Performance evaluation of atom search optimization algorithm performance in optimal operation of multi-reservoir systems and single reservoirs under sedimentation (Case study: Dez Dam)

Due to the scarcity of water resources and increasing human demand, the optimal operation of reservoirs has become one of the most important issues in the world. In this regard, the correct and optimal operation of dams is one of the most efficient tools for water resources management. Due to the large number of decision variables, relationships, and constraints in the problems, reservoir management, and optimization have many complexities. Therefore, many researchers in their research have paid special attention to this issue. In this research, a new algorithm namely Atom Search Optimization (ASO), which is derived from the concepts of molecular dynamics, will be developed for multi-reservoir water resource systems. On the other hand, failure to implement appropriate policies to protect the soil has led to an important phenomenon of erosion in the lands above the reservoirs, one of the negative consequences of which will be sedimentation. The transfer and accumulation of suspended sediments, in turn, will reduce the useful volume of the reservoir, which is neglected in most reservoir optimization issues. However, in this study, the optimal operation of a single reservoir dam is investigated by the Atom Search Optimization algorithm in terms of monthly sediment yield.

First, the performance of the atom search algorithm on mathematical benchmark functions will be investigated. Then, by performing sensitivity analysis to logically determine the effective coefficients of the algorithm and selecting the appropriate number of particles and the number of iterations of each operation, the performance of the algorithm on conventional systems of four and ten reservoirs is analyzed. In order to supply water downstream of Dez Dam, considering the important issue of monthly sediment flow in the reservoir, the atom search algorithm and four other common algorithms are used. The results are modified by selecting the objective function value criteria, RMSE, MAE, NSE, and PBIAS values and prioritized using TOPSIS and Modified-TOPSIS ranking techniques.

The performance of the atom search algorithm on conventional systems of four and ten reservoirs is analyzed, which shows the results of 95.33% with an absolute optimal solution of four reservoirs, ie 308.29, and 89.67% with an absolute answer of ten reservoirs, 1194.44. Also, by comparing the atom search algorithm and four common Salp Swarm Algorithm (SSA), Sine Cosine Algorithm (SCA), Particle Swarm Optimization (PSO), and Genetic Algorithm (GA) in a single reservoir system under deposition, the absolute superiority of the ASO search algorithm was demonstrated.

The use of an atom search algorithm in solving optimization problems in the field of water resources management is recommended, especially in terms of the effectiveness of soil protection and sedimentation of reservoirs.