Papers

We prove the existence of a nematic phase in a model of $l\times w$ hard rectangles on the square lattice with two allowed orientations and a large aspect ratio $k:=l/w$. The proof is based on a two-scale cluster expansion method developed previously by Disertori–Giuliani for hard rods in 2D and Disertori–Giuliani–Jauslin for hard plates in 3D. Our main contributions lie in explicitly tracking the constants and parameters and completing the arguments left implicit in these works. Hence, the proof produces a sufficient set of quantitative conditions from which estimates for the required aspect ratio can be extracted. A non-optimized evaluation of these conditions yields the bound $k\ge 10^{72}$. Although it vastly overshoots the numerical prediction, $k_{\min}=7$, our result appears to be the first rigorous estimate of the aspect ratio required for the formation of a nematic phase in hard rectangle systems.

Two-sided threshold crossings arise in discrete monitoring systems whenever intervention is triggered by departure from an operating band. This paper develops a finite-state first-passage framework based on a biased random walk with symmetric absorbing barriers. To remove parity gaps of the raw hitting time, we analyse a parity-corrected lifetime through even- and odd-state Markov skeletons. Adjacent-row likelihood-ratio inequalities verify the monotonicity condition and establish increasing failure rate and new-better-than-used properties. These ageing properties yield log-concave survival probabilities and geometric persistence bounds. Substochastic-matrix formulas are derived for survival, hazard, mean lifetime, state-conditioned remaining useful life, and finite-horizon crossing risk. A renewal-cost criterion then converts the first-passage distribution into an optimal preventive-intervention schedule. Model parameters are obtained by matching the drift and variance of observed increments to a two-point random-walk approximation, with a rolling-window extension for nonstationary regimes. Numerical experiments compare exact calculations with Monte Carlo diagnostics, moment-matched inverse-Gaussian and Weibull benchmarks, serially correlated increments, calibration perturbations, and fixed-interval policies. A synthetic remote-patient-monitoring example illustrates how discrete physiological deviations can be mapped to transparent risk scores and personalized review schedules. The framework provides an auditable link between discrete stochastic dynamics, remaining-lifetime prediction, and condition-based intervention, while separating mathematical validation from clinical validation.

Training data constrains optimizer geometry through the covectors visible to a declared information channel. We study how such partial information determines a full positive cometric relative to a reference and which degrees of freedom remain unidentified. Our central result resolves full-column-rank positive-definite compression under affine-invariant Riemannian geometry. The compression map is a split-Hadamard metric submetry and admits an explicit unique completion that is the affine-invariant nearest full geometry realizing a visible target and yields exact full-to-visible variational reduction. When the channel moves, the completions form a gauge-invariant rank stratification of the positive-definite cone. Its closed-form pullback pair metric separates visible-metric motion from subspace rotation through a reference-mismatch weight, yields an explicit positive-semidefinite multi-direction Gram matrix, and exposes the precise singularity of reference-valued modes. The mechanism is explained by a metric theorem equating ball submetry, attained fiber distance, and lossless reduction of every monotone radial visible decision problem. A smooth split-Hadamard theorem supplies coherent information sheets, proximal commutation, and solution-wise gradient-flow lifting. The positive-definite realization also gives closed-form prior-data shrinkage. Diagonal and block optimizer families reduce to relative-interior conic image tests with valid facial certificates, while deterministic and finite-sample bounds quantify recovery of the visible geometry and its subspace. Together these results characterize exact reduction, reference-dependent completion, and structured expressivity for the stated finite-dimensional affine-invariant model.

