Adaptive Cuckoo Search Algorithm for Short-Term Fixed-Head Hydrothermal Scheduling Problem with Reservoir Volume Constraints

This paper present three versions of Cuckoo Search Algorithm (CSA) including conventional Cuckoo Search Algorithm (CSA), modified CSA (MCSA) and adaptive CSA (ACSA) for solving the fixed head short-term hydrothermal scheduling (ST-HTS) problem where the reservoir volume constraints and nonconvex fuel cost function of thermal unit as well as the power losses in transmission line are taken into account. Among the applied methods, ACSA is first developed in the study by performing two modifications on second new solution generation via the action of an alien egg to be abandoned. In the ACSA, all initial solutions or all solutions at the end of the previous iteration are evaluated and sorted into two kinds of solution, good solutions with lower fitness function and bad solutions with higher fitness function. The implementation of the first new solution generation first via Lévy flights in the ACSA is carried out similarly to that in MCSA. However, at the second new solution generation the ACSA evaluates the current solutions to choose the best one and use the information of the best one with a random solution to generate the second new solutions via the action of an alien egg to be abandoned. In addition, the probability of an alien egg discovery is considered an adaptive variable, which is set to the largest value at the beginning and decreased as the iteration is increased. Due to the adaptive value of the parameter, the ACSA can search an optimal solution but the trial runs are significantly decreased compared to CSA and MCSA. The performance of the ACSA is validated by testing on two systems and comparing with CSA, MCSA and other existing methods available in the paper.