Alphabeta Math
Pipeline-generated
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.

✓ 3 results · all verified · 2 also independently AI-judged
Every result on this page is machine-checked by a proof checker and read in full and owner-audited; the judge is an additional, independent cross-model AI review of the proofs. The 1 not AI-judged were verified by owner audit (typically over a confirmed judge false positive), not failures.

Differential algebraic equations and consistent initialization — Examples

1 · Prerequisites

2 · Summary

These examples compute a consistent current and a backward-Euler step in a simple circuit, differentiate a singular-Jacobian index-two DAE to expose residual amplification and inconsistent initial data, and derive the index-three constrained-pendulum conditions. The higher-index examples mark the limits of the index-one residual and time-stepping guarantees.

3 · Logical flowchart

4 · Definitions, theorems and proofs

None yet.

5 · Examples, counterexamples and false statements

ExampleConstruction: AI-adaptedVerification: AI-adaptedprecheck passOpen item page →

A capacitor-resistor circuit has one consistent algebraic current

Example

Let v(t) be the capacitor voltage and i(t) the series current in a circuit with capacitance C>0, resistance R>0, and source voltage u(t). Kirchhoff's current law and voltage law give the semi-explicit system

Cv′=i,0=Ri+v−u(t).

The algebraic equation determines i=(u(t)−v)/R. With C=R=1, u(t)=1, and v(0)=0, the consistent initial current is i(0)=1, and the reduced ODE is v′=1−v. One backward-Euler step of size h gives

v1=h1+h,i1=11+h.

For h=1, this is (v1,i1)=(1/2,1/2). The pair v(0)=0,i(0)=0 is inconsistent for this same source.

Facts & Assumptions

Given: C,R>0, the displayed circuit equations, the constant source and initial voltage in the example.

[F1]

Semi-explicit index-one DAEs and consistent initial data: a consistent initial pair must satisfy the algebraic equation at the initial time.

[F2]

Backward Euler step equation and algebraic residual: for an ODE Y′=f(t,Y), backward Euler uses the endpoint equation Y1=Y0+hf(t1,Y1).

[F3]

A regular index-one DAE has a local reduced ODE branch: if the derivative of the algebraic equation with respect to its algebraic variable is invertible, it defines a local branch and reduced ODE.

Verification

1.1givenF3algebra

In the circuit, f(t,v,i)=i/C and g(t,v,i)=Ri+v−u(t), so gi=R≠0. Thus the selected algebraic branch is i=(u(t)−v)/R, and the reduced ODE follows from [F3].

1.2givenF1algebra

For C=R=1, u(0)=1, and v0=0, the algebraic residual at i0=1 is 1⋅1+0−1=0, whereas it is 1⋅0+0−1=−1 at i0=0. By [F1], the first pair is consistent and the second cannot be the initial value of a solution.

2.1givenF2step 1.1algebra

Apply [F2] to the reduced ODE v′=1−v from step 1.1. With v0=0 and h>0, the endpoint equation v1=h(1−v1) gives v1=h/(1+h). Recovering the current from the algebraic branch gives i1=1−v1=1/(1+h), so v1−v0=hi1 and i1+v1=1. For h=1, both values equal 1/2.

3.1givenstep 1.2step 2.1algebra∎

The exact reduced solution v(t)=1−e−t, i(t)=e−t satisfies v′=i and i+v=1, and it starts at the consistent pair (0,1). Thus the example's initialization and one-step calculations satisfy both circuit equations.

ExampleConstruction: AI-adaptedVerification: AI-adaptedprecheck passjudge pass (gpt-6-luna)Open item page →

An index-two DAE can have zero algebraic residual and an arbitrary algebraic variable

Example

Consider the scalar semi-explicit system

x′=z,0=g(x,z):=x.

Its algebraic Jacobian is gz=0, so it is not regular index one. In the convention of Differentiation index for a regular local semi-explicit DAE, differentiating x=0 once gives z=0, and differentiating again gives z′=0; the system has differentiation index two. The algebraic residual is g(x,z)=x: at every candidate (0,M), it is zero for any M, although the algebraic variable of a solution must be z=0. Thus no residual-to-z-error bound holds near (0,0). The initial pair (0,1) satisfies the algebraic equation but is not consistent with a solution, while (1,0) violates the algebraic equation and cannot be an initial value.

Facts & Assumptions

Given: The displayed DAE on R2.

[F1]

Semi-explicit index-one DAEs and consistent initial data: a semi-explicit DAE has differential equation x′=f(t,x,z) and algebraic equation g(t,x,z)=0; algebraic consistency means g=0.

