Intermittent control: a computational theory of human control

Area of Science:

• Human motor control
• Cybernetics
• Control theory

Background:

• Continuous control models advance human motor control understanding.
• Existing models overlook intermittent feedback, triggered responses, and refractory periods.
• Human control systems involve significant time delays.

Purpose of the Study:

• To present event-driven intermittent control as a unifying framework for human motor control.
• To explain phenomena not accounted for by continuous control models.
• To demonstrate the biological-cybernetic advantages of intermittent control.

Main Methods:

• Developed an event-driven intermittent controller based on a continuous-time predictive control model.
• Incorporated features like sampling, system matched hold, intermittent prediction, and event triggering.
• Modeled human operator behavior under various conditions, including double stimulus tracking and quiet standing.

Main Results:

• Intermittent control explains a wider range of human operator behaviors than continuous control.
• The model naturally explains refractoriness and bimodal stabilization distributions.
• Under specific conditions (small thresholds, regular sampling), intermittent control mimics continuous control.

Conclusions:

• Event-driven intermittent control offers a more comprehensive framework for human motor control.
• This approach reconciles continuous and intermittent control hypotheses.
• Intermittent predictive control is computationally advantageous for systems with time delays.