Exact Constraint Enforcement in Physics-Informed Extreme Learning Machines using Null-Space Projection Framework

2026-01-16

AI Summary

This work proposes NP-PIELM, a novel approach to physics-informed extreme learning machines (PIELM) that overcomes the limitations of traditional penalty-based formulations, which suffer from limited accuracy and high sensitivity to hyperparameters due to approximate enforcement of boundary and initial conditions. By introducing a nullspace projection in the coefficient space, NP-PIELM reformulates the constrained optimization problem as an unconstrained least-squares problem, enabling exact satisfaction of boundary conditions at discrete collocation points. This method achieves strict enforcement of boundary constraints within the PIELM framework for the first time—without requiring penalty parameters, dual variables, or problem-specific constructions—while preserving the single-shot training efficiency inherent to ELMs. Numerical experiments demonstrate that NP-PIELM delivers high accuracy and robustness across elliptic and parabolic partial differential equations, complex geometries, and mixed boundary conditions.

Technology Category

• Constraint Satisfaction and Optimization: Constraint Learning and Acquisition
• Machine Learning: Learning with Manifolds
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems

Abstract

Physics-informed extreme learning machines (PIELMs) typically impose boundary and initial conditions through penalty terms, yielding only approximate satisfaction that is sensitive to user-specified weights and can propagate errors into the interior solution. This work introduces Null-Space Projected PIELM (NP-PIELM), achieving exact constraint enforcement through algebraic projection in coefficient space. The method exploits the geometric structure of the admissible coefficient manifold, recognizing that it admits a decomposition through the null space of the boundary operator. By characterizing this manifold via a translation-invariant representation and projecting onto the kernel component, optimization is restricted to constraint-preserving directions, transforming the constrained problem into unconstrained least-squares where boundary conditions are satisfied exactly at discrete collocation points. This eliminates penalty coefficients, dual variables, and problem-specific constructions while preserving single-shot training efficiency. Numerical experiments on elliptic and parabolic problems including complex geometries and mixed boundary conditions validate the framework.

Problem

• Physics-informed extreme learning machines
• boundary conditions
• initial conditions
• constraint enforcement
• penalty methods

Innovation

• null-space projection
• physics-informed extreme learning machine
• exact constraint enforcement
• coefficient manifold
• unconstrained least-squares