Robust H_∞ Control For Uncertain Systems With Both Measurement Data And Control Data Missing

The network control system has been applied to various fields with the continued integration between control technology, computer technology, network communication technology and other academic fields over the last decade. The introduction of network control system brings many advantages, such as low maintenance cost, high flexibility and reliability and so on. However, it will inevitably have some problems, since the data transmit in the network, such as network delay, data loss and so on. In addition, large-scale systems in modern industrial, social and economic fields is usually composed of many subsystems in accordance with certain combinations, the network control not only be used in the subsystem, but also between subsystems. At present, for the study of the data loss problem of networked control systems, most existing achievements are built on the premise of a single system, and only measuring data lose. In this paper, a controller design problem is researched under the presence of both measurement data and control data missing with uncertain dropouts probabilities.(1) For a class of linear discrete system with both measurement d ata and control data missing whose dropouts’ probabilities are unknown, the packet dropouts are modeled as a Bernoulli random binary switching sequence with an unknown conditional probability distribution that is assumed to be in an interval, an observer-based controller is proposed to make the closed- loop system exponentially stable in the sense of mean square and achieve t he prescribed H? performance. Sufficient conditions are derived for the existence of controller with Lyapunov stability theory and linear matrix inequality method. A numerical example is also provided to demonstrate the validity of the proposed design scheme.(2) For a class of uncertain linear discrete system with both measurement data and control data missing whose dropouts’ probabilities are unknown, the packet dropouts are modeled as a Bernoulli random binary switching sequence with an unknown conditional probability distribution that is assumed to be in an interval, using the method of single exponential smoothing to forecast on measurement data, an observer-based predictive controller is proposed to make the closed- loop system exponentially stable in the sense of mean square and achieve t he prescribed H? performance. Sufficient conditions are derived for the existence of controller with Lyapunov stability theory and linear matrix inequality method. A numerical example is also provided to demonstrate the validity of the proposed design scheme.(3) Combined with the network control system and large system characteristics, a class of uncertain large-sacle systems with both measurement data and control data missing whose dropouts’ probabilities are unknown is studied. the packet dropouts are modeled as a Bernoulli random binary switching sequence with an unknown conditional probability distribution that is assumed to be in an interval, using the method of single exponential smoothing to forecast on measurement data for every subsystem, an observer-based predictive controller is proposed to make the closed- loop system exponentially stable in the sense of mean square and achieve t he prescribed H? performance. Sufficient conditions are derived for the existence of controller with Lyapunov stability theory and linear matrix inequality method. A numerical example is also provided to demonstrate the validity of the proposed design scheme.