A JAX-based differentiable model of pressure vacuum swing adsorption accelerates cyclic steady-state simulation by 20x via Newton iteration and produces a better Pareto front with IPOPT than NSGA-II in two orders of magnitude less time on a post-combustion capture benchmark.
hub
On the implementation of an interior-point filter line-search algorithm for large-scale nonlinear programming
20 Pith papers cite this work, alongside 7,875 external citations. Polarity classification is still indexing.
20
Pith papers citing it
7,875
external citations · Crossref
hub tools
citation-role summary
background 2
method 2
citation-polarity summary
representative citing papers
An approximate IPTR framework for linearly constrained optimization uses low-rank projector updates to cut per-iteration cost while preserving feasibility and convergence guarantees, with experiments showing 2.48x speedup.
Four Hessian-informed trust-region filter variants using low- and high-fidelity surrogates reduce iterations and black-box evaluations by up to an order of magnitude on 25 benchmarks and five engineering cases while lowering tuning sensitivity.
A framework for directly optimizing proton treatment plans using an extended probability of lesion origin (POLO) model in low-grade glioma patients.
The paper presents ModelPredictiveControl.jl, an open-source Julia toolkit for model predictive control including nonlinear, economic, and successive linearization variants, illustrated with CSTR and inverted pendulum simulations and benchmarked against MATLAB.
SupPy toolbox applies superiorization to three physics problems and shows improved outcomes over feasibility-seeking alone, including viable plans on infeasible sets.
DiffSlack introduces learnable slack variables and a damped Gauss-Newton projection to create a differentiable layer that enforces hard nonlinear inequality constraints in neural network outputs.
An automated microliter-scale multi-dataset INST 13C-MFA workflow in C. glutamicum on ethanol yields robust net fluxes but variable pool sizes, with the glyoxylate shunt playing a central role.
A WLaSDI-based framework creates noise-robust latent surrogates for PDE-constrained optimization, deriving direct and adjoint gradients to achieve up to five orders of magnitude speedup on radiative transfer, Vlasov-Poisson, and Burgers benchmarks.
A Riemannian L-BFGS method with adapted Cauchy-point bound handling outperforms classical interior-point and L-BFGS-B solvers on mixed manifold-plus-bounds problems by orders of magnitude.
A complete open-data pipeline reconstructs realistic US transmission grids from OSM and EIA sources and produces publicly released models that solve AC-OPF for 88% of single-state cases at peak load.
Piecewise M-stationarity is equivalent to B-stationarity for MPCCs under MPCC-ACQ and reduces the cost of verifying stationarity for NCP-based algorithms.
A fully discrete strain-based model for continuum robot dynamics via Lie group variational integrators, combined with an EKF-based observer for states and disturbances, validated on hardware.
Hybrid ME-DDP variants combine deterministic DDP with inverse-Hessian sampling to improve success rates over pure DDP and MPPI in robotic navigation under non-convex costs.
A filter line search SQP algorithm reduces iterations and computation time for nonconvex SOS programs compared to prior methods.
A trust-region funnel algorithm for gray-box optimization achieves global convergence to first-order critical points and performs comparably or better than the classical trust-region filter method.
Converting an 800k-line C++ mathematical library to C++20 modules is feasible with moderate effort and yields compile-time savings inside the library but no clear trend for downstream users.
Overlapping Schwarz decomposition for nonlinear OCPs achieves local linear convergence with rate improving exponentially with overlap size, based on exponential decay of sensitivity for primal and dual solutions.
An open-sourced Unified Autonomy Stack fuses LiDAR, radar, vision and inertial data with sampling-based planning and control barrier functions to deliver resilient autonomy on aerial and ground robots in challenging real-world settings.
A hybrid MILP-NLP-complementarity decomposition solved via spatial/temporal ADMM yields up to 13x speedup on unbalanced AC power flow-constrained DES design for networks with 55 loads, with maximum 0.61% optimality gap.
citing papers explorer
-
Accelerating Simulation and Optimisation of Cyclic Adsorption Processes with Differentiable Programming
A JAX-based differentiable model of pressure vacuum swing adsorption accelerates cyclic steady-state simulation by 20x via Newton iteration and produces a better Pareto front with IPOPT than NSGA-II in two orders of magnitude less time on a post-combustion capture benchmark.
-
Scalable First-Order Interior Point Trust Region Algorithms for Linearly Constrained Optimization
An approximate IPTR framework for linearly constrained optimization uses low-rank projector updates to cut per-iteration cost while preserving feasibility and convergence guarantees, with experiments showing 2.48x speedup.
-
Trust-region filter algorithms utilizing Hessian information for gray-box optimization
Four Hessian-informed trust-region filter variants using low- and high-fidelity surrogates reduce iterations and black-box evaluations by up to an order of magnitude on 25 benchmarks and five engineering cases while lowering tuning sensitivity.