[F2]

Differentiation index for a regular local semi-explicit DAE: differentiate the algebraic constraint along the differential equation; the index is the least positive differentiation count at which the regular prolonged equation determines the algebraic variables and their first derivatives.

[A1]

For the index calculation, x∈C2 and z∈C1 on a nondegenerate interval, and the constraint holds at every time in that interval.

Verification

1.1givenF1algebra

Here f(t,x,z)=z and g(t,x,z)=x, so gz=0 at every point. The regular index-one hypothesis gz invertible therefore fails.

2.1A1F2step 1.1algebra

Along any solution, differentiating g(x(t),z(t))=x(t)=0 and substituting x′=z gives z=0. Differentiating this relation gives z′=0. The original constraint does not involve z; the first differentiated relation determines z, and the second determines its first derivative. By [F2] the differentiation index is two.

3.1step 2.1algebra

For every real M, the candidate (x,z)=(0,M) has algebraic residual g(0,M)=0, while step 2.1 shows that a solution has z=0. Taking arbitrarily small M≠0 rules out any finite constant C for which ∣z∣≤C∣g(x,z)∣ throughout a neighborhood of (0,0).

3.2F1A1step 2.1algebra

The pair (0,1) satisfies g(0,1)=0, but a solution would require z(0)=0 by step 2.1, so it is not a consistent initial pair for a solution. The pair (1,0) has g(1,0)=1≠0, so it fails even algebraic consistency.

4.1givenstep 2.1algebra∎

The constant functions x(t)=0,z(t)=0 satisfy both equations, so the system has a solution with the consistent initial pair (0,0). The three boundary calculations above concern only this fixed index-two system and imply no index-one numerical guarantee.

ExampleConstruction: AI-adaptedVerification: AI-adaptedprecheck passjudge pass (gpt-6-luna)Open item page →

The constrained pendulum has differentiation index three

Example

For q,v∈R2, multiplier λ∈R, length ℓ>0, and gravitational acceleration g0>0, consider

q′=v,v′=−g0e2−λq,0=12(qTq−ℓ2),

where e2=(0,1)T. The position constraint, tangency condition, and acceleration constraint determine the allowed initial position, velocity, and multiplier. On the constraint manifold the multiplier is

λ=∥v∥22−g0q2ℓ2.

This is a regular index-three DAE in the convention of Differentiation index for a regular local semi-explicit DAE.

Facts & Assumptions

Given: The pendulum equations, ℓ>0, and g0>0.

[F1]

Semi-explicit index-one DAEs and consistent initial data: a DAE solution must satisfy its algebraic position constraint at every time.

[F2]

Differentiation index for a regular local semi-explicit DAE: the index is the least number of differentiations of the algebraic constraint needed locally to determine the first derivative of the algebraic variable through a regular prolonged equation.

[A1]

The variables have the derivatives used below, the constraint satisfies qTq=ℓ2 along the trajectory, and ℓ>0.

Verification

1.1givenF1A1algebra

Differentiating qTq=ℓ2 once and using q′=v gives qTv=0, the tangency condition. Differentiating a second time gives 0=∥v∥22+qTv′=∥v∥22−g0q2−λqTq. Using [A1], this equation solves uniquely for λ=(∥v∥22−g0q2)/ℓ2.

2.1givenA1step 1.1algebra

Differentiate ∥v∥22−g0q2−λℓ2=0, the second-derivative constraint from step 1.1. Since q′=v and v′=−g0e2−λq, this gives 0=2vTv′−g0v2−λ′ℓ2=−3g0v2−2λqTv−λ′ℓ2=−3g0v2−λ′ℓ2, where the last equality uses the tangency condition in step 1.1. Since ℓ>0, this determines λ′=−3g0v2/ℓ2 with nonzero coefficient −ℓ2 on the highest algebraic derivative.

3.1F2step 1.1step 2.1algebra

The original constraint does not contain λ; its first derivative yields tangency but not λ, its second derivative determines λ, and its third derivative determines λ′. The prolonged equations are regular because the multiplier coefficients are −ℓ2 at each of these latter solves. Therefore the least differentiation count in [F2] is three.

4.1givenA1step 1.1algebra∎

For example, q0=(ℓ,0), v0=(0,s), and λ0=s2/ℓ2 satisfy the position and tangency constraints, and the multiplier formula in step 1.1 gives λ0. Thus this tuple satisfies the initial conditions exposed by the first two differentiations.

Sources