IJCAS

In order to achieve accurate control of uncertain mechanical systems, nonlinearity and uncertainty are two major challenges. To address these challenges, this paper proposes a “dynamic” performance metric, denoted as ϑˆ, which differs from the previous performance metric ϑ, to describe the constraint tracking error in a general way. Based on the metric ϑˆ, a new robust control method is proposed to improve the control performance. The control design process consists of two stages. In the first stage, a robust control scheme is designed to provide guaranteed performance in the presence of uncertainty. This control scheme includes a tunable parameter that can be selected within a certain range. In the second stage, an intelligent multi-agent method is employed to determine the optimal control parameter. This method considers multiple performance indexes, each of which involves different design considerations that may be opposed to each other. Adjusting parameters may improve one performance index while reducing another. Therefore, the intelligent multi-agent method is used to optimize performance indexes and obtain the optimal control parameter. The proposed design method is implemented on a mobile robot, and its performance is verified by comparing with other methods.

Keywords: Constraint tracking error, multi-agent method, robust control, uncertain mechanical systems.