We introduce the ideal zero-product probability, a new probabilistic invariant associated with a finite commutative ring $R$, defined by \[ \zeta_k(R) = \frac{ |\{(I_1,\ldots,I_k)\in\mathcal I(R)^k : I_1\cdots I_k=(0)\}| } {|\mathcal I(R)|^k}, \] where $\mathcal I(R)$ denotes the set of all ideals of $R$. This quantity gives the probability that the product of $k$ independently and uniformly chosen ideals is the zero ideal. We establish its basic properties, including invariance under ring isomorphisms, monotonicity in $k$, and multiplicativity with respect to finite direct products. Explicit formulas are obtained for finite fields, finite Boolean rings, finite chain rings, and finite principal ideal rings. For finite chain rings, we prove that the invariant depends only on the Loewy length. We also derive closed formulas using inclusion–exclusion and bounded compositions, together with generating functions. Finally, we study the asymptotic behaviour of $\zeta_k(R)$ and prove that it converges to $1$ as $k\to\infty$. We also compare the ideal zero-product probability with the classical zero-product probability of ring elements on finite chain rings, obtaining explicit formulas and showing that the two invariants capture fundamentally different structural information.

A universal cover is a convex set that covers all sets of diameter 1 after some rotation, reflection, and translation. First, we show that a regular dodecahedron that circumscribes a ball of diameter 1 is not a universal cover in $\mathbb R^3$, answering a question of Chakerian [K]. Second, we improve the bounds on the minimum volume of a universal cover in $\mathbb R^3$ to $(0.545193,0.655984)$. Third, we give an infinite family of centrally symmetric polytopes in $\mathbb R^d$ with $d(d+1)$ facets that circumscribe the unit ball and are universal covers. This is the maximum possible number and proves a conjecture of Makeev [M2]. Finally, on the least number of facets of a centrally symmetric polytope in $\mathbb R^d$ that circumscribe the unit ball and is not a universal cover, we generalize Makeev's lower bound [M1] and improve his upper bound [M3].

This book is about the mathematical foundations of data science. 1. Introduction 2. Curses, Blessings, and Surprises in High Dimensions 3. Singular Value Decomposition and Principal Component Analysis 4. Linear Regression and Regularization 5. Graphs, Networks, and Clustering 6. Nonlinear Dimension Reduction and Diffusion Maps 7. Linear Dimension Reduction via Random Projections 8. Optimization for Data Science 9. Classification 10. A Mathematical Introduction to Deep Learning 11. Large Sample Limit of Graph Laplacians 12. Community 13. Concentration of Measure and Gaussian Analysis 14. Matrix Concentration Inequalities 15. Compressive Sensing and Sparsity 16. Low-Rank Matrix Recovery

This study introduces and investigates generalized q-Morgan-Voyce polynomials and their specific cases, including the first and second kinds of q-Morgan-Voyce polynomials, q-Horadam-Morgan-Voyce polynomials, and Fibonacci-type q-Morgan-Voyce polynomials. The generalized q-Morgan-Voyce polynomials are defined by the recurrence relation featuring the specific q-power q^{n-2} and a negative sign, formulated as M_{n}(x,q) = (x+1+q)M_{n-1}(x,q) - q^{n-2}M_{n-2}(x,q) for n >= 2. The generating functions and explicit expressions for these polynomials are established by utilizing the Fibonacci operator and the binomial theorem. Furthermore, the study extends these polynomials to negative indices and derives the corresponding explicit formulas. Summation formulas and determinantal presentations of the generalized polynomials and their special cases are also comprehensively provided. Finally, the q-Cassini's formula for the generalized q-Morgan-Voyce polynomials is established through the definition of specific square matrices and their subsequent recurrence relations.

We develop a discrete Beurling–Kato theory for bounded operators and relate it to a Hardy–Sobolev functional calculus on Stolz domains. Our central class is that of Ritt operators. We prove that a bounded operator is Ritt if and only if it admits a bounded Hardy–Sobolev calculus on a Stolz domain. The construction is based on a logarithmic reproducing formula and uniform bounds for the associated logarithmic kernels. We then derive Kato-type results and characterise the Ritt property by Beurling–Kato defects formulated in terms of powers of the operator. This yields a discrete theory parallel in spirit to the sectorially bounded holomorphic semigroup setting, but intrinsically global in nature. We also discuss examples and sharpness phenomena, and prove that the Ritt property is preserved under convex combinations of powers and under positive domination on Banach lattices.

