GPU-Accelerated Verification of Machine Learning Models for Power Systems

Computational tools for rigorously verifying the performance of large-scale machine learning (ML) models have progressed significantly in recent years. The most successful solvers employ highly specialized, GPU-accelerated branch and bound routines. Such tools are crucial for the successful deployment of machine learning applications in safety-critical systems, such as power systems. Despite their successes, however, barriers prevent out-of-the-box application of these routines to power system problems. This paper addresses this issue in three key ways. First, we reformulate several key power system verification problems into the canonical format utilized by modern verification solvers. Second, we enable the simultaneous verification of multiple verification problems (e.g., checking for the violation of all constraints simultaneously, and not by solving individual verification problems). To achieve this, we introduce an exact transformation that converts a set of potential violations into a series of ReLU-based neural network layers. This allows verifiers to interpret these layers directly, and determine the “worst-case” violation in a single shot. Third, power system ML models often must be verified to satisfy power flow constraints. We propose a dualization procedure which encodes linear equality and inequality constraints (such as power balance constraints and line flow constraints) directly into the verification problem in a manner which is mathematically consistent with the specialized verification tools. To demonstrate these innovations, we verify problems associated with data-driven security constrained DC-OPF solvers. We build and test our first set of innovations using the α, β-CROWN solver, and we benchmark against Gurobi 10.0. Our contributions achieve a speedup that can exceed 100x and allow higher degrees of verification flexibility.