In order to handle that one production task simultaneously requires multiple processors to process it,multiprocessor task scheduling in a multi-stage hybrid flowshop with transportation times is proposed.This problem has been shown to be NP-hard. For this, an integer programming model is constructed with the optimization objective of minimizing the maximum completion time (makespan).A mixed discrete particle swarm algorithm is developed combined with an improved particle swarm algorithm, a modified genetic algorithm and a simulated annealing algorithm. Firstly, the relevant behaviors of particle swarm algorithm are improved to avoid the premature convergence of this algorithm. Next,crossover and mutation operators of genetic algorithm are introduced to further enhance the excellent individuals in the particle swarm algorithm and genetic algorithm. Finally, a simulated annealing algorithm isapplied to perform local search for the obtained particle swarm so as to get solutions with higher quality. The comparison and analyses between the proposed algorithm and some existing algorithms show that the developed hybrid particle swarm algorithm has better performance.