---

A constrained version of the Big Bang-Big Crunch algorithm for the efficient solution of the optimal reservoir operation problems is proposed in this paper. Big Bang-Big Crunch (BB-BC) algorithm is a new meta-heuristic population-based algorithm that relies on one of the theories of the evolution of universe namely, the Big Bang and Big Crunch theory. An improved formulation of the algorithm named Constrained Big Bang-Big Crunch (CBB-BC) is proposed here and used to solve the problems of reservoir operation. In the CBB-BC algorithm, all the problems constraints are explicitly satisfied during the solution construction leading to an algorithm exploring only the feasible region of the original search space. The proposed algorithm is used to optimally solve the water supply and hydro-power operation of “Dez” reservoir in Iran over three different operation periods and the results are presented and compared with those obtained by the basic algorithm referred to here as Unconstrained Big Bang–Big Crunch (UBB–BC) algorithm and other optimization algorithms including Genetic Algorithm (GA), Ant Colony Optimization (ACO) and Particle Swarm Optimization (PSO) and those obtained by Non-Linear Programming (NLP) technique. The results demonstrate the efficiency and robustness of the proposed method to solve reservoir operation problems compared to alternative algorithms.

---

This paper utilizes the Colliding Bodies of Optimization (CBO), Enhanced Colliding Bodies of Optimization (ECBO) and Vibrating Particles System (VPS) algorithms to optimize the reservoir system operation. CBO is based on physics equations governing the one-dimensional collisions between bodies, with each agent solution being considered as an object or body with mass and ECBO utilizes memory to save some historically best solutions and uses a random procedure to escape from local optima. VPS is based on simulating free vibration of single degree of freedom systems with viscous damping. To evaluate the performance of these three recent population-based meta-heuristic algorithms, they are applied to one of the most complex and challenging issues related to water resource management, called reservoir operation optimization problems. Hypothetical 4 and 10-reservoir systems are studied to demonstrate the effectiveness and robustness of the algorithms. The aim is on discovering the optimum mix of releases, which will lead to maximum benefit generation throughout the system. Comparative results show the successful performance of the VPS algorithm in comparison to the CBO and its enhanced version.

---

Parallel Cellular Automata (PCA) previously has been employed for optimizing bi-objective reservoir operation, where one release is used to meet both objectives. However, if a single release can only be used for one objective, meaning two separate sets of releases are needed, the method is not applicable anymore. In this paper, Multi-Step Parallel Cellular Automata (MSPCA) has been developed for bi-objective optimization of single-reservoir systems’ operation. To this end, a novel cellular automata formulation is proposed for such problems so that PCA’s incapability when dealing with them will be overcome. In order to determine all releases throughout the operation period, in each iteration – unlike PCA – two updates take place so as to calculate releases individually. Since a bi-objective problem in Dez reservoir (in southern Iran) has been solved by PCA in earlier works, the same data is used here. The results are given for a 60-months operation period, and to evaluate this method, the results of Non-Dominated Sorting Genetic Algorithm (NSGAII) is also given for the same problem. The comparison shows MSPCA, beside remarkable reduction in computational costs, gives up solutions with higher quality as well.