Geometric stabilization of virtual linear nonholonomic constraints

Alexandre Anahory Simoes,Anthony Bloch,Leonardo Colombo,Efstratios Stratoglou

2024-11-04

Abstract:In this paper, we give sufficient conditions for and deduce a control law under which a mechanical control system converges exponentially fast to a virtual linear nonholonomic constraint that is control invariant via the same feedback control. Virtual constraints are relations imposed on a control system that become invariant via feedback control, as opposed to physical constraints acting on the system. Virtual nonholonomic constraints, similarly to mechanical nonholonomic constraints, are a class of virtual constraints that depend on velocities rather than only on the configurations of the system.

Optimization and Control

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to design a control law so that the mechanical control system can converge rapidly and exponentially to a virtual linear nonholonomic constraint, and this constraint remains invariant through the same feedback control. Specifically:

1. **Problem Background**:
- **Virtual Nonholonomic Constraints**: These constraints are relationships imposed on the control system and are made invariant through feedback control. Unlike physical constraints, virtual nonholonomic constraints depend on the velocities of the system rather than just the configuration.
- **Existing Research**: Previous studies have explored how to make certain types of constraints (such as virtual holonomic constraints) invariant through feedback control and have been applied in fields such as robot motion control.

2. **Specific Problem**:
- The goal of the paper is to find a control law such that the system can operate under a given virtual linear nonholonomic constraint, and when the system already satisfies these constraints, this control law can ensure that the trajectory of the system still conforms to these constraints.
- The author hopes to prove that there is a broader control law that can drive the system to satisfy these virtual nonholonomic constraints, and if the system already satisfies these constraints at a certain point, this control law is consistent with the previously known unique control law.

3. **Main Contributions**:
- A new control law \( u^* \) is proposed. This control law can not only make the system satisfy the virtual nonholonomic constraint but also make the constraint conditions approach zero at an exponential rate.
- The validity of the proposed mathematical results is verified through specific examples and simulations.

4. **Theoretical Basis**:
- Geometric methods are used to describe virtual nonholonomic constraints, and the concept of controlled invariant distribution is introduced.
- The Levi - Civita connection in Riemannian geometry and other related concepts are utilized to derive the specific form of the control law.

5. **Formula Representation**:
- The virtual nonholonomic constraint is defined by the fiber - linear function \( \hat{\mu}_b : TQ \to \mathbb{R} \), where \( \hat{\mu}_b(q, \dot{q})=\mu_b^i(q)\dot{q}_i \).
- The expression of the control law \( u^* \) is:
\[
u^*_a = C_{ab}(-\hat{\mu}_b - G(\hat{\mu}_b))
\]
where \( C_{ab} \) is the inverse matrix of the matrix \( C_{ab}=\mu_b(Y_a) \), and \( G \) is the geodesic vector field.

In summary, this paper aims to design an effective control law through geometric methods and control theory so that the mechanical control system can rapidly and stably satisfy the virtual linear nonholonomic constraint.