/PGA

Code for the paper "A Penalty-Based Guardrail Algorithm for Non-Decreasing Optimization with Inequality Constraints"

Primary LanguagePython

Code for the paper "A Penalty-Based Guardrail Algorithm for Non-Decreasing Optimization with Inequality Constraints". The PGA has been implemented on three application domains: an N-dimensional optimization problem, a simplified model of a district heating system, and a model derived from learned deep neural networks.