In this thesis, we present an expanded account of "particle swarm optimization algorithm for flow shop group scheduling problem with sequence dependent setup time" based on some articles by Schaller et.al (2000), Lian et.al (2006), Liu et.al (2008) and Kuo et.al (2009). It is well known that the group scheduling in sequence dependent flow shop problem is a branch of production scheduling. In this case, a major setup on a machine is needed for switching between the groups and a negligible or minor setup is needed on the machines for switching between the jobs of the same group. it is proved that this problem in the state of sequence dependent setup time is NP-hard and current algorithms even for moderate size problem cannot be solved to guarantee optimality. Thus, the optimal solution must be estimated with the heuristic algorithms. Particle Swarm Optimization PSO algorithm iired by social behavior of bird flocking or fish schooling. This algorithm has been developing rapidly and applied widely in the optimization problems since it was introduced, as it is easily understood and realized. PSO algorithm, are applied for scheduling problem such as permutation flow shop scheduling problem and are quite effective in solving this problem. The search space in PSO algorithm is continuous, thus for use this algorithm for the problem with the discreet search space e.g. scheduling problems, must be changed. An encoding scheme based on ranked order value (ROV) is developed which converts the continuous position value of particles in PSO to job permutation and group permutation. In this thesis, for improvement the solutions and do not trap in the local optima, the dynamic PSO hybrids with the neighborhood search algorithm, called Individual Enhancement, and applied for group scheduling in flow shop problem with sequence dependent setup time to minimize total completion time (i.e., makespan) and minimize total flow time. The performance of the algorithm is compared with the available algorithms in literature based on available test problems. The results show that the proposed algorithm has a superior performance compared to the available ones in literature.