Switching Control of DC Motor Using Multiple Fuzzy Cognitive Network Models

Abstract

A new DC motor multiple models switching control architecture is proposed in this paper, which is based on the framework of Fuzzy Cognitive Network (FCN) system modeling. A FCN is an operational extension of a Fuzzy Cognitive Map which incorporates proven stability and guaranteed exponentially-fast error convergence to zero during its training and supports at the same time the continuous interaction with the system it describes. In the proposed approach the network assumes functional interconnection weights instead of plain values and the acquired knowledge, during its training, is actually stored in multiple polynomial weight forms. Multiple models carry information associated with different areas of DC Motor operation leading to multiple local inverse FCN control actions, each one associated with the corresponding process model. The model that best approximates the plant in every time instant is determined through a switching rule based on a performance index. The incorporated multiple models may adapt and change their shape online enhancing the overall performance.