitions: A heuristic switching strategy is implemented. Actuators are allowed to be active for a duration T, followed by a mandatory rest period Ť. Crucially, a brief overlap period Î is introduced between switching sets to ensure smooth transitions and prevent abrupt changes in control effort and system output. This overlap is designed to maintain continuity and robustness.
  6. Thermal Modeling: A first-order linear differential equation models actuator temperature, relating it to the squared control effort (as a proxy for current). This allows for the analysis of thermal profiles under different control strategies.

Datasets

Not discussed in the paper. The evaluation is based on simulations.

Key Insights and Ideas

  • Switching Doesn't Always Mean Less Heat: Counter-intuitively, the study reveals that actuator switching, while intended to reduce heating, can lead to higher thermal accumulation. This is because fewer active actuators need to exert a larger control effort, generating more heat per actuator.
  • Overlap is Crucial for Smoothness: Introducing a brief overlap period (Î) between switching actuator sets is vital for smooth transitions. Without it, abrupt changes in control effort lead to noticeable spikes in the system's output trajectory.
  • Trade-off Between Performance and Thermal Management: There's a clear trade-off. Increasing the penalty on control effort (via the R matrix in the LQG cost) reduces heat generation but can slightly compromise tracking accuracy.
  • Actuator Selection as a Constraint Satisfaction Problem: The problem of choosing actuator subsets is framed as an optimization problem that balances tracking cost with an operation cost (weighted by β) and a hard constraint on the number of active actuators.

Key Results

  • Accurate Tracking: The proposed LQG controller with Kalman filtering successfully tracks desired trajectories, both constant and sinusoidal, even in the presence of noise.
  • Smooth Switching: The overlapping switching strategy (Case 2) significantly reduces transients and spikes in the system output compared to non-overlapping switching (Case 1), achieving a lower mean squared tracking error (0.0084 cm² vs. 0.01 cm²).
  • Thermal Behavior:
    • Full actuation leads to lower individual actuator temperatures than switching when no explicit control effort limitation is applied (Figure 11).
  • Increasing the penalty on control effort (Case 5) effectively reduces actuator temperatures, but at the cost of slightly reduced tracking performance (Figure 12).
  • Actuator Usage Management: The switching strategy ensures that actuators get rest periods, preventing continuous use and potentially extending their lifespan, even if instantaneous temperatures are higher.

Limitations and Future Work

  • Actuator Physical Limitations: The current model assumes actuators can deflect without physical limits. Future work will incorporate constraints on maximum displacement to make the model more realistic.
  • Experimental Validation: The study is based on simulations. Future work will involve experimental validation on real ESAA systems.
  • Long-Term Operation and EMI: Future work will also consider long-term operation effects and electromagnetic interference (EMI).
  • Wider Range of Conditions: More extensive evaluations across a wider range of loading conditions and actuator configurations are planned.
  • Real-time Actuator Selection: The current framework implements real-time selection, but further exploration of its nuances and potential improvements is ongoing.

Summary updated

Abstract

Electromagneticsoft actuator arrays (ESAAs) combine compliance with fast, controllable actuation and scalability, providing a promising foundation for the development of interconnected soft actuator arrays inspired by the structure and function of biological muscles. In this work, we present a control framework and an actuator selection strategy for an artificial soft muscle composed of ESAAs to enable accurate reference tracking. Since directly measuring the states of each ESA is often impractical in real-world applications, we first design a Kalman filter-based observer to estimate all system states from available observations. Using these estimates, we develop a Linear Quadratic Gaussian (LQG) controller to achieve reference tracking. Since thermal buildup from constant use can damage the actuators, we consider whether switching between different subsets of active actuators could offer thermal relief. While actuator switching intuitively suggests reduced heating by providing resting periods, our investigation reveals that this strategy can lead to higher thermal accumulation compared to the continuous mode. This is because we need substantially larger control effort when we have fewer active actuators in the switching mode, which, in the absence of effective active cooling, fail to provide sufficient heat dissipation during operation. Simulation results are presented to demonstrate the effectiveness of the proposed method in achieving the trajectory objective and to explore how switching affects the system’s thermal profile, revealing a trade-off between tracking performance and heat generation.
BibTeX
@inproceedings{zolfaghari2025actuator,
  author = {Hussein Zolfaghari and Nafiseh Ebrahimi and Xaq Pitkow and Mohammadreza Davoodi},
  title = {{Actuator Selection and Control of an Array of Electromagnetic Soft Actuators}},
  booktitle = {Electronics},
  year = {2025},
  doi = {10.3390/electronics14183682},
}
Figure