Tracking Controller Design for Nonlinear Dynamic Systems via Fuzzy Association Rule Reduction

One of the conventional significant criteria to be considered in real-time control applications is the computational complication of the controllers. The major weakness of fuzzy data mining is that after applying fuzzy data mining on the quantitative data, the number of extracted fuzzy association rules is very enormous. When many connection rules are obtained, the value of them will be reduced. This paper presents structures and a systematical progress method for two degree of freedom fuzzy controllers with nonhomogenous dynamics with respect to the two input channels. In this paper, singular value decomposition (SVD)-based complexity reduction technique and a reduced modified fuzzy logic controller (MFLC) are proposed. We introduce an approach to reduce and summarize the extracted fuzzy association rules after fuzzy data mining. Matlab/Simulink software is used to simulate the mathematical model of the Gun Turret-Barrel dynamic system. Different controllers are applied to the system. A comparison is introduced among the system performances under the control of each of FLC, a reduced FLC (in the number of membership functions and rules by using SVD method), and a reduced MFLC (reduction in the number of rules and modification in the shape of the membership functions by observing the system performance). The analysis points out that the proposed MFLC can ensure better control system performance with respect to the reference input in comparison with other controllers.