This paper investigates the topological stability of random dynamical systems . Our main goal is to extend the classical result of Walters \cite{Walters77} to the random setting by employing the notions of random continuum-wise expansivity and shadowing property. We prove that any random dynamical system that is random $cw$-expansive and satisfies the random shadowing property is randomly topologically stable. Furthermore, in the random setting we establish the topological invariance of these properties under conjugacy and analyze the relationship between topological transitivity and the periodic shadowing property.

We prove a sharp extrapolation bound for $k$-sparse Fourier sums with arbitrary real frequencies and no separation assumption. For every $g(t)=\sum_{j=1}^k v_je^{i\lambda_jt}$ with $\lambda_j\in\mathbb{R}$ and every $x\geq 1$, \begin{align*} |g(x)|\le k\exp(O(k\operatorname{arcosh} x))\|g\|_{L^2[-1,1]}. \end{align*} Near the endpoint at $x=1+\delta$, this gives an exponential factor of $\exp(O(k\sqrt\delta))$, improving on the $\exp(O(k^2\log k\cdot\delta))$ estimate of Chen and Price (ICALP 2019), and the exponent is optimal up to a constant factor. Our theorem has algorithmic applications to recovering gapless Fourier-sparse signals: it improves the resolution of cluster-location algorithms and the sample complexity of Fourier interpolation algorithms. In addition, it gives sharp bounds on the error transfer of sparse Fourier models just outside the training interval.

The VC-dimension is a fundamental measure of the complexity of a set system. In this paper, we introduce and study a colorful variant of VC-dimension that captures the behavior of set systems on colored ground sets. By studying this new notion, we obtain a variety of geometric results. First, we prove that separable abstract convexity spaces with Radon number $D$ admit a Tverberg theorem with Tverberg number $O(D^2 r \log r)$. This bound significantly improves the $O(Dr^2\log r)$ bound of Alon and Smorodinsky from SODA'26 and is the first quasi-linear bound in $r$, in which the dependence on $D$ is not super-exponential. Second, we prove the first colorful $k$-wise Tverberg theorem for separable abstract convexity spaces. Using this theorem, we obtain a colorful selection lemma with $O(D^3)$ colors, an uncolored selection lemma for subsets of size $O(D^3)$, a weak $\varepsilon$-net theorem with nets of size $O_D(\varepsilon^{-O(D^3)})$, and a $(p,q)$-theorem with exponent of $\mathrm{poly}(D)$. All these quantitative bounds are significantly better than the best previously known general bounds for abstract convexity spaces. Finally, we extend our method to obtain a colorful Tverberg theorem for unions of convex sets, generalizing the uncolored theorem of Alon and Smorodinsky (SODA'26).

Probability distributions are central to information theory, statistical inference, and modern probabilistic learning. Maximum entropy selects a probability state under prescribed constraints, but it does not specify how that state is reached, how probability is transported, or how dissipation and external information exchange are accounted for along the path. We develop a path-dependent entropic Lagrangian calculus that extends static state selection to probability-path evolution through restricted generators, upper-limit history terms, and explicit balance–entropy port routing. The construction yields the thermal state relation, conservative probability balance, and nonnegative production under standard mobility closure. Its KL/Shannon sector recovers maximum-entropy and Bayesian laws as stationary no-flux states, while time-dependent information potentials separate internal dissipation from supplied information power. Composable information and structural potentials control tails, sparsity, robustness, regularization, and nonlocal multimodality without changing the accounting architecture. Two numerical examples verify mass conservation, energy decomposition, and the total free-energy ledger.

We study univariate polynomials with coefficients in an idempotent semifield and their factorization. We do not assume the idempotent semifield under consideration to be totally ordered, in contrast with most of the existing work on this topic. We notably determine when a polynomial splits into linear factors, and when its associated polynomial function does so. These results lead us to characterize algebraically closed idempotent semifields – those in which every polynomial function splits. We prove in particular that every complete idempotent semifield is algebraically closed. We also relate algebraic closedness to the properties of preradicability and radicability and to the existence of solutions to polynomial equations or inequalities.

