Scheduling short-term hydrothermal generation using evolutionary programming techniques

The authors present a novel evolutionary programming (EP) based algorithm for the short-term hydrothermal scheduling problem. To more realistically represent the relationship between the generation and amount of water discharge for hydroaggregates, the generation models of the hydro plants as well as thermal plants are often expressed as nonlinear and nonsmooth curves with prohibited areas. The advantage of the proposed algorithm is that it is capable of determining the global or near global optimal solutions to such an optimisation problem with multiple local minima. The developed algorithm is illustrated and tested on two model systems. The test results are compared with those obtained using gradient search, simulated annealing and genetic algorithm methods. Numerical results show that the proposed EP-based algorithm can provide accurate solutions within a reasonable time.