| Fuzzy control bases on a rule control and the basic idea implements human control experience by computer,wihich is Fuzzy contol rules that are component of linguistic viables. Fuzzy control as an important branch of intelligent control, has been widely used in many fields. However, fuzzy control over many of the problems faced are restricted its further development. For instance, "rule explosion" in fuzzy control; the control precision of fuzzy control and how the performance is superior to the traditional controller and son on. In light of these problems, this paper attempts to start membership function that a useful attempt to slove these problems.Membership function in general fuzzy control is determined on the basis of experience, which often will appear roughness in fuzzy control. It merely reflects the fact membership function is a change trend in membership in the state variables and strictly speaking it should be obscure. because it is not known how the membership function more suited to their state variables.Therefore, in the whole control process, we should hope that the membership function itself is not a static process, and the hope that it is a dynamic process, which changes the trend is improve the performance of the fuzzy control.Tone operators has a direct impact on the size of the membership function in fuzzy mathematic. Therefore, this paper will attempt to make use of tone operators in fuzzy control, and use tone operators to change the membership function of state variables. The membership rate and sliding mode control will be applied to tone operators in order to adjust the control variable when the state variables are changed. In paper based on above the idea it is made some exploratory research in fuzzy control algorithm. Its main contents are :1. Defines fuzzy sets of tone operators. If fuzzy sets of tone operators compose of fuzzy sets in infinit membership function by tone operators, the classical sets and fuzzy sets can be seen as the two special circumstances in fuzzy sets of tone operators.2. Put forward the concept of variable membership function. If the general membership function in the control process itself sees as a static process. Tone operators will then on the membership function can be seen as a dynamic process. And a similarity of the dynamic membership functions can be called variable membership function.3. Designed algorithm of fuzzy control with sliding model variable membership function on the three stage inverted pendulum. The memberships of state variables and the control are the map relation and tone operators are changed by the idea of sliding mode control in order to adjust the control variable. Last this algorithm is used on the three inverted pendulum simulation and it is or not effect that the algorithm is certificated.4. Designed algorithm of sliding model fuzzy control with variable membership function on the quadruple inverted pendulum. According to the characteristic of the interpolation in fuzzy control, it is changed that the corresponding input membership functions by tone operators, and the partitions of the universe of state variables are changed by the idea of sliding mode control in order to adjust the control variables.Last the algorithm is or not effect by the simulink of the quadruple inverted pendulum. |
PDF Full Text Request