Exploring the efficiency of harmony search algorithm for hydropower operation of multi-reservoir systems: A hybrid cellular automat-harmony search approach

Multi-reservoir systems are one the important infrastructures due to their role in the energy supply for human beings. Construction of these systems are very costly making their operation a delicate task due to their significant economic impacts. Optimal operation of hydropower reservoir systems is a complex task due to the non-convexity and nonlinearity of the problem involved. Conventional methods often fail to tackle the complexity of the problem while modern heuristic algorithms lack efficiency when solving this problem. This is further amplified when population based heuristic methods are to be used for large scale multi-reservoir real-time operation problems, where efficiency of the solution method is vital. This paper explores the hybridization of a newly proposed method namely Cellular Automata with the wellknown Harmony Search algorithm for efficient solution of multi-reservoir hydropower operation problems. The HS method is embedded into a CA framework in which the CA is used to breakdown the large scale reservoir system operation into a series of small scale sub-problem with a size equal to the number of reservoirs in the system. HS method is then used to solve each sub-problem and the results are passed to the CA method. The proposed method is used to solve a nonlinear version of the well-known four reservoir problem and the results are presented and compared with the existing results.