Papers

In this work, we propose an efficient and transparent green learning pipeline to address the automatic modulation classification (AMC) problem. This pipeline aims to enable receivers to blindly identify the modulation modes of the incoming signals in a computationally efficient way with a small model size. Our method includes the following steps. First, the input signal is transformed into a precise representation through the sparse coding method. Second, various features are extracted from the sparse coding representation with the statistics from the input signal. Third, the classification subspace is hierarchically partitioned with a tree structure to achieve a lightweight model size with good prediction accuracy. The experimental results demonstrate the effectiveness and efficiency in classifying the modulated features and representation of received signals. Compared to lightweight deep learning models, the number of model parameters is reduced by 41%, while the usage of Floating Point Operations (FLOPs) is only $\mathcal{O}(10^{-4})$ of the blind waveform recognition without pre-arranged knowledge of incoming waveforms.

Reliable, affordable electricity remains inaccessible to over 600 million people in sub-Saharan Africa (SSA), where islanded hybrid microgrids combining renewable generation, battery storage, and diesel backup offer a viable electrification pathway. This paper presents a multiobjective, multiperiod optimization framework for the design, sizing, and dispatch of such systems, with a case study for a remote community in Kenya. System sizing is optimized over a one-year horizon and dispatch over a representative day, both at hourly resolution. The formulation jointly minimizes lifecycle levelized cost of energy (LCOE), emissions, lost load, and dumped energy, while maximizing renewable penetration. Seven optimization algorithms are benchmarked; particle swarm optimization (PSO) achieves the best trade-off between runtime (63 s) and solution quality (normalized objective 0.146) and is used for subsequent analyses. The optimal configuration of solar PV, wind, lithium-ion battery storage, and diesel backup achieves a normalized LCOE of 0.46 USD per kWh with over 94 percent renewable penetration, outperforming alternatives. Pareto fronts highlight trade-offs between cost, emissions, and reliability, showing that cost-only optimization yields inferior outcomes. Sensitivity analyses identify fuel prices and discount rates as the most influential parameters in SSA contexts. A break-even distance analysis shows microgrids are economically competitive with grid extension at the study site. The dispatch model produces day-ahead schedules that are robust to short-term uncertainty, though extended wind lulls increase diesel reliance. This work fills a critical gap by providing a comprehensive multiobjective design and dispatch framework tailored to SSA resource, economic, and operational conditions.

Decarbonising heavy-duty railway networks requires maximising the capacity of existing electrical infrastructure. Integrating heavy freight alongside fast passenger services exposes the hard physical limits of conventional AC traction networks, causing severe localised power quality degradation, phase unbalance, and low-voltage behaviour that triggers protective substation tripping. Because hardware upgrades are highly capital-intensive, software-based Energy Management Strategies (EMS) offer a potentially viable alternative. This systematic review synthesises the literature on algorithmic energy management for grid-constrained multi-train AC railway networks, classifying the reviewed studies along three axes: algorithm family, operational scope, and constraint coupling. Three findings emerge consistently. First, single-train trajectory optimisation, however mathematically refined, cannot represent the coupled electrical interactions that increasingly define network capacity on mixed-traffic networks. Second, while multi-train Train-Track-Power (TTP) simulations capture these interactions, the algorithm families used to solve them face well-documented trade-offs between computational tractability and constraint flexibility; predictive and distributed methods, including hierarchical model predictive control and decomposition-based schemes, narrow this trade-off substantially, but electrical fidelity and network-scale real-time operation have been demonstrated largely in separate studies rather than together. Third, the literature increasingly identifies a gap between mathematically optimal speed profiles and operationally executable ones, particularly on networks operated by human drivers rather than Automatic Train Operation systems. The review delineates where current methods succeed, where they fail, and which directions the literature has identified as open.

