Design of type-2 fuzzy logic controllers using meta-heuristic optimization algorithms for rotary inverted pendulum / Mukhtar Fatihu Hamza

Mukhtar , Fatihu Hamza (2017) Design of type-2 fuzzy logic controllers using meta-heuristic optimization algorithms for rotary inverted pendulum / Mukhtar Fatihu Hamza. PhD thesis, University of Malaya.

Preview

PDF (Thesis PhD) Download (3305Kb) | Preview

Abstract

The type-2 fuzzy set (T2FS) is introduced to circumvent the limitations of the type-1 fuzzy set (T1FS). The main characteristic of type-2 fuzzy logic controller (T2FLC) is that, its Membership functions (MFs) are fuzzy. Therefore, it has more degree of freedom in designing varieties of systems with uncertainties. The control performance of type-1 fuzzy logic controller (T1FLC) can be improved by T2FLC because it has the advantage of footprint of uncertainty (FOU) that can be used to improve the corresponding MFs. Searching the suitable values of parameters and structure of type-2 fuzzy logic systems is a complex task. Many types of meta-heuristic algorithms have been proposed in the literature to solve this complex problem. This is presently attracting tremendous attention from researchers in this area of research. The Genetic algorithm (GA), Particle Swarm Optimization (PSO) and Cuckoo Search (CS) were considered in this research as three different paradigms that can be used in designing the optimized T2FLCs. In this research, the design of an optimized Interval Type-2 Fuzzy Proportional Derivative Integral Controller (IT2FPIDC) in cascade form for Rotary Inverted Pendulum (RIP) system is reported. The Type-1 Fuzzy Proportional Derivative Integral Controller (T1FPIDC) in cascade form was also designed using the same procedure for IT2FPIDC for fair comparisons. The parameters of the T1FPIDC and IT2FPIDC are optimized using GA, PSO and CS. This was also done to enable the comparisons between these types of meta-heuristic optimization algorithms. The goal is to stabilized the RIP at upright unstable equilibrium position and control it to follow a desired time varying trajectory. The performance indexes considered for the proposed controllers are steady state error (

Actions (For repository staff only : Login required)