We consider a diffuse interface model for the mixture of two incompressible fluids driven by transport noise. Under suitable abstract assumptions, we prove the existence of a global weak martingale solution as well as the pathwise uniqueness of a global strong solution in the two dimensional case. This is the first result addressing a unified framework derived from the works by Brzeźniak & Motyl and Mikulevicius & Rozovskii, on the study of stochastic partial differential equations.

We compare four methods for the construction of algebraic cubature rules on planar elements, whose boundary is tracked by splines. The methods, that we have developed over the last two decades, are based on Green theorem together with some cornerstones of polynomial approximation theory: Gaussian quadrature,Tchakaloff theorem, discretized Chebyshev expansion (hyperinterpolation), Fekete-like interpolation. We discuss their advantages and drawbacks in view of the application to curved polytopal element methods. We have also made freely available at a single site the corresponding open-source Matlab codes.

We generalize the geometric construction of vertex operator algebra (VOA) bundles and their associated automorphic forms from the elliptic modular curve to arbitrary Fuchsian groups $ \Gamma \subset \mathrm{PSL}_2(\mathbb{R})$. A sharp topological dichotomy emerges regarding the existence of a holomorphic weight-$2$ quasi-automorphic generator $E_2^\Gamma$. When $\Gamma$ has a cusp, we construct $E_2^\Gamma$ via the analytic continuation of parabolic Eisenstein series and prove that the space of quasi-automorphic forms is a free polynomial extension, allowing the algebraic setup of the genus one theory, including the quasi-VOA structure and the characterization of strict automorphic forms via a lowering operator. Conversely, when $\Gamma$ is cocompact of genus $g\ge 2$, Atiyah theorem on holomorphic connections rigorously obstructs the existence of $E_2^\Gamma$. For this obstructed case, we provide dimension formulas that link the shortage in lifting quasi-automorphic forms to the failure of quasi-primarity within the VOA, fully resolving the torsion-free case and conjecturing the extension to groups with elliptic points.

This geometric study evaluates a compact seven-petal, 6:5-resonant tulip-shaped orbit constellation for lunar south-pole positioning, navigation, and timing (PNT) and direct-to-Earth relay services. The tulip-shaped orbits are compared against elliptical lunar frozen orbit (ELFO) constellations over the NASA LunaNet Service Volume 2 (SV2), covering lunar latitudes south of -75 degrees. We compare a six-satellite tulip baseline with a minimum-cost five-satellite variant; both use the same shared three-body orbit and differ only in satellite count and along-track phasing. Performance is scored against three Initial Operating Capability C (IOC-C) metrics: line-of-sight (LOS) link availability, Lunar Augmented Navigation System (LANS) geometric dilution of precision (GDOP) below 6, and daily extravehicular activity (EVA) usable-PNT windows. Both tulip constellations satisfy all three IOC-C metrics across SV2. The six-satellite configuration meets requirements with wide margin: 75% worst-point daily GDOP availability and 18 hours of daily EVA support. The five-satellite variant also passes, but with thinner margin: 44% availability and 10 hours of EVA support. Unlike the ELFO configurations, each spacecraft in the tulip-shaped constellation maintains continuous Earth line of sight, providing persistent geometric opportunity for direct single-hop Earth relay. Because all spacecraft share a single three-body orbit, initial phasing and post-failure reconstitution reduce to along-track drift maneuvers between neighboring orbits, with screening-level estimates indicating low maneuver cost.

