Chaotic Arc Adaptive Grasshopper Optimization

The grasshopper optimization algorithm (GOA) has become one of the most widely used methods, regardless of its shortcomings in its performance and algorithm design. To further harmonize exploration and exploitation in the GOA, this paper proposed an enhanced variant of the primary method by incorporating chaos theory to address the problems of slow convergence, low solution accuracy, and stagnation. Structural code analysis and previous works reveal that the grasshopper algorithm's internal function is one of these shortcomings' primary sources. The proposed GOA-based method employed an adaptive arc function instead of the grasshopper algorithm's internal function to strengthen and balance the global exploration capabilities and local exploitation capacities without any expensive modifications. Additionally, the chaotic mapping strategy was adopted to update the individual positions of agents iteratively. The individuals near the current optimal solution were disturbed to regenerate the optimal solution and enhance the swarm's diversity. This idea improves the global inspection capacity and discourages falling into local optima. To verify the proposed technique's effectiveness, a comprehensive set of twenty-seven benchmark cases and three engineering design problems were used for validation. We compared the proposed GOA-based method with the WOA, SCA, GOA, DA, CSSA, CGSCA, CLPSO, GL25, and OBWOA. Additionally, we tested the designed GOA against LSHADE, SHADE, EBOwithCMAR, SaDE, MPEDE, and EPSDE. Simulation results demonstrated that the algorithm was substantially superior to the original technique. Its global optimization competence, search accuracy, and convergence performance were notably improved. The results expose the regulation of this internal factor, which significantly affects the quality of the results. Further information about this research and assistance with any request is available at http://aliasgharheidari.com.