Reservoir optimization models are often large-scale, having a complex, nonlinear, multi-dimensional structure, which poses a challenge for classical methods for their solution. This has encouraged the researchers to focus on Meta-heuristic which due to their flexible and adaptive nature have been successful in solving a plethora of real-life optimization problems. This study brings forward an implementation and comparison of six well-known Meta-heuristics namely: Simulated Annealing, Genetic Algorithms, Particle Swarm Optimization, Differential Evolution, Artificial Bee Colony, and Cuckoo Search and an integrated version of these algorithms with dynamic programming for optimizing the reservoir operations policy. In addition, two adaptive variants of DE named: FCADE2 and SaDE are also considered for the comparison. The case study considered for Mula reservoir supplying water to Major Irrigation Project on River Mula (Godavari basin), Ahmednagar district, Maharashtra, India. The objective is to determine the optimum release policy for Mula reservoir. Performance of the algorithms is analysed on two data sets (1) single year and (2) 30-years.

储层优化模型通常是大规模的，具有复杂的非线性多维结构，这对经典的求解方法提出了挑战。这鼓励了研究人员将注意力放在元启发式算法上，由于其灵活和自适应的特性，已经成功地解决了许多现实生活中的优化问题。这项研究提出并比较了六种著名的元启发式算法：模拟退火，遗传算法，粒子群优化，差分进化，人工蜂群和布谷鸟搜索，以及这些算法的集成版本和用于优化的动态程序设计水库运行政策。此外，还考虑了DE的两个自适应变体FCADE2和SaDE进行比较。考虑了为印度马哈拉施特拉邦艾哈迈德纳加尔地区的穆拉河（戈达瓦里盆地）的大型灌溉项目供水的穆拉水库的案例研究。目的是确定穆拉水库的最佳释放策略。在两个数据集（1个单年和（2）30年）上分析算法的性能。