Critique: Lie Brackets and “Sequencing” (Misinterpretation of Actuation)
Critique: Lie Brackets and “Sequencing” (Misinterpretation of Actuation)
Summary of Concern
The article articles/nonlinear-control-insights.qmd suggests that “Lie bracket analysis” formalizes the biomechanical concept of sequencing (proximal deceleration). Specifically, it claims that proximal deceleration is a “necessary control action to trigger non-holonomic acceleration… effectively ‘steering’ the drift field”.
This conflates Nonholonomic Control (generating motion in unactuated directions via commutators, e.g., parallel parking) with Holonomic Coupling (momentum transfer in a fully actuated chain). In a linked chain like the arm-club system, proximal deceleration transfers energy to distal segments via inertial coupling terms in the mass matrix (\(M_{ij}\)) and Coriolis terms (\(C\)). This is a direct dynamic effect visible in the equations of motion \(\ddot{q} = M^{-1}(\tau - C\dot{q} - G)\). It does not require Lie Brackets (\([f,g]\)) to explain.
Invoking Lie Brackets implies the system is Underactuated in a way that requires complex maneuvers to generate motion in forbidden directions. But the clubhead is fully reachable via the 3 actuated joints. “Sequencing” is about Efficiency (Energy Transfer), not Reachability (Geometry).
Location
- Page:
articles/nonlinear-control-insights.qmd - Section: Expanded Nonlinear Control Tools and Their Relevance -> Lie Bracket Analysis
- Claim: “Proximal deceleration is not an aesthetic choice but a necessary control action to trigger non-holonomic acceleration… effectively ‘steering’ the drift field.”
Nature of the Issue
- Theoretical Misclassification: Treating a Dynamic phenomenon (Kinetic Chain Energy Transfer) as a Geometric phenomenon (Lie Bracket Generation).
- Misuse of “Non-holonomic”: The golf swing constraints are holonomic (joints). The system is fully actuated in configuration space. There are no non-holonomic velocity constraints (like a rolling wheel) preventing direct acceleration.
Why This Is a Problem
- Obscures the Mechanism: It hides the simple inertial explanation (Angular Momentum Conservation / \(M^{-1}\) coupling) behind unnecessary and potentially incorrect differential geometry.
- False Necessity: It claims this sequencing is “necessary… to trigger acceleration”. It is not necessary for acceleration (you can accelerate with pure torque). It is necessary for efficient acceleration or speed amplification.
Evidence / References
- Putnam, C. A. (1993). “Sequential motions of body segments in striking and throwing skills.” (Explains sequencing via interaction moments, i.e., Coriolis/Inertial terms, not Lie Brackets).
- Spong, Hutchinson, Vidyasagar. Robot Modeling and Control. (Dynamics of multi-link arms are derived via Lagrange/Newton-Euler, not Lie Brackets).
Severity
- Medium/High: It makes the theory look pretentious and technically confused to a robotics expert.
Suggested Remedies
- Remove “Lie Bracket” reference regarding sequencing.
- Replace with “Inertial Coupling Analysis”: Explain sequencing via the off-diagonal terms of the mass matrix (\(M_{ij}\)) and the Coriolis vector (\(C\)). Proximal deceleration creates an inertial force \(-M_{distal,proximal} \ddot{q}_{proximal}\) that accelerates the distal segment. This is the standard, correct mechanical explanation.