Control of rotary inverted pendulum system parameters with moving sliding mode control method

The rotary inverted pendulum system is one of the most preferred systems in control applications. In order to demonstrate the success of a new control method, the system is tried to be stabilized for the upper equilibrium point where the system is unstable. In this study, the system is controlled by three different control methods. Firstly, the system is controlled by a sliding mode control method with a classical moving sliding surface. In the first method, the unknown coefficients are calculated by genetic algorithm through the equations of motion of the system with a program in Matlab. As the second control method, fuzzy logic based moving sliding mode control method was used. In this method, the slope of the sliding surface is calculated by fuzzy logic method. The boundary values of the membership functions in the structure of fuzzy logic were found by genetic algorithm. Finally, adaptive neural fuzzy inference based moving sliding mode control structure is preferred in the control of the system. Here, the adaptive neural fuzzy inference method is used to calculate the slope of the moving sliding surface. Genetic algorithm was used to calculate the coefficients in the adaptive neural fuzzy inference method. When the results are analysed, the fuzzy logic based moving sliding mode control method is the method that provides the control in the shortest time with approximately 1 second. If the torque magnitude is desired to be low, in this case, the classical moving sliding mode control method has been the control method that provides the lowest torque value with 12 Nm. In cases where the maximum amperage drawn by the system is important, the classical moving sliding mode control method gave the best result with 2.5 amperes.