Long-term beamforming (LTBF) is a widely-used scalable alternative to instantaneous multi-user MIMO processing that leverages slowly varying spatial channel statistics. VLSI implementations require matrix inversion that become computationally challenging for massive MIMO systems with large number of antennas. In this work, we show that dominant interferers significantly degrade the numerical conditioning of the LTBF covariance matrix, leading to severe performance loss in finite-precision implementations of polynomial and conjugate gradient (CG) based inversion methods. To address this issue, we propose a subspace nulling approach that operates solely on long-term channel statistics and acts as an implicit preconditioning step for LTBF. By projecting the received signal onto the orthogonal complement of the dominant interference subspace, the proposed method reduces the eigenvalue spread of the covariance matrix and improves numerical stability. Through ray-tracing simulations in a realistic 5G scenario, we demonstrate that the proposed method substantially reduces the number of CG iterations required to achieve near-optimal performance across floating-point and fixed-point implementations while preserving the low-overhead nature of LTBF.

Accurate dietary assessment is critical for precision nutrition, yet most image-based methods rely on a single pre-consumption image and provide only coarse, meal-level estimates. These approaches cannot determine what was actually consumed and often require restrictive inputs such as depth sensing, multi-view imagery, or explicit segmentation. In this paper, we propose a simple vision-language framework for food-item-level nutritional analysis using paired before-and-after eating images. Instead of relying on rigid segmentation masks, our method leverages natural language prompts to localize specific food items and estimate their weight directly from a single RGB image. We further estimate food consumption by predicting weight differences between paired images using a two-stage training strategy. We evaluate our method on three publicly available datasets and demonstrate consistent improvements over existing approaches, establishing a strong baseline for before-and-after dietary image analysis.

This paper develops a sparsity-promoting integral concurrent learning (SP-ICL) adaptation law for a linearly parametrized uncertain nonlinear control-affine systems. The unknown parameters are learned using ICL with sparsity-promoting $\ell_1$ regularization. The use of $\ell_1$ regularization for sparsity promotion is common in system identification and machine learning; however, unlike existing approaches, this paper develops an online parameter update law that integrates the regularization penalty with ICL via sliding modes. We show via non-smooth Lyapunov analysis that the trajectories of the closed-loop system are ultimately bounded under the SP-ICL update law. Simulations verify the effectiveness of the sparsity penalty in the SP-ICL update law on recovering sparse dynamics during trajectory tracking.

This paper first demonstrates that applying standard incremental nonlinear dynamic inversion (INDI) with incremental control allocation (ICA) to input nonaffine systems relies on an untenable linear approximation of the actuator model. It then shows that avoiding this issue, while retaining the static control allocation paradigm, generally requires solving a nonlinear programming (NLP) problem. To address the associated online computational challenges, the paper subsequently presents a supervised learning-based approach. Numerical experiments on an example problem validate the identified limitations of standard INDI + ICA for input nonaffine systems, while also demonstrating that the proposed learning-based method provides an effective and computationally tractable alternative.

This paper studies the co-design of actuators, sensors, and communication in the distributed setting, where a networked plant is partitioned into subsystems with sub-controllers interacting with other sub-controllers. The objective is to jointly minimize control cost (i.e., LQ cost) and material cost (i.e., number of actuators, sensors, and communication links used). We approach this using an evolutionary algorithm that selectively prunes a baseline dense LQR controller, and provide convergence and stability analyses. Our approach is validated in simulations; it substantially outperforms naive pruning (over 80% in most cases), and performs similarly as greedy pruning but with much-improved scalability.

