How statement and proof provenance work
The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.
- Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
- AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
- AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.
These labels describe origin, not correctness: citations and verification chips remain separate evidence.
Existence, uniqueness and sign ambiguity of the quaternion lift of a rotation path
Statement
Let be a nondegenerate open interval, let be of class (The orthogonal group and the rotation group ) with body angular velocity in the sense of The body angular velocity of a rotation path: is skew, determines a unique continuous , and , let , and let satisfy , with and as in The quaternion rotation map : conjugation on the imaginary quaternions, the unit sphere , and quaternion lifts. Then:
- There is exactly one quaternion lift of with , namely the unique solution on of the initial value problem
- Every lift of satisfies or ; in particular is the lift with initial value , and .
Facts & Assumptions
Given: The rotation path , its body angular velocity , the times , and the unit quaternion with .
for ; the assignment restricts to a surjective homomorphism whose kernel is , and ; in particular is the identity (The quaternion rotation map : conjugation on the imaginary quaternions, the unit sphere , and quaternion lifts, Definition and Remarks).
For one has , , and because ; the product of two quaternion curves is again . The unit and inverse assertions use , , and (The product rule for quaternion and matrix curves, multiplicativity of the quaternion norm, and the conjugation formula clauses 1-2, The quaternions : real quadruples with componentwise addition and an explicit multiplication formula matching the table on , is a division ring that is not commutative, hence not a field: for , while and , The quaternion rotation map : conjugation on the imaginary quaternions, the unit sphere , and quaternion lifts).
The body angular velocity is continuous and uniquely characterized by , where is the matrix (The body angular velocity of a rotation path: is skew, determines a unique continuous , and clauses 2-3).
For continuous , with and : , and is with ; moreover the initial value problem , with has exactly one solution on (The quaternion lift of a prescribed continuous angular velocity: keeps invariant, its rotation path has body angular velocity , and the initial value problem has a unique solution on the whole interval clauses 1-2).
A linear initial value problem on with continuous coefficient satisfies the Picard-Lindelöf hypotheses: being linear in the state variable with the continuous matrix coefficient , the field is continuous and locally Lipschitz in the state variable, locally uniformly in time (Local Lipschitz continuity in the state variable, locally uniform in time and parameters, First-order systems, initial value problems, and solutions on intervals, The vector space of by matrices over a field, with entrywise operations); hence any two solutions through the same data agree on every common interval containing the initial time, and there is a unique maximal solution (Picard-Lindelöf local existence and uniqueness for first-order systems, Every Picard–Lindelöf initial value problem has one maximal solution on an open interval).
If is continuous on and is an interval, then the image is order-convex: for with the whole segment lies in (The image of an interval under a continuous real function is order-convex, hence an interval, and the image of a closed bounded interval is a closed bounded interval clause 1).
Proof
The body angular velocity is continuous by [F3], so [F4] applies with , and : there is exactly one curve of class with on and .
Both and are curves solving the same linear initial value problem , : for this is the characterization in [F3], and for it is clause 1 of [F4] applied to the solution of step 1.1. By [F5] the coefficient is continuous and the equation satisfies the Picard-Lindelöf hypotheses, so two solutions through the same data agree on every common interval containing ; the interval is such a common interval, hence . Thus is a lift of with .
Let be any lift of and put , a product of two curves and hence by [F2]. By [F2], takes values in ; by the homomorphism property and kernel of [F1], , and this identity element of is ; hence for every . The real part is continuous and takes its values in ; by [F6] applied to the interval its image is order-convex, so it cannot contain both and (the whole segment between them would lie in a two-point set), and is constant. At , is when and when ; thus the constant value gives in the first case and in the second. Since , these are the only cases, so every lift is one of . Finally, is a curve into with initial value and by [F1]. This proves the uniqueness with each prescribed initial sign in clause 1 and the full assertion of clause 2.
Remarks
-
The two lifts and solve the same differential equation with opposite initial data and define the same rotation path; the sign is therefore not determined by . Selecting one sign requires an initial choice, and no construction on this page selects a preferred sign.
-
The statement is conditional on the countable-choice hypothesis of the published covering example
ex-su-two-to-so-three-as-a-covering-homomorphismrecorded in The quaternion rotation map : conjugation on the imaginary quaternions, the unit sphere , and quaternion lifts, which is used through the covering and kernel facts of [F1] — that is, through the identification of with and the kernel . It is not used in the differential-equation part of the proof, which is choice-free. -
Nothing in the statement asserts that the covering is nontrivial, that a closed rotation loop has a closed lift, or that any global continuous sign selection exists; those questions are treated on the companion examples page.
Depends on
The quaternion rotation map : conjugation on the imaginary quaternions, the unit sphere , and quaternion lifts
The product rule for quaternion and matrix curves, multiplicativity of the quaternion norm, and the conjugation formula
The body angular velocity of a rotation path: is skew, determines a unique continuous , and
The quaternion lift of a prescribed continuous angular velocity: keeps invariant, its rotation path has body angular velocity , and the initial value problem has a unique solution on the whole interval
The orthogonal group and the rotation group
Local Lipschitz continuity in the state variable, locally uniform in time and parameters
First-order systems, initial value problems, and solutions on intervals
Picard-Lindelöf local existence and uniqueness for first-order systems
Every Picard–Lindelöf initial value problem has one maximal solution on an open interval
The vector space of by matrices over a field, with entrywise operations
The quaternions : real quadruples with componentwise addition and an explicit multiplication formula matching the table on
is a division ring that is not commutative, hence not a field: for , while and
The image of an interval under a continuous real function is order-convex, hence an interval, and the image of a closed bounded interval is a closed bounded interval
Dependency tree · 2 levels
Sources
- J. Sola, Quaternion kinematics for the error-state Kalman filter (arXiv:1711.02508), Section 2.4.2 (q = e^{omega t/2} solves qdot = (1/2) q (x) omega), printed pages 21-22, and Section 2.4.6 (the double cover: q and -q encode the same rotation), printed page 24
- N. Trawny and S. I. Roumeliotis, Indirect Kalman Filter for 3D Attitude Estimation (University of Minnesota, TR-2005-002), Section 1.1 (a quaternion and its negative describe the same rotation) printed page 2, and Section 1.5 (quaternion time derivative), printed pages 9-10
- A. W. Knapp, Lie Groups Beyond an Introduction, Digital Second Edition (2023), Introduction, Problems 7-9 (kernel {+I,-I} of the homomorphism SU(2) -> SO(3)), printed pages 20-21