Memory Pattern Identification for Feedback Tracking Control in Human-Machine Systems.

Area of Science:

• Human-computer interaction
• Cognitive science
• Control theory

Background:

• Manual control tasks engage nondeclarative memory, specifically procedural memories encoding motor skills.
• This study investigates the characteristic patterns within these procedural memories during a one-dimensional tracking task.
• Data from human subjects controlling dynamic systems with varying fractional orders were analyzed.

Purpose of the Study:

• To identify and characterize memory patterns in human operators during manual control tasks involving visual input.
• To analyze the procedural memory patterns associated with controlling dynamic systems.
• To investigate the stability of the human-machine system based on identified memory patterns.

Main Methods:

• A finite impulse response (FIR) controller was fitted to human subject data.
• Pattern analysis was performed on the fitted FIR controller parameters.
• The FIR model was reduced, and stability analysis of the closed-loop human-machine system was conducted.

Main Results:

• The finite impulse response (FIR) model successfully identified and represented patterns in human procedural memory.
• A characteristic time scale of approximately 650 ms was observed for the procedural memory pattern before decay.
• The fitted controller demonstrated stability for systems with a fractional order less than or equal to 1.

Conclusions:

• The proposed FIR model-based control scheme effectively characterizes the linear properties of human manual control across different fractional orders.
• This research supports a biofidelic approach to modeling human manual control based on visual feedback.
• Applications include developing shared-control systems and monitoring human operator states.