The dual root lattice $A_n^*$ is an important lattice in quantization, coding, and estimation. It can be represented as the projection of the integer lattice $\mathbb{Z}^{n+1}$ onto the $n$-dimensional hyperplane whose coordinates sum to zero. This representation makes $A_n^*$ particularly natural for quantizing simplex-constrained data, such as histograms and probability distributions. This paper studies the closest-point problem for $A_n^*$: given a query vector $y$, find the lattice point $x\in A_n^*$ minimizing $\|y-x\|^2$. The fastest previously known method is the linear-time algorithm of McKilliam, Clarkson, Smith, and Quinn (MCSQ), which employs bucket sort as a core operation. We present a faster linear-time algorithm. The key observation is that the closest-point objective depends on the rounding residuals only through two prefix aggregates: a count and a residual sum. Hence the elements inside each bucket never need to be sorted, stored, or traversed. This replaces the linked-list traversal and pointer chasing of MCSQ with a single bucketing pass over two flat arrays with counting-sort-style accumulates. A scaled objective further makes most of the computation exact integer arithmetic, and when the input coordinates are rationals with a common denominator, for example, histograms or empirical distibutions, the entire algorithm becomes exact and integer-only. Experiments on an Intel Core i9-13900H show speedups of about $1.8\times$ to $3.0\times$ over MCSQ for $n=2,\dots,100$, with larger gains at higher dimensions. The proposed algorithm is also noticeably faster than Conway and Sloan methods for other root lattices, including $A_n$, $D_n^*$, and $E_8$. An open-source implementation is available in the "fanstar" project.

This paper presents a data-driven method for modeling the pressure response of a hydraulic clutch control circuit. The system consists of a variable-force solenoid, accumulator, pressure regulator valve, and latch valve, and exhibits nonlinear behavior caused by hysteresis, latch transitions, and actuator dynamics. A baseline model using commanded current variables captured the general pressure response but failed to represent hysteresis and latch behavior accurately. The input vector was therefore extended with current derivative information, and several classifiers were tested to separate latch-related operating regimes before fitting Gaussian Process regression models to the resulting partitions. Nonlinear SVC and gradient boosting produced the highest latch-classification accuracy, and nonlinear SVC was selected for the final local-regression pipeline. The proposed approach was evaluated on unseen ramp-rate data and compared against a physics-based Amesim model. The machine-learning model reproduced the measured pressure response and hysteresis behavior more accurately than the physics-based simulation for the tested operating conditions. These results suggest that machine-learning plant models can complement physics-based hydraulic models during hardware development and controller calibration when representative test-stand data are available.

Given two dense multivectors of the Clifford algebra $C\ell(V, Q)$ with $N=2^{p+q}$ coefficients, the fastest known classical algorithms compute their geometric product in $O(N^{\omega/2})$ arithmetic operations, where $\omega$ denotes the matrix multiplication exponent. I show that, under amplitude encoding, a quantum computer executes the geometric product in $O(\operatorname{polylog} N)$ time, using logarithmic space with sublogarithmic circuit depth. This exponential speedup establishes Clifford multiplication as a quantum primitive, providing an efficient computational foundation for quantum geometric algorithms and relativistic simulations.

This paper investigates the backward dynamics of hyperspace systems induced by continuous interval maps. Focusing on the hyperspace of continua, we provide a structural characterization of the $\alpha$-limit sets of backward branches for interval subcontinua. A central result of this work is the proof that the special $\alpha$-limit set of any nondegenerate subinterval is always closed. This reveals a striking topological contrast with classical single-point dynamics, where special $\alpha$-limit sets need not be closed. Finally, we prove that if a special $\alpha$-limit set contains two nondegenerate periodic continua with disjoint orbits, then the base map has a horseshoe and, consequently, positive topological entropy.

In waveguide dynamics and moving-load problems (e.g., high-speed trains), critical velocities indicate the onset of strong vibration amplification. In systems that are invariant in the direction of motion, these velocities can be identified from dispersion relations as points where the phase and group velocities of a propagating mode coincide. Finding such points indirectly by tracing dispersion curves can be cumbersome and potentially unreliable for multimodal systems with complex branch interactions. We present a direct method for computing critical velocities in such scenarios, specifically in the context of semi-analytical methods. Starting from a polynomial parameter-dependent eigenvalue problem for the wavenumber-frequency relation, incorporating the additional condition of equal phase and group velocities yields a singular polynomial multiparameter eigenvalue problem that can be linearized and solved using established algorithms. The proposed approach enables the simultaneous computation of all critical points without requiring the tracing of dispersion curves. Its performance is demonstrated by several benchmark problems, confirming the accurate and robust identification of critical velocities.

