Abstract: Based on adaptive adjustment of the inertia weight in the particle swarm optimization algorithm，this paper addressed the issue of the algorithm easily falling into local optima and being difficult to escape， inspired by the principle of sardine being stimulated to accelerate swimming to avoid death， a collision strategy is used to simulate external stimuli and enhance its ability to the algorithm's ability to break free from local optima. In order to balance the ability of global exploration and local fine search， a nonlinear adaptive adjustment braking operator is introduced to adjust the velocity of particles accordingly. The improved particle swarm optimization algorithm is applied to multi-dimensional function optimization ， and the experimental results show that the adaptive particle swarm optimization algorithm with strategy of collision and braking has better algorithm efficiency and global optimization ability compared to the standard particle swarm optimization algorithm. 

Key words: collision； braking； particle swarm optimization algorithm； local extremum； inertia weight