Partial constraint satisfaction approaches for optimal operation of a hydropower system

Abstract

Optimal operation models for a hydropower system using partial constraint satisfaction (PCS) approaches are proposed and developed in this study. The models use mixed integer nonlinear programming (MINLP) formulations with binary variables. The models also integrate a turbine unit commitment formulation along with water quality constraints used for evaluation of reservoir downstream water quality impairment. New PCS-based models for hydropower optimization formulations are developed using binary and continuous evaluator functions to maximize the constraint satisfaction. The models are applied to a real-life hydropower reservoir system in Brazil. Genetic Algorithms (GAs) are used to solve the optimization formulations. Decision maker's preferences towards power production targets and water quality improvements are incorporated using partial satisfaction constraints to obtain compromise operating rules for a multi-objective reservoir operation problem dominated by conflicting goals of energy production, water quality and consumptive water uses.