Summary: The particle swarm optimization (PSO) algorithm has certain disadvantages; for instance, the convergence viscosity of the algorithm is reduced in the post evolution phase, the optimization search efficiency is also reduced, the algorithm is easy to be inserted with a local extremum during the calculation of a complex problem of a high-dimensional multiple extremum, and the convergence thereof is low. To compensate for the PSO disadvantage, we propose a particle swarm optimization of the comprehensive improvement strategy, which is a simple particle swarm optimization with a dynamic adaptive hybridization of the extremum disturbance and the ecds-PSO algorithm. This new comprehensively improved particle swarm algorithm discards the particle velocity and reduces the PSO from the second-order to a first-order difference equation. The evolutionary process is controlled by the particle position variables only. The hybridization operation of increasing the extremum disturbance and introducing a genetic algorithm can accelerate the particles to overstep the local extremum. The mathematical derivation and the plurality of a comparative experiment provide the following information: the improved particle swarm optimization is a simple and effective optimization algorithm that can enhance the algorithm accuracy, the convergence viscosity and the ability of avoiding the local extremum, and effectively reduce the calculation complexity.