Information Fusion For Multi-sensor Time-delayed System

In practice, systems are always under time delay. That means the state of system is not only depends on the current state, but also depends on the previous states. The application of time-delay system has a very strong background in all areas of our life, such as biological, ecological, chemical, mechanical, population dynamics, network transmission, computer information and data transmission and so on. At the same time as the development of multi-sensor technology, communication technology, an increasing number of networks are introduced into control systems, which enables sensors distribute into a complex or dangerous operating environment. However, a common character of these systems is the existence of noises, and data loss. These factors make the information provided by one single sensor are difficult to meet the requirement of state estimation. Using multi-sensor measurement data can improve the accuracy of state estimation, thereby reducing the processing errors.In this paper, we study the information fusion algorithm of multi-sensor time-delayed systems with measurement data loss. First, introduce filtering methods and common information fusion algorithms which are widely used in real systems, then based on this knowledge propose Kalman filterand information fusion algorithm for multi-sensor time-delayed systems with measurement data loss.This study focused on two aspects: 1. Based on optimal information fusion criterion weighted by matrix, diagonal matrix, and scalar, we propose an optimal Kalman information fusion filter for time-delay systems with random data loss. Then we study the stability of this Kalman filter system. The cross-covariance matrix of filtering errors between any two-sensor subsystems is derived for time-delay system. Distributed fusion estimator could improve accuracy and easy for fault detection and separation. 2. Based on this model and the state augmentation method, the sufficient condition for the existence of the robust filter, which guarantees the stability of the filtering error system and H? constraints, is given via Lyapunov theory and linear matrix inequality (LMI) . The corresponding H? robust filter is designed via solving LMI. Then we propose an optimal information fusion algorithm to deal with these data.