A novel hybrid differential particle swarm optimization based on particle influence

In a global optimization problem, the particle swarm optimization algorithm finds the global optimal solution through the movement of particles. Each particle's position update is guided by the global optimum (Gbest) and the personal optimum (Pbest). However, this approach tends to ignore sub-optimal and promising individuals. In this paper, a novel hybrid differential particle swarm optimization (DPSO-PI) algorithm based on particle influence is proposed. DPSO-PI identifies the most influential potential individuals (non-global optimal and individual optimal) in the population based on the influence size of the particles. These potential particles undergo differential evolution, facilitating information exchange among sub-optimal or promising individuals in the population. DPSO-PI dynamically balances global and local search by adjusting the learning weights during evolution. Further, an adaptive jump-out strategy based on the hyperbolic tangent function is employed to prevent convergence to the local optima. Finally, some experiments were conducted between DPSO-PI and eight state-of-the-art algorithms on the CEC2017 benchmark functions. The experimental results demonstrate that DPSO-PI can effectively regulate population diversity and avoid falling into local optima.