Comparison between closed-loop and partial open-loop feedback control policies in long term hydrothermal scheduling

Abstract

Stochastic dynamic programming has been extensively used in the optimization of long term hydrothermal scheduling problems due to its ability to cope with the nonlinear and stochastic characteristics of such problems, and the fact that it provides a closed-loop feedback control policy. Its computational requirements, however, tend to be heavy even for systems with a small number of hydroplants, requiring some sort of modeling manipulation in order to be able to handle real systems. An alternative to closed-loop optimization is an approach that combines a deterministic optimization model with an inflow forecasting model in a partial open-loop feedback control framework. At each stage in this control policy, a forecast of the inflows during the period of planning is made, and an operational decision for the following stage is obtained by a deterministic optimization model. The present paper compares such closed-loop and partial open-loop feedback control policies in long term hydrothermal scheduling, using a single hydroplant system as a case study to focus the comparison on the feedback control performance. The comparison is made by simulation using data from historical and synthetical inflow sequences in the consideration of three different Brazilian hydroplants located in different river basins. Results have demonstrated that the performance of the partial open-loop feedback control policy is similar to that of the closed-loop control policy, and is even superior in dry streamflow periods.

An Application of Genetic Fuzzy Systems to the Operation Planning of Hydrothermal Systems

“…Therefore, the operating cost is represented by a single cost function of the non-hydraulic sources (Carvalho & Soares, 1987). The future cost is a terminal condition, used in the optimization models of the operation, to balance the water use during the planning horizon and its future use (Martinez & Soares, 2002). Equality (2) represents the restriction of meeting the demand of electricity in the time range t. The total generation from the hydroelectric system is represented by equation (3), given by the sum of the functions of hydraulic production of each hydroelectric power plant.…”

Detailed long‐term hydro‐thermal scheduling for expansion planning in the Nordic power system

“…This process can be resembled in rolling horizon simulator scheme, as demonstrated by the authors in [23,24], combining short-term decisions and medium-term strategy computation in sequence. This type of scheme, which is often linked to the framework of model predictive control, was proposed for the LTHTS problem in [25][26][27]. A more generic controltheoretic framework for power dispatch and storage dynamics was presented in [28].…”

“…Some preliminary results obtained using an early version of the proposed method on synthetic test data were reported in [30]. Moreover, we add to the work in [25,26] by introducing uncertainties in the scheduling, and to [27] by applying the model to a realistic system.…”

Análise do erro de previsão de vazões mensais com diferentes horizontes de previsão

“…Isto torna o planejamento da operação um problema de otimização de grande porte, não linear e, consequentemente, estocástico. Outras abordagens consideram de forma implí-cita o modelo estocástico para as vazões como é feito em (Martinez and Soares, 2002). Nestas abordagens, os modelos para as vazões são utilizados para obter previsões.…”