Abstract: Aiming at the existent question that serial PSO algorithms has big calculation and speed slow
we propose a parallel PSO algorithms based on DSP parallel system. The algorithm adopts the information transfer model and single population Master-Slave model in 4 DSP parallel system.The initialize setting is realized by DSP-A
the adapt value of every particle are parallel computed by DSP-B
DSP-C and DSP-D.Finally
by DSP-A to compare to the adapt value of every particle and it’s individual extremum
and to choice better individual extremum and compare to choice best solution of entire population
and the update every particle position and speed.The simulation experiments show that the algorithms is compare to serial PSO algorithms that get 2.525 acceleration and 63.13% deficiency Keywords:Target tracking; Parallel PSO algorithms; DSP; Parallel system