Aiming at the problem of insufficient convergence ability of particle swarm optimization and easy to fall into local optimum, a novel particle swarm optimization based on scout learning strategy is proposed. The algorithm first uses the topology structure to construct the particle population, and then uses the local search ability and global search ability of the joint factor equalization algorithm, and improves the speed and position formula of the algorithm through the scout learning strategy to generate candidate solutions. The results of Wilcoxon rank sum test and CEC2017 benchmark function test show that this novel particle swarm algorithm has better convergence ability, optimal solution precision and algorithm stability, indicating that the performance of the algorithm has been improved.