Improved Sparrow Search Algorithm Fused with Multiple Strategies

The original sparrow search algorithm suffers from drawbacks such as slow convergence speed, reduced population diversity, inadequate optimization accuracy, and easily falling into local extremum. To overcome these shortcomings, this study presents an enhanced sparrow search algorithm (GCSSA) that fused hybrid method. GCSSA enhances population richness and diversity by introducing the elite reversal strategy for initializing the population. The incorporation of the golden sine strategy in the update mechanism of discovery locations balances local optimization and global search capabilities, thereby improving convergence speed. Furthermore, the introduction of cauchy mutation in the update mechanism of joiner positions interferes with the present solution, improving the algorithm on attaining global optima and preventing premature convergence. Simulation experiments are conducted on 10 benchmark functions, compared with PSO, GWO, MFO, original SSA and the other two improved SSA. Experiment results demonstrate that the GCSSA exhibits superior global search capability and robustness in escaping local optima.