Comparison of Fuzzy Identification Schemes for Robust Control Performance of an Adaptive Fuzzy Controller

Fuzzy identification schemes play a vital role in the control performance of an adaptive fuzzy controller. Non-linear uncertain systems are difficult to model and control. A good fuzzy model of an uncertain non-linear system can be guaranteed if the adaptation mechanism is computationally efficient as well as robust in the face of noisy data coming from the sensors. The prediction accuracy of the fuzzy model depends on the quality of learning provided by the identification algorithm. In adaptive fuzzy control problems the control performance is heavily dependent on the parameters estimates produced by the identification scheme. In this paper the controller develops an inverse model of the plant online. The problem of inverse model identification becomes more challenging when external disturbances and plant delays are present in the control loop. In this research work two different computationally efficient fuzzy identification schemes are discussed. They are used for estimating the rule confidences of Fuzzy Relational Models and are based on the probabilistic learning approach. Both the classes of learning schemes are compared on the basis of robustness, rate of convergence and dependence on other controller parameters such as learning rate and forgetting factor. The control objective is a tracking problem when the plant under control is non-linear and controlled output is corrupted by sensor noise.