We present an empirical case study of cinema SDR-to-HDR mapping using ASC StEM2, a rare common-source dataset containing EXR scene-referred images and matched SDR/HDR cinema release masters from the same ACES-based mastering workflow. Based on pixel-wise statistics over all 18,580 frames of the test film, we construct a three-domain comparison involving EXR source data, SDR release masters, and HDR release masters to characterize their luminance and color structural relationships within this controlled workflow. In the luminance dimension, SDR and HDR masters exhibit a highly stable global monotonic correspondence, with geometric structure remaining largely consistent overall; sparse and structured deviations appear in self-luminous highlights and specific material regions. In the color dimension, the two masters remain largely consistent in hue, with saturation exhibiting a redistribution pattern of shadow suppression, midtone expansion, and highlight convergence. Using EXR as a scene-referred anchor, we further define a pixel-level decision map that operationally separates EXR-closer recovery regions from content-adaptive adjustment regions. Under this operational definition, 82.4% of sampled image regions are classified as EXR-closer recovery, while the remainder require localized adaptive adjustment. Rather than claiming a universal law for all cinema mastering pipelines, the study provides an interpretable quantitative baseline for structure-aware SDR-to-HDR analysis and for designing learning-based models under shared-source mastering conditions.

We study what is, to our knowledge, the first tractable multistage ex-ante distributionally robust regret optimization (DRRO) formulation for stochastic control. We consider finite-horizon LQR with common stage-law ambiguity, where disturbances are independent across time but drawn from the same unknown stage law whose mean and covariance lie in a Gelbrich ball around nominal moments. Unlike the benign single-stage quadratic setting, the nominal controller is generally not regret-optimal: reuse of the stage law makes past disturbances informative for future decisions. Despite the general hardness of DRRO, we show that, over affine disturbance-feedback policies, the multistage DRRO-LQR problem admits an exact semidefinite programming reformulation. An optimal controller in this class is the nominal LQR controller plus a strictly causal empirical-mean correction. We also characterize worst-case moment pairs and show that, for the DRRO-optimal policy, they are not unique. Portfolio liquidation experiments show that DRRO substantially reduces worst-case regret relative to DRO and the nominal controller, with comparatively modest increases in worst-case cost, and exhibits a learning effect: its correction matrices empirically approach the corresponding coefficients of the oracle controller that knows the true disturbance law in hindsight.

This paper presents a theoretical convergence analysis for an opinion-action coevolution model that integrates the opinion updating rule of the Hegselmann-Krause model with a utility-based decision-making mechanism. The model is reformulated into an augmented state-space representation, where the state matrix induces a time-varying social interaction digraph. The convergence analysis is grounded on two existing theoretical findings that establish convergence for the Hegselmann-Krause type of models and containment control systems with multiple stationary leaders, respectively. Results indicate that, if the structure of the interaction digraph stabilizes within finite time, the model either converges to consensus, where all agents' opinions and actions reach an identical state, or exhibits clustering, where some opinion nodes act as stationary leaders while the remaining nodes approach the convex hull formed by the leaders. Numerical simulations are then provided to validate the theoretical results.

In this paper, we study four-strategy conservative replicator dynamics induced by constant payoff matrices. We establish necessary and sufficient conditions for coexistence to occur by associating the payoff matrix with its digraph, revealing exactly five distinct digraph classes governing the global behavior. We further show that, whenever the dynamics is coexistent, every non-equilibrium trajectory in the relative interior of the simplex is a Lyapunov-stable periodic orbit. Together with the classification of the boundary phase portraits, these results provide a complete characterization of the global dynamics in the four-strategy case with coexistence.

Electric vertical takeoff and landing (eVTOL) aircraft operating in high-density urban airspace must maintain safe separation through tactical conflict resolution, yet the energy cost of such maneuvers has not been systematically quantified. This paper investigates how conflict-resolution maneuvers under the Modified Voltage Potential (MVP) algorithm affect eVTOL energy consumption. Using a physics-based power model integrated within a traffic simulation, we analyze approximately 71,767 en route sections within a sector, across traffic densities of 10-60 simultaneous aircraft. The main finding is that MVP-based deconfliction is energy-efficient: median energy overhead remains below 1.5% across all density levels, and the majority of en route flights within the sector incur negligible penalty. However, the distribution exhibits pronounced right-skewness, with tail cases reaching 44% overhead at the highest densities due to sustained multi-aircraft conflicts. The 95th percentile ranges from 3.84% to 5.3%, suggesting that a 4-5% reserve margin accommodates the vast majority of tactical deconfliction scenarios. To support operational planning, we develop a machine learning model that estimates energy overhead at mission initiation. Because conflict outcomes depend on future traffic interactions that cannot be known in advance, the model provides both point estimates and uncertainty bounds. These bounds are conservative; actual outcomes fall within the predicted range more often than the stated confidence level, making them suitable for safety-critical reserve planning. Together, these results validate MVP's suitability for energy-constrained eVTOL operations and provide quantitative guidance for reserve energy determination in Advanced Air Mobility.

