Tracking control of time-varying signal is a very challenging problem for the network environment applications. An adaptive control strategy based on the inverse of fuzzy singleton model is proposed in the paper. The fuzzy singleton model is a designed equivalent system instead of the fuzzy clustering model of the controlled process. Following an invertibility condition, a collection of predicted control actions are derived from the iterated inverse fuzzy singleton model. Thus, the data dropout and time delays in the network are compensated by means of these predicted values. To enhance control performance, the adaptive control strategy is adopted. Since the method is started from the inputs and outputs of the process, it is actually a data-based solution which is very suitable to the processes with blurred mechanism. Compared with other two control algorithms, the proposed control algorithm exhibits good accuracy, high efficiency, and fast tracking features. Simulations in the data dropout and time-delay cases have verified the effectiveness of the method.
Tracking control of time-varying signal is a very challenging problem for the network environment applications. An adaptive control strategy based on the inverse of fuzzy singleton model is proposed in the paper. The fuzzy singleton model is a designed equivalent system instead of the fuzzy clustering model of the controlled process. Following an invertibility condition, a collection of predicted control actions are derived from the iterated inverse fuzzy singleton model. Thus, the data dropout and time delays in the network are compensated by means of these predicted values. To enhance control performance, the adaptive control strategy is adopted. Since the method is started from the inputs and outputs of the process, it is actually a data-based solution which is very suitable to the processes with blurred mechanism. Compared with other two control algorithms, the proposed control algorithm exhibits good accuracy, high efficiency, and fast tracking features. Simulations in the data dropout and time-delay cases have verified the effectiveness of the method.