Uncertainty management via online scheduling for optimal short-term operation of cascaded hydropower systems

摘要

The efficient utilization of available resources plays a significant role in the operation of cascaded hydropower systems. Scheduling of such systems may be used to determine power generation and generating unit decisions that optimize economic performance subject to prevailing constraints. The highly volatile and unpredictable nature of market and weather conditions give rise to uncertainties in electricity prices and water inflows, which may significantly impact the power generation schedules for these systems. Computed schedules may also give rise to significant variation in the generating unit commitments in response to slight fluctuations in electricity pricing, a phenomenon referred to as "nervousness". This article focuses on the development and implementation of a novel rolling horizon online scheduling scheme capable of handling uncertainties in key system and model parameters. Different strategies for mitigating the undesirable effects of nervousness are also investigated. These approaches are demonstrated through application to several case studies.