The proposal for this study is to obtain an adaptive controller by feedback states for fourth order time-varyng multivariable systems or with parametric uncertainty. Alternatively the difficulties and limitations involving the usual methods of synthesis of optimal controllers for this class of systems, through the investigation of an algorithm based on particles swarm optimization with good tuning of parameters, we seek to define a flexible structure for applications that require decision making and online estimation of the feedback matrix parameters. The controller developed should provide, for every execution, not only system stability, but at the same time, provides compliance to the performance indexes set, without the need to perform complex and expensive mathematical calculations, eliminating thus the subjective components of analysis and design