We study the Service-Aware Multi-Commodity Flow (SAMCF) problem, in which demand is elastic and governed by a logit choice model while routing is subject to hard capacity constraints. In a centralized, system-optimal setting, the network operator jointly determines how much demand to serve and how to route it. We formulate the SAMCF as a nonlinear program and propose an iterative fixed-point heuristic that alternates between solving an inelastic MCF via column generation and updating demand from the resulting service levels. Two linear approximations based on piecewise-linear demand functions and McCormick envelopes serve as benchmarks, while a piecewise-linear outer-approximation of the demand function is used to provide valid lower bounds. Computational experiments on public transport instances show that the heuristic finds near-optimal solutions in under two seconds - orders of magnitude faster than the benchmark methods - while matching their solution quality on all instances they can solve within a ten-minute time limit.

We investigate Krylov spread complexity for states evolving under time-dependent Hamiltonians. For periodically driven systems, we formulate the problem within Floquet theory and show how the Magnus expansion provides a systematic approximation when the Floquet Hamiltonian is not available in closed form. We then extend this framework beyond periodic driving and demonstrate that, in addition to the globally truncated Magnus expansion, a piecewise Magnus expansion provides a reliable method when the global expansion loses convergence or accuracy. Our results provide practical tools for analyzing complexity growth in a broad class of time-dependent quantum systems.

We study synchronous Boolean dynamics on finite loopless directed graphs in which a vertex is active at the next time step exactly when its number of active in-neighbors is prime. We call these systems Pulse Graphs. Let $L(n)$ denote the largest attractor period realizable on $n$ vertices. Exhaustive enumeration gives \[ L(1),\ldots,L(5)=1,1,1,3,9. \] Our main result determines the exponential order of the maximum period: \[ 2^{n-3}-1\leq L(n)\leq2^n-n-1 \qquad(n\geq5). \] The lower bound is obtained by implementing a maximal-length affine feedback register using prime-count logic gates. For $n\geq6$, the construction is loopless, has maximum in-degree five, and uses only $O(n)$ edges. For odd $n\geq7$, a period-3 control module improves the lower bound to \[ 3(2^{n-4}-1). \] For complete directed graphs, we derive an exact update formula, classify all attractors as fixed points or complement two-cycles, prove that every orbit reaches its eventual attractor within three updates, and count the attractors explicitly. We also derive the activation probability under independent random inputs. For sparse random directed graphs, the associated prime-Poisson mean-field map undergoes a nondegenerate fold at \[ c_\ast\approx3.824963, \qquad \rho_\ast\approx0.368241, \] with local bistability immediately above the threshold.

We prove Krattenthaler's conjecture from 2001 about the $1/3$-phenomenon for lozenge tilings of semiregular hexagons. In a first step we reduce the problem to the case of regular hexagons. In a second step we further reduce the question to a special case already covered in the literature. This is achieved by vast application of the celebrated Zeilberger Algorithm and the Holonomic Ansatz.