-
Direct optimization of the probability of lesion origin in proton treatment planning for low-grade glioma patients
A framework for directly optimizing proton treatment plans using an extended probability of lesion origin (POLO) model in low-grade glioma patients.
-
ModelPredictiveControl.jl: advanced process control made easy in Julia
The paper presents ModelPredictiveControl.jl, an open-source Julia toolkit for model predictive control including nonlinear, economic, and successive linearization variants, illustrated with CSTR and inverted pendulum simulations and benchmarked against MATLAB.
-
GPU-accelerated superiorization on constrained physical problems with SupPy
SupPy toolbox applies superiorization to three physics problems and shows improved outcomes over feasibility-seeking alone, including viable plans on infeasible sets.
-
DiffSlack: Learning under Nonlinear Inequality Constraints via Learnable Slack Variables
DiffSlack introduces learnable slack variables and a damped Gauss-Newton projection to create a differentiable layer that enforces hard nonlinear inequality constraints in neural network outputs.
-
Automated multi-dataset INST $^{13}$C metabolic flux analysis at microliter scale reveals robust fluxes but variable metabolite pools in $Corynebacterium~glutamicum$
An automated microliter-scale multi-dataset INST 13C-MFA workflow in C. glutamicum on ethanol yields robust net fluxes but variable pool sizes, with the glyoxylate shunt playing a central role.
-
Time-Dependent PDE-Constrained Optimization via Weak-Form Latent Dynamics
A WLaSDI-based framework creates noise-robust latent surrogates for PDE-constrained optimization, deriving direct and adjoint gradients to achieve up to five orders of magnitude speedup on radiative transfer, Vlasov-Poisson, and Burgers benchmarks.
-
A Riemannian quasi-Newton algorithm for optimization with Euclidean bounds
A Riemannian L-BFGS method with adapted Cauchy-point bound handling outperforms classical interior-point and L-BFGS-B solvers on mixed manifold-plus-bounds problems by orders of magnitude.
-
Building Power Grid Models from Open Data: A Complete Pipeline from OpenStreetMap to Optimal Power Flow
A complete open-data pipeline reconstructs realistic US transmission grids from OSM and EIA sources and produces publicly released models that solve AC-OPF for 88% of single-state cases at peak load.
-
Piecewise M-Stationarity and Related Algorithms for Mathematical Programs with Complementarity Constraints
Piecewise M-stationarity is equivalent to B-stationarity for MPCCs under MPCC-ACQ and reduces the cost of verifying stationarity for NCP-based algorithms.
-
Discrete Geometric Modeling and Extended State Estimation of Continuum Robots
A fully discrete strain-based model for continuum robot dynamics via Lie group variational integrators, combined with an EKF-based observer for states and disturbances, validated on hardware.
-
Beyond Pure Sampling: Hybrid Optimization Mechanisms for Non-Convex Model Predictive Control
Hybrid ME-DDP variants combine deterministic DDP with inverse-Hessian sampling to improve success rates over pure DDP and MPPI in robotic navigation under non-convex costs.
-
On the Practical Implementation of a Sequential Quadratic Programming Algorithm for Nonconvex Sum-of-squares Problems
A filter line search SQP algorithm reduces iterations and computation time for nonconvex SOS programs compared to prior methods.
-
A trust-region funnel algorithm for gray-box optimization
A trust-region funnel algorithm for gray-box optimization achieves global convergence to first-order critical points and performs comparably or better than the classical trust-region filter method.
-
Experience converting a large mathematical software package written in C++ to C++20 modules
Converting an 800k-line C++ mathematical library to C++20 modules is feasible with moderate effort and yields compile-time savings inside the library but no clear trend for downstream users.
-
On the Convergence of Overlapping Schwarz Decomposition for Nonlinear Optimal Control
Overlapping Schwarz decomposition for nonlinear OCPs achieves local linear convergence with rate improving exponentially with overlap size, based on exponential decay of sensitivity for primal and dual solutions.
-
The Unified Autonomy Stack: Toward a Blueprint for Generalizable Robot Autonomy
An open-sourced Unified Autonomy Stack fuses LiDAR, radar, vision and inertial data with sampling-based planning and control barrier functions to deliver resilient autonomy on aerial and ground robots in challenging real-world settings.
-
Distributed Energy System Design including Unbalanced AC Power Flow for Large LV Networks with ADMM
A hybrid MILP-NLP-complementarity decomposition solved via spatial/temporal ADMM yields up to 13x speedup on unbalanced AC power flow-constrained DES design for networks with 55 loads, with maximum 0.61% optimality gap.