- Parallel optimization offers faster and more efficient problem-solving by reducing computational resource usage and execution time. By integrating multiple techniques such as evolutionary algorithms and swarm-based optimization, it enables more effective exploration of the search space and facilitates the attainment of optimal solutions within shorter time frames. Differential Evolution (DE) is a powerful and relatively recent evolutionary algorithm. However, its performance is often limited because most applications rely on a single mutation operator with fixed parameter values. To address this limitation, this study proposes a parallel framework that executes three DE algorithms simultaneously, with dynamic selection among three mutation strategies. Each algorithm runs independently on separate computational units, and the best solution identified is shared across units to accelerate convergence and improve efficiency. In the proposed approach, a mutation pool consisting of three mutation operators and a parameter pool with three predefined values are established. During evolution, mutation operators and parameter values are randomly selected from these pools to generate trial vectors, allowing the algorithm to exploit the complementary strengths of different strategies. The effectiveness of the proposed algorithm was evaluated on 24 widely used benchmark functions. Experimental results demonstrate significant improvements in both convergence speed and solution quality compared with traditional DE and non-DE algorithms. These findings indicate that the proposed parallel DE framework is highly competitive and provides a promising direction for solving complex optimization problems.

Keywords: optimization; global optimization; differential evolution; evolutionary algorithm; mutation operator; Parameter   

                    pool; parallel techniques; termination rules.