Intelligent Control Systems and Fuzzy Controllers. II. Trained Fuzzy Controllers, Fuzzy PID Controllers

摘要

The problems of control systems intellectualization are observed. The necessity of intellectualization of a wide range of systems and control methods is proved. The hierarchy of levels of intellectual control observed and comparison analysis of different artificial intelligence devices given. Importance of target setting's automation problems' solving in control systems is pointed out, as well as intellectualization of anthropocentric systems, including the ones based on fuzzy logic and case-based reasoning. The logical-linguistic, analytical, learned and PID fuzzy controllers are considered, based on fuzzy logics of Zadeh. An overview of the Mamdani-type controllers, controllers based on TS-model and the ANFIS architecture, using neural network structure is provided. The conditions of optimality and stability of control systems with Mamdani fuzzy controllers are analyzed. The Sugeno dynamic models and the ANFIS adaptive models and the methods of learning developed on the basis of fuzzy controllers are considered. The structure of a Mamdani fuzzy controller and its implementation by means of the Simulink is described. An example of application of Simulink to determine the optimal parameters of a fuzzy controller is shown. The examples of the fuzzy controllers use are given.