Identification and Fuzzy Controller Design for Nonlinear Uncertain Systems with Input Time-Delay

This paper considers the identification and fuzzy controller design for nonlinear uncertain systems in presence of unknown input time-delay. Firstly, a time-delay Takagi-Sugeno-Kang (TSK) type fuzzy neural system (TDFN) is proposed to identify a class of nonlinear input time-delay systems. The input-output signals of nonlinear systems are used to identify the system dynamics and unknown time-delay, and then construct the system in the form TSK-fuzzy time-delay model. Each fuzzy rule has a corresponding linear system with an input time-delay as the consequent part. Based on parallel distribution compensation (PDC) approach, the Smith predictor compensation and dominate pole assignment technique are then adopted to design the fuzzy PID controller. Several simulations are shown to demonstrate the effectiveness and control performance of the proposed approach. mode controller [4]. Neural network based controller were introduced to treat a class of nonlinear time-delay systems [6, 7] Recently, intelligent systems have attracted great attention from both academic and industrial communities. They have been successfully used in a wide variety of applications [3, 9-10, 12-18]. In particular, the neuro-fuzzy systems combining the advantages of fuzzy logic systems and neural networks have become a very active subject in many scientific and engineering areas, such as, model reference control problems, PID controller tuning, signal processing, etc. [10, 12-18]. To date, the fuzzy neural system has been used an alternative approach to conventional controller techniques for complex control problems. These systems possess the properties of a parallel computation scheme, easy to implement, fuzzy logic inference system, and parameter convergence [12-17]. The fuzzy rules can be designed and trained from linguistic information and numeric data. That is, it is then easy to design a fuzzy neural system to achieve a satisfactory level of accuracy by manipulating the network structure and learning algorithm of the fuzzy neural system.