Number Of Fuzzy Systems Research Articles

Adaptive Synchronization of Unknown Chaotic Systems Using Mamdani Fuzzy Approach

The synchronization problem for a general-class of unknown chaotic systems is addressed. Fuzzy systems in Mamdani type are employed to provide an approximate model of the master chaotic system using an adaptive approach. Within this, a number of fuzzy systems are utilized to approximate the unknown nonlinear functions of the master system. An approximate model similar to the master system is constructed which is the slave system. The error dynamics between the master and slave chaotic systems is used to build a suitable control input and fuzzy systems’ parameters adaptive laws to force the slave system to be synchronized with master system. The stability of the overall synchronization system is derived based on Lyapunov stability theory. Its shown that under appropriate assumptions, the proposed approach guarantees the boundness and asymptotic convergence of synchronization errors to a small neighborhood of origin. An extensive simulation study is performed on both Duffing and Rössler chaotic systems to show the effectiveness of the proposed scheme.

Fuzzy model-based condition monitoring and fault diagnosis of a diesel engine cooling system

This paper describes a fuzzy model-based diagnostic system and its application to the cooling system of a diesel engine. The aim is to develop generic cost-effective knowledge-based techniques for condition monitoring and fault diagnosis of engine systems. A number of fuzzy systems have been developed to model the cooling system components. Residuals are generated on line by comparison of measured data with model outputs. The residuals are then analysed on line and classified into a number of fuzzy classes symptomatic of potential system conditions. A fuzzy rule-based system is designed to infer a number of typical fault conditions from the estimated state of the valve and patterns in the residual classes. The ability to diagnose certain faults in the system depends on the state of the thermostatic valve. The diagnostic systems have been tested with data obtained by experimental simulation of a number of target fault conditions on a diesel generator set test bed. In five test cases for separate cooling system operating conditions, the diagnostic system's successful diagnosis rate ranged between 73 and 97.7 per cent of the test data.

Fuzzy model-based condition monitoring and fault diagnosis of a diesel engine cooling system

This paper describes a fuzzy model-based diagnostic system and its application to the cooling system of a diesel engine. The aim is to develop generic cost-effective knowledge-based techniques for condition monitoring and fault diagnosis of engine systems. A number of fuzzy systems have been developed to model the cooling system components. Residuals are generated on line by comparison of measured data with model outputs. The residuals are then analysed on line and classified into a number of fuzzy classes symptomatic of potential system conditions. A fuzzy rule-based system is designed to infer a number of typical fault conditions from the estimated state of the valve and patterns in the residual classes. The ability to diagnose certain faults in the system depends on the state of the thermostatic valve. The diagnostic systems have been tested with data obtained by experimental simulation of a number of target fault conditions on a diesel generator set test bed. In five test cases for separate cooling system operating conditions, the diagnostic system's successful diagnosis rate ranged between 73 and 97.7 per cent of the test data.

The f*‐Divergences in the Sequential Observation of Fuzzy Information Systems

On the basis of the f*‐Divergence for fuzzy information systems, this article presents a sequential selection method for a fixed number of fuzzy systems. f*‐Divergence is a measure of the quantity of information concerning the state space provided by the fuzzy system when the a priori probability distribution is defined on the space. The method described by the author determines a procedure which maximises the “Terminal” f*‐Divergence.