Research On RS-MEA-Based Fuzzy Control Strategy In The Inverted Pendulum

The inverted pendulum is a kind of typically nonlinear, multi-variable, unstable, strong-coupling and fast system. Control circle have paid more attention to its research for many years because it is influenced by nonlinear factors and difficult to control. The research of the inverted pendulum can come down to the research of the multi-variable and absolutely unstable system, so its control methods and ideas have comprehensive purposes to the general industry process, and the research of control methods also possess introductive significances.At first, this paper comments control problems and common control strategies of the inverted pendulum. Secondly, basic ideas and piths of Rough Sets (RS, in short) are deeply analyzed, and its current application in intelligent control system is summarized. The theoretic pith, main arithmetic, its application and development of Mind Evolutionary Algorithm (MEA, in short) are mainly introduced.Finally, aiming at dynamic characteristics of the inverted pendulum, this paper searches the cut-in point of RS, MEA and fuzzy control, and proposes a new kind of intelligent control strategy-RS-MEA-based fuzzy control strategy to control the inverted pendulum.This strategy adopts the merit of T-S fuzzy model which can express dynamic characteristic of nonlinear system, and sets up the fuzzy inquiry table. Moreover, this paper brings forward a new fast inquiry method-storage in row and subscript inquiry method. This method cangreatly improve the speed of rule inquiry and spend the same inquiry time on every rule. Knowledge reduction and inconsistent degree are used to reduce fuzzy rule table, which avoids some problems such as vast search space and rule blast because of too much input variables. Simultaneously, MEA, which is superior to Genetic Algorithm in compute efficiency and astringency, is used to optimize parameters of fuzzy control.Control experiments of the inverted pendulum show that, comparing with PID and fuzzy control strategies, this RMBFC strategy possesses simpleness, validity. It not only reduces the number of fuzzy rules, quicken optimizing speed of MEA and improve system fastness, but also avoids the disadvantage of general fuzzy controller that input variables are reduced at random from subjective view because disposal of analyzed data is objective. At the same time, yawp caused by strong coupling relation between variables is eliminated due to getting rid of abundant condition attributes, which makes rapid speed and robustness of system improved. Moreover, this control strategy can well control two-stage inverted pendulum.