The Algorithm And Application Of Distributed Information Fusion Based On Fieldbus

Over the last 20 years, information fusion technology has undergone rapid development, which made the exchange of information easier and faster. However, with the further rapid technological development, various fields growing complexity makes industrial application or monitoring environmental problems facing information overload. It urgent need for more practical information fusion algorithm to a plethora of information for rapid digestion, interpretation and assessment, to improve the automatic information processing and complete the estimation and decision, at the same time, it also promote the multi-sensor information processing technology develop rapidly. Information fusion technology is of great significance because it is the key of information getting technology. Distributed Information fusion technology because of its simple structure, high accuracy, reliability, and the advantages of structured, has become a key information fusion research direction and applied in many field. In this paper, we study the algorithm of distributed information fusion based on Fieldbus and achieve better results. Main content are as follow:1. Analyze the basic concept of information Fusion, the system features and algorithm. Present the model of distributed information fusion based on Fieldbus and detailed analyze the distributed Information Fusion principle and characters.2. On the basis of the analysis of basic Kalman filter and strong-tracking Kalman filter, we propose a modified strong-tracking Kalman filter and a modified strong-tracking Kalman filter distributed information fusion algorithm. In this method, we adopted modified strong-tracking Kalman filter on the bottom information processing, and the smallest variance estimation error method used on the top. Fusion results have high precision and a very strong real-time tracking capability for urgent change signal. Simulation shows the validity and reliability of this method.3. Based on the analysis of BP Neural Network, we integrated modified strong-tracking Kalman filter distributed information fusion and BP Neural Network and propose modified strong-tracking Kalman filter Neural Network distributed information fusion algorithm. We used modified strong-tracking Kalman for original fusion and its result processed by BP neural network, which is reliable and efficient. We combined this method with Fieldbus data transmission network for environmental monitoring. Simulation shows the validity and reliability