The Swarm Algorithm Framework (SWAF) contains a compact Multiagent Optimization System (MAOS) and a Problem Formulater. The MAOS works on the well-defined landscape of the optimization task, and the Problem Formulater manipulates the landscape based on certain knowledge, including the feedback information from the MAOS, for facilitating the solving process of the MAOS.

The work on the Problem Formulater: an adaptive relaxing rule for handling equality constraints of numerical optimization problem (NOP) has been integrated.