الفهرس 
Chapter 1 IntroductionOnly 14 pages are availabe for public view
 |  | from | 139 |  |  |
| | | from | 139 | | |
Abstract
Grasshopper Optimization Algorithm (GOA), which is one of the recent metaheuristic optimization algorithms, mimics the natural movements of grasshoppers in swarms seeking food sources. Some deficiencies exist in the original version of GOA such as slow convergence speed, and the original GOA may get quickly stuck into optima. For tackling the drawbacks of the original GOA, enhanced versions of GOA have been proposed to deal with the optimization problems more effectively. In the current thesis, two strategies have been integrated into GOA: the grouping mechanism of non-linear ‘c’ parameters and the mutation mechanism. Moreover, two different groups of non-linear ‘c’ parameters have been suggested in the grouping mechanism. Integrating the grouping mechanism into GOA can update the grasshoppers’ positions within a limited local area, whereas the diversity of agents can be improved by considering the mutation mechanism. Eight Novel-Variants GOA (NVGOAs) are proposed to address the deficiencies of the original GOA. where two variants NVGOA1_1 and NVGOA2_1 represent the impact of each proposed group of ‘c’ parameters. Another two variants NVGOA3 and NVGOA4 represent the impact of the mutation mechanism with two different values of probability. Moreover, four variants: NVGOA1_2, NVGOA1_3, NVGOA2_2, and NVGOA2_3 represent the combination of the two proposed mechanisms. First, the comparison between each proposed variant and the original GOA has been conducted. Then, the performance of the best-recorded NVGOA variants has been tested against CEC-2017 benchmark functions and compared with six state-of-the-art optimization algorithms. Moreover, the Wilcoxon Signed-Rank test has been employed to exhibit the efficiency of the proposed variants. As well, comparative analysis with previous enhancements of GOA has been conducted against the best recorded NVGOA variants. The results of all these analyses demonstrated the success and efficiency of the proposed NVGOA variants to solve numerical optimization problems. Concerning demand side management in smart grids, the proposed NVGOA variants have been applied to schedule the loads in three areas: residential, commercial, and industrial to decrease the daily operating costs and peak demand. The overall results show that the proposed NVGOA algorithms are effective solutions to address the flaws of the original GOA and can get high-quality solutions for different optimization problems.