Hybrid Fuzzy Logic Control to Stabilize an Inverted Pendulum from ArbitraryInitial Conditions

摘要

The purpose of this work was to develop a controller that could perform thehighly discontinuous and nonlinear task of balancing the inverted pendulum from an arbitrary set of initial conditions. Fuzzy logic was chosen as the control technique because of its ability to deal with nonlinear systems, as well as its intuitive nature. The rule base depends on intuition and logic, rather than an exact mathematical model. This makes it more robust to changes in the model, and also gets rid of the need to solve nonlinear differential equations or optimality conditions. Using a set of fuzzy logic controllers, linked in the right way, the primary research objectives were accomplished. In addition, the adjustable desired energy allows for different tip masses and arm lengths to be controlled equally as well with a simple number adjustment on the interactive animation window. This makes the controller very robust. There are many options for future research on the inverted pendulum using fuzzy logic. The current controller could be improved by further adjustment of the membership functions. This adjustment could be performed manually by experimenting and checking the results. An optimization routine using Genetic Algorithms could also be used. This would involve letting the computer run simulations and compare results automatically. (KAR) p. 47.