Smoothing Sliding Mode Control for Robust Actuation

This direction is merged into another direction: chattering mitigation. Its last version stays readable and citable here.

Version 1 of 128 papers40 conceptsfrozen 2026-10-04

Sliding Mode Control (SMC) is a powerhouse of robust control theory. Its core strength lies in its ability to force a system's state onto a predefined "sliding manifold" and keep it there, regardless of external disturbances or modeling inaccuracies. However, this robustness traditionally comes at a steep price: chattering.

Chattering is a high-frequency oscillation caused by the controller's need to switch instantaneously across the sliding surface. In a perfect mathematical world, this switching happens at infinite frequency. In the real world, actuators have physical limits, and this rapid "on-off" behavior leads to mechanical wear, thermal stress, and system instability.

A growing body of research focuses on "smoothing" these transitions—replacing the harsh, discontinuous logic of traditional SMC with continuous approximations that preserve robustness while protecting physical hardware.

The Geometry of the Sliding Manifold

To understand why chattering happens, imagine a marble on a curved surface. Traditional SMC acts like a pair of invisible hands that constantly slap the marble back toward a specific line (the sliding manifold) whenever it drifts. If the marble is to the left, the controller pushes right; if it's to the right, it pushes left.

The mathematical heart of this "slap" is the signum function, $\text{sgn}(s)$, which outputs $1$ for any positive value and $-1$ for any negative value. As Shilin Zuo explains in Neural Network-based Sliding Mode Control for Permanent Magnet Synchronous Motor, this discontinuous jump is what creates the high-frequency vibrations that plague motor controllers.

Softening the Blow with Continuous Functions

The most direct way to eliminate chattering is to replace the discontinuous $\text{sgn}(s)$ function with a smooth, continuous approximation. This creates a "boundary layer" around the sliding surface. Inside this layer, the control signal transitions smoothly rather than jumping instantly.

Several researchers have explored different mathematical "flavors" of this smoothing:

  • Hyperbolic Tangent ($\tanh$): Ying Xiong, Mou Chen, and Qingxian Wu utilized the $\tanh$ function for swept-wing aircraft in Robust tracking control of variable swept-wing near space vehicle based on disturbance observers. The $\tanh$ function mimics the shape of the sign function but provides a differentiable curve, ensuring smoother actuator responses. Similarly, Xi Ma and colleagues applied this to spacecraft attitude control to prevent mechanical degradation.
  • Sigmoid Functions: In the context of micromachining, En‐Chih Chang and Rong‐Ching Wu used a Sigmoid Sliding Mode Function to minimize thermal stress on power transistors. Musheng Deng also employed sigmoid-based reaching laws in motor drives to balance convergence speed with torque smoothness.
  • Arctangent ($\arctan$): For robot manipulators, Danni Shi and Jinhui Zhang proposed an arctangent-based reaching law. This maintains high gain when errors are large but provides a gentle "landing" as the system approaches the target.

Integration as a Low-Pass Filter

A more sophisticated approach to chattering suppression involves changing where the switching logic is applied. Instead of applying the discontinuous signal directly to the actuator, researchers apply it to the derivative of the control signal.

In Pitch angle control with fault diagnosis and tolerance for wind turbine generation systems, Yiran Shi and Shoutao Li demonstrate that the integration of this derivative acts as a natural low-pass filter. The "jumps" are absorbed by the integrator, resulting in a smooth output signal. This technique is particularly valuable for sensitive industrial actuators, such as the UAV electric braking systems studied by Yuren Li and Liangbo Tian, where high-frequency chattering could lead to premature hardware failure.

Handling Physical Limits: Saturation and Performance

While smoothing solves the chattering problem, real-world systems also face actuator saturation—the physical limit of how much force or torque a motor can actually provide.

Dongjae Lee and H. Jin Kim addressed this for multirotor drones by using hyperbolic tangent functions to allow commands to saturate smoothly. This prevents the abrupt instability that occurs when a controller tries to "push" harder than the hardware allows.

In Robust Adaptive Finite-Time Tracking Control, Chao Chen and Guibing Zhu took this a step further by using a Gaussian error function to model torque limits. This replaces the sharp "cliff" of a hard limit with a smooth, differentiable curve, making it much easier to prove the system's stability using Lyapunov analysis.

Why It Works: The Intuition of Tunable Transitions

The common thread across these papers is the move from binary logic to tunable transitions. By adjusting the "hardness" of a sigmoid or the width of a boundary layer, engineers can trade off between two competing goals:

  1. Aggressive Robustness: A steeper transition (closer to a sign function) rejects disturbances faster but increases chattering.
  2. Hardware Longevity: A shallower transition (smoother curve) protects the actuators but may allow for slightly larger tracking errors.

As Daoyi Dong and Ian R. Petersen note in their work on quantum systems, these sigmoid-like functions allow a system to operate at maximum power when far from the target and naturally transition into a linear, smooth regime as it converges. This "best of both worlds" approach is what makes smoothed sliding mode control the preferred choice for modern, high-performance robotics and aerospace applications.

Papers in this version

History & citationVersion 1 · 2026-10-04

Merged into another direction · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_smoothing_sliding_mode_control_for_robus_840e05c4,
  title = {Smoothing Sliding Mode Control for Robust Actuation},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/smoothing-sliding-mode-control-for-robust-actuation/rdr_840e05c487c1448ba06f2a00b19e59bd},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-21},
}

This exact version (version 1, frozen text):

@misc{lacuna_smoothing_sliding_mode_control_for_robus_840e05c4_v1,
  title = {Smoothing Sliding Mode Control for Robust Actuation},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/smoothing-sliding-mode-control-for-robust-actuation/rdr_840e05c487c1448ba06f2a00b19e59bd/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-21},
}