A computational study of Ising‐based solvers for discrete landscape exploration in process optimization

ABSTRACT Conceptual process design combines discrete configuration choices with continuous operating decisions, often yielding difficult mixed‐integer nonlinear or simulation‐based optimization problems. This work presents an exploratory computational assessment of Ising‐based solvers, simulated annealing, quantum annealing, and entropy computing, as candidate methods for discrete process design subproblems within a sequential process systems engineering workflow. Two case studies, an ionic‐liquid reactor‐separator network and a drug‐substance manufacturing process, compare these methods with deterministic branch‐and‐bound strategies. Deterministic methods rapidly recover optimal discrete solutions, whereas Ising‐based solvers generate distributions of feasible candidate configurations. Downstream evaluation further shows that discrete‐subproblem objective rankings may differ from rankings based on the integrated process objective, revealing limitations of sequential decomposition when discrete and continuous decisions are strongly coupled. The study characterizes current solver behavior, practical QUBO/Ising reformulation requirements, and implementation considerations for applying emerging Ising‐based optimization tools in process systems engineering.