Optimal Power Extraction from Active Particles with Hidden States

Area of Science:

• Physics
• Statistical Mechanics
• Active Matter

Background:

• Active particles exhibit complex dynamics driven by internal energy sources.
• Feedback control offers a pathway to manipulate and harness energy from active matter systems.
• Observational limitations, such as inaccessible self-propulsion forces, pose challenges in controlling active particles.

Purpose of the Study:

• To identify generic protocols for optimal power extraction from a single active particle.
• To explore power extraction under continuous feedback control with limited observation.
• To establish theoretical bounds for active matter engine performance.

Main Methods:

• Utilized a Bayesian approach combined with the Onsager-Machlup path integral formalism.
• Applied the framework to specific models: free run-and-tumble and active Ornstein-Uhlenbeck dynamics in 1D.
• Assumed only spatial trajectory is observable, not instantaneous self-propulsion force.

Main Results:

• Derived generic protocols for optimal power extraction.
• Demonstrated positive work extraction even in models with time-symmetric trajectories and zero informational entropy production.
• Established theoretical performance bounds for active particle systems.

Conclusions:

• Optimal power extraction is achievable with feedback control under partial observation.
• The developed protocols provide a benchmark for realistic active matter engines.
• Insights into energy harvesting from non-equilibrium systems are provided.