This paper investigates the leader-following consensus problem for a class of multi-agent systems subject to adversarial attack-like external inputs. To address this, we formulate the robust leader-following control problem as a global coalitional min-max zero-sum game using differential game theory. Specifically, the agents' control inputs form a coalition to minimize a global cost function, while the attacks form an opposing coalition to maximize it. Notably, when these external adversarial attacks manifest as disturbances, the designed game-theoretic control policy systematically yields a robust $H_\infty$ control law. Addressing this problem inherently requires solving a high-dimensional generalized algebraic Riccati equation (GARE), which poses significant challenges for distributed computation and controller implementation. To overcome these challenges, we propose a two-fold approach. First, a decentralized computational strategy is devised to decompose the high-dimensional GARE into multiple uniform, lower-dimensional GAREs. Second, a dynamic average consensus-based decoupling algorithm is developed to resolve the inherent coupling structure of the robust control law, thereby facilitating its distributed implementation. Finally, numerical simulations on the formation control of multi-vehicle systems with feedback-linearized dynamics are conducted to validate the effectiveness of the proposed algorithms.

In this work, we study a discrete Schrödinger bridge problem with partial marginal observations. A main difficulty compared to the classical Schrödinger bridge formulation is that our problem is not strictly convex and standard Sinkhorn-type methods cannot be directly applied. To address this issue, we propose a scalable computational method based on an entropic proximal scheme. Furthermore, we develop a framework for this problem that includes duality results, characterization of the optimal solutions, and an observability condition that determines when the optimal solution is unique. We validate the method on the problem of estimating contamination in a water distribution network, where the partial marginals correspond to measured pollutant concentrations at the sensor locations. The experiments were conducted on a laboratory-scale water distribution network.

We propose a cooperative sensing framework for mmWave ISAC networks in which a target is sensed by its nearest BS while opportunistically exploiting bistatic echoes from neighboring BSs. Cooperation requires no dedicated resources or exchange of sensing results, and is realized via non-coherent echo-power combining. Using stochastic geometry, we characterize sensing/communication coverage and rates and, for the first time, the cooperative sensing meta-distribution to quantify reliability across targets. Results show substantial sensing gains with limited communication loss and improved high-reliability tail, increasing the fraction of targets meeting stringent reliability guarantees crucial for safety-critical applications.

We consider event-triggered linear-quadratic Gaussian (LQG) control when sensor updates are transmitted over an i.i.d. packet-erasure channel. Although the optimal controller in a standard LQG setup is available in closed form, choosing when to transmit remains computationally and analytically difficult because packet drops randomize packet delivery and couple scheduling decisions with the estimation-error dynamics, making direct dynamic-programming solutions impractical. By certainty equivalence, the co-design problem becomes choosing a binary send/skip sequence that balances control performance and communication cost. We derive a closed-form expansion of the error covariance as precomputable Gramian terms scaled by a survival factor that depends only on the number of transmission attempts on each interval. This converts the problem into an unconstrained binary program that we linearize exactly via running attempt counters and a one-hot encoding, yielding a compact MILP well suited to receding-horizon implementation. On the linearized Boeing-747 benchmark, a model predictive control (MPC) scheduler lowers cost while attempting far fewer transmissions than a one-shot baseline across channel success rates.