We develop a theory of adelic loop groups on the universal one-dimensional solenoid \(S^1_{\mathbb Q}=(\mathbb R\times\widehat{\mathbb Z})/\mathbb Z_{\mathrm{diag}}\), the compact abelian group whose Pontryagin dual is \(\mathbb Q\) rather than \(\mathbb Z\). We introduce the adelic projective line \(\mathbb{CP}^1_{\mathbb Q}\), its ring of Laurent–Puiseux series, and holomorphic vector bundles defined by solenoidal clutching data. We prove that its Picard group is naturally isomorphic to the additive group \(\mathbb Q\). The paper establishes a scalar Wiener–Birkhoff factorization theorem, a matrix Wiener lemma, exact factorization for ordered triangular and small-norm cocycles, a density theorem for factorable matrix loops in the Wiener algebra \(\mathfrak W_{\mathbb Q}\), and a Birkhoff–Grothendieck splitting theorem in the pro-algebraic category. These results lead to the Solenoidal Birkhoff–Grothendieck conjecture, asserting that every \(g\in \mathrm{GL}*n(\mathfrak W*{\mathbb Q})\) admits a factorization \(g=h_-^{-1}\operatorname{diag}(\chi_{q_1},\ldots,\chi_{q_n})h_+\), where \(h_\pm\in\mathrm{GL}*n(\mathfrak W^\pm*{\mathbb Q})\) and \(q_i\in\mathbb Q\). We also develop the Kahler, Grassmannian, and Morse–Bott geometry of adelic loop groups in the spirit of Pressley–Segal. Finally, we compare the theory with perfectoid geometry. The Fargues–Fontaine curve provides a non-archimedean structural counterpart of \(\mathbb{CP}^1_{\mathbb Q}\) at the level of rational slope data, Kedlaya's slope theory supplies a (p)-adic analogue of Wiener–Birkhoff factorization, and the Fargues–Fontaine classification provides a proved perfectoid model for the matrix splitting problem formulated here. This comparison yields a Harder–Narasimhan reformulation of the Solenoidal Birkhoff–Grothendieck conjecture.

Graphs that change over time arise throughout applications, but there is no single standard way to decompose them into smaller pieces. In this paper, we propose a systematic categorical method for doing so. The main idea is to combine structured decompositions, which generalize graph decompositions, such as tree-decompositions, with persistent narratives, which model time-varying data as diagrams. We prove that, under suitable categorical hypotheses, any static theory of decompositions can be lifted to a corresponding temporal theory. As case studies, we apply this construction to time-varying graphs and recover natural temporal analogues of ordinary tree-width, complemented tree-width, and the tree-independence number.

Input–output feedback linearization fails at decoupling singularities, where the decoupling matrix loses rank, the relative degree is lost, and the linearizing control becomes unbounded. This paper develops a singularity-robust trajectory-tracking controller for square nonlinear control-affine systems that tracks through isolated decoupling singularities with bounded control. The method recasts tracking as real-time arc-length homotopy continuation, equivalently a continuous-time Newton/Davidenko flow, and replaces the inverse decoupling matrix by the least-norm Moore–Penrose solution of an augmented matrix $A=[\Lambda\mid b]$, where $b$ is the homotopy direction. A transversality condition $w^T b \ne 0$, with $w$ in the left null space of the decoupling matrix, keeps the augmented matrix full row rank through a generic rank-one loss. The resulting flow agrees with feedback linearization away from the singular set, tracks with $O(1/k)$ error, and re-locks after each crossing. The theory also characterizes the reflection-versus-branch-crossing dichotomy at Whitney folds and relates the reflection case to a Filippov sliding mode. Extensions cover dynamic relative-degree-one minimum-phase systems and arbitrary relative degree via filtered-error reduction. The core results are formalized and machine-checked in the Lean proof assistant, including the multi-parameter (higher-corank) repair criterion, the uniqueness of the sliding motion at the fold, and the machine-checked treatment of Filippov solutions. Simulations include a redundant 2-DOF manipulator, relative-degree-one and relative-degree-two plants, and a dual-active-bridge series-resonant DC/DC converter, where the method performs bounded inversion across buck/boost and resonance singularities while preserving zero-voltage soft switching.

We study second-order optimality conditions for optimal control problems with integral cost. We consider extremals obtained by concatenating finitely many regular arcs and prove both a necessary condition for weak local optimality and a sufficient condition for strong local optimality within this class. The key object is the Jacobi curve, a curve of Lagrangian subspaces encoding the second variation along the reference extremal. In the regular case, this curve is smooth and can be identified with the tangent spaces to a suitable field of extremals. For piecewise regular extremals, the field of extremals is only piecewise smooth, and the associated Jacobi curve has discontinuities at the switching times. This makes the relation between vertical intersections, conjugate points, and optimality more delicate. We develop this discontinuous Jacobi-curve framework and show how it yields effective second-order tests. The case of one-dimensional free-control systems is studied in detail as an illustrative class in which piecewise regular extremals arise naturally.