Operational planning of hydrothermal systems based on a fuzzy-PSO approach

Reservoir Operation Rules (ROR) are functions that determine the operating volume of each reservoir in order to establish a coupled behavior between hydroelectric power plants. This paper proposes the implementation of ROR through a hybrid approach based on fuzzy systems and Particle Swarm Optimization (PSO). Fuzzy inference systems of the Mamdani type are used to estimate the operating volume of each hydroelectric plant, which is based on the value of the stored energy in the hydroelectric system. In order to represent the particular behavior of each reservoir for the operation of the system in an optimized configuration, a fuzzy system for each hydroelectric plant was designed. The PSO algorithm is used to adjust the membership functions that represent the consequent of the linguistic rules of the fuzzy system. A computational model for simulating the operation of hydroelectric systems is also used to implement the proposed ROR, based on developed Fuzzy-PSO (ROR-FPSO) systems, and to compare them to the Parallel Operation Rules (ROR-PO), to the operation rules based on Mathematical Functions (ROR-MF) and to the operational rules based on Takagi-Sugeno Fuzzy Systems (ROR-TSFS). The results illustrate the effectiveness of the ROR-FPSO, which maximizes the hydroelectric benefits associated to hydrothermal generation system, when compared to other ROR already found in literature.