Safety-critical control of uncertain, adaptive systems often relies on conservative, worst-case uncertainty bounds that limit closed-loop performance. Online conformal prediction is a powerful data-driven method for quantifying uncertainty when truth values of predicted outputs are revealed online; however, for systems that adapt the dynamics without measurements of the state derivatives, standard online conformal prediction is insufficient to quantify the model uncertainty. We propose Staggered Integral Online Conformal Prediction (SI-OCP), an algorithm utilizing an integral score function to quantify the lumped effect of disturbance and learning error. This approach provides long-run coverage guarantees, resulting in long-run safety when synthesized with safety-critical controllers, including robust tube model predictive control. Finally, we validate the proposed approach through a numerical simulation of an all-layer deep neural network (DNN) adaptive quadcopter using robust tube (MPC), highlighting the applicability of our method to complex learning parameterizations and control strategies.

This letter studies pilot design for orthogonal frequency-division multiplexing-based time-division duplex (TDD) systems under a sliding-window latest-slot recovery framework that jointly exploits delay-Doppler sparsity across recent slots. Under contiguous-subband and fairness constraints, this viewpoint naturally leads to a geometry-aware time-frequency joint pilot assignment. We show that effective patterns should balance grid coverage and redundant-collinearity suppression, with an additional symmetry-avoidance refinement when complete collinearity elimination is infeasible. Based on these principles, we formulate a mixed-integer construction method compatible with practical TDD allocation. Numerical results show that minimum-coverage-radius and collinearity-control (MCC) pattern improves both surrogate geometry metrics and latest-slot recovery performance.

This paper investigates the design of chirp-layered superposition coding for LoRa, where an additional waveform is linearly superposed on a standard LoRa transmission with minimal impact on the LoRa demodulation process. We first show that any non-zero superposed signal perturbs the output of the standard dechirp-and-DFT demodulator, and then characterize the class of superposed waveforms that minimize this degradation under a given power budget. In particular, we show that a high spreading factor (high-SF) LoRa waveform superposed on a low-SF signal (e.g., SF12 on SF7) can be designed so that its impact on the standard LoRa demodulation remains small. As a result, within each low-SF symbol interval, the high-SF segment can be treated as a quasi-narrowband carrier that conveys an additional BPSK stream. We derive analytical error-rate expressions for both the low-SF LoRa layer and the superposed high-SF layer, and validate them through Monte Carlo simulations. The proposed chirp-layered superposition coding scheme improves the spectral efficiency of LoRa-based links and uses a relatively simple transceiver architecture.

We develop a framework for studying and quantifying the risk of cascading failures in time-delay consensus networks, motivated by a team of agents attempting temporal rendezvous under stochastic disturbances and communication delays. To assess how failures at one or multiple agents amplify the risk of deviation across the network, we employ the Average Value-at-Risk as a systemic measure of cascading uncertainty. Closed-form expressions reveal explicit dependencies of the risk of cascading failure on the Laplacian spectrum, communication delay, and noise statistics. We further establish fundamental lower bounds that characterize the best-achievable network performance under time-delay constraints. These bounds serve as feasibility certificates for assessing whether a desired safety or performance goal can be achieved without exhaustive search across all possible topologies. In addition, we develop an efficient single-step update law that enables scalable propagation of conditional risk as new failures are detected. Analytical and numerical studies demonstrate significant computational savings and confirm the tightness of the theoretical limits across diverse network configurations.

From a maneuverability perspective, the main advantage of tilting multirotor UAVs lies in the dynamic variability of the feasible executable wrench, which represents a key asset for physical interaction tasks. Accordingly, cant-angle selection should be optimized to ensure high performance while avoiding abrupt variations and preserving real-world feasibility. In this context, this work proposes a lightweight control framework for star-shaped interdependent cant-tilting hexarotor UAVs performing interaction tasks. The method uses an offline-computed look-up table of zero-moment force polytopes to identify feasible cant angles for a desired control force and select the optimal one by balancing efficiency and smoothness. The framework is integrated with a geometric full-pose controller and validated through Monte Carlo simulations in MATLAB/Simulink and compared against a baseline strategy. The results show a significant reduction in computation time, together with improved pose-tracking performance and competitive actuation efficiency. A final physics-based simulation of a complete wall inspection task in Simscape further confirms the feasibility of the proposed strategy in interacting scenarios.

Realistic modeling of scattering from curved metallic bodies - such as vehicles and roadside structures - is essential for cellular and vehicular channel modeling as well as radar applications. A practical approach is to approximate curved surfaces with planar facets and apply ray-tracing with diffraction methods; however, accuracy depends critically on both geometric discretization and diffraction modeling. This work investigates ray-tracing-based modeling of near-field scattering from curved bodies, both in the backscattering and in the forward (shadow) region; in the ray-tracing tool, diffraction is modeled according to the Uniform Theory of Diffraction (UTD), extended with vertex diffraction and double-bounce interactions, including a heuristic combination of edge and vertex diffraction. A discretization strategy linking facet size to local curvature and wavelength is proposed to balance geometric fidelity, diffraction modeling, and efficiency. Validation is initially performed against analytical solutions and full-wave simulations for canonical geometries (sphere and circular cylinder). Furthermore, the practical applicability of the approach is demonstrated for a realistic vehicle by comparison with bistatic measurements in the backscattering region and full-wave simulation in the shadow region. The results demonstrate that no universal discretization strategy exists: fine meshes are beneficial for accurate backscattering prediction, while coarser discretizations can provide more efficient and accurate shadow region prediction. The proposed extended diffraction framework provides a computationally efficient framework for vehicular propagation and integrated sensing and communication (ISAC) channel modeling.

This paper addresses practical tracking control for robotic vehicles with pivoted unidirectional actuators. A widely recognized obstacle to tracking control for multirotors is the free-fall singularity — when commanded to free-fall, the vehicle's attitude target becomes undefined. This singularity is a pathology pertinent to all robotic vehicles with pivoted and gimbaled unidirectional actuators. This paper addresses the planar instance of the general theoretical problem: the case of a vehicle with pivoted unidirectional actuation. Starting from a baseline robust controller that assumes unconstrained inputs, we redesign the control law to be compatible with the pivoted actuator. This is accomplished by driving the output of the pivoted actuator to a ball centered at the target input value. The baseline controller's guarantees are recovered in a practical sense. The theory is illustrated with a simulation example.

Incentive design constitutes a foundational paradigm for influencing the behavior of strategic agents, wherein a system planner (principal) publicly commits to an incentive mechanism designed to align individual objectives with collective social welfare. This paper introduces the Regret-Minimizing Adaptive Incentive Design (RAID) problem, which aims to synthesize incentive laws under information asymmetry and achieve asymptotically minimal regret compared to an oracle with full information. To this end, we develop the RAID algorithm, which employs a switching policy alternating between probing (exploration) and estimate-based incentivization (exploitation). The associated type estimator relies only on a weaker excitation condition required for strong consistency in least squares estimation, substantially relaxing the persistence-of-excitation assumptions previously used in adaptive incentive design. In addition, we establish the strong consistency of the proposed type estimator and prove that the incentive obtained asymptotically minimizes the planner's average regret almost surely. Numerical experiments illustrate the convergence rate of the proposed methodology.

We provide a practical relaxation of Willems' fundamental lemma for discrete-time linear time-invariant (single-input-single-output) systems. Instead of maintaining conventional Willems' persistency of excitation condition in the behavioral theory, we reformulate the problem in terms of signal generators, hence going back to the dynamical systems theory. We discuss the relationship between the persistency of excitation order and the dimension of the signal generator. Furthermore, we identify a necessary and sufficient condition on the signal generator that can generate informative input–output data for almost all systems and initial conditions. This formulation accommodates both a broad range of conventional persistently exciting input designs and a class of shorter inputs that violate Willems' condition, including sinusoidal sequences with fewer frequencies. Finally, the signal generator perspective allows a natural extension to continuous-time systems.

Metal artifacts from high-attenuation implants severely degrade CT image quality, obscuring critical anatomical structures and posing a challenge for standard deep learning methods that require extensive paired training data. We propose a paradigm shift: reframing artifact reduction as an in-context reasoning task by adapting a general-purpose vision-language diffusion foundation model via parameter-efficient Low-Rank Adaptation (LoRA). By leveraging rich visual priors, our approach achieves effective artifact suppression with only 16 to 128 paired training examples reducing data requirements by two orders of magnitude. Crucially, we demonstrate that domain adaptation is essential for hallucination mitigation; without it, foundation models interpret streak artifacts as erroneous natural objects (e.g., waffles or petri dishes). To ground the restoration, we propose a multi-reference conditioning strategy where clean anatomical exemplars from unrelated subjects are provided alongside the corrupted input, enabling the model to exploit category-specific context to infer uncorrupted anatomy. Extensive evaluation on the AAPM CT-MAR benchmark demonstrates that our method achieves state-of-the-art performance on perceptual and radiological-feature metrics . This work establishes that foundation models, when appropriately adapted, offer a scalable alternative for interpretable, data-efficient medical image reconstruction. Code is available at https://github.com/ahmetemirdagi/CT-EditMAR.

Gray-box RC models are widely used to enable energy-efficient model predictive control (MPC) in buildings. However, estimating RC parameters remains difficult, as conventional optimization-based algorithms are prone to local minima, rely heavily on good initial guesses, and incur high computational cost. To address these issues, we propose the Estimator from Scratch, a novel neural parameter estimation approach that embeds the physical equations into a neural network's training process to estimate RC parameters. To further improve estimation accuracy and eliminate dependence on an initial guess, we extend this approach by pretraining the neural network on data from multiple source buildings, the Pretrained Estimator. We benchmark both methods against a genetic-algorithm-based RC estimator and a fully black-box neural network. All methods are evaluated across eight simulated and three real-world buildings for two RC configurations, for both prediction accuracy and closed-loop MPC performance, the latter only for the simulated buildings. The Pretrained Estimator achieves the best prediction performance among all RC-based methods, particularly with little training data, and yields the lowest and least variable MPC costs across buildings and benchmarks. These results position the Pretrained Estimator as a robust, computationally efficient, and initial-guess-free alternative for RC parameter estimation, with potential to extend to other control-oriented dynamical systems.

We study local sensitivity of structured ARX-based data-driven predictive control. Although predictor estimation is linear in the ARX parameters, the lifted multi-step predictor used in MPC depends on them implicitly, which complicates both uncertainty propagation and task-aware regularization. We derive a local first-order linearization of this implicit predictor map. The resulting Jacobian yields both an approximate control-relevant prediction uncertainty term and a task-dependent sensitivity metric for shaping kernel regularization. Numerical results show that the proposed analysis is most useful in weak-excitation regimes, where baseline SS regularization already provides substantial robustness gains and the proposed sensitivity shaping yields a further smaller improvement.

Direct collocation for Bolza optimal control yields discrete Karush-Kuhn-Tucker (KKT) points, while practical solvers expose only discrete quantities such as primal-dual iterates, reduced Hessians, and Jacobians. This creates a gap between continuous second-order optimality theory and what can be certified from solver output. We develop an a posteriori certification framework that bridges this gap. Starting from a discrete KKT solution, we reconstruct piecewise polynomial state, control, and costate trajectories, evaluate residuals of the dynamics, boundary, and stationarity conditions, and derive a computable lower bound for the continuous second variation. The bound is expressed as the discrete reduced curvature minus explicit residual-dependent correction terms. A positive bound yields a sufficient certificate for continuous second-order sufficiency and provides quantitative information relevant to local growth and trust-region sizing. The constants entering the certification inequality are conservatively estimable from reconstructed discrete data. The resulting test is operationally verifiable from collocation outputs and naturally supports adaptive mesh refinement through residual decomposition. We also outline an extension to path inequalities with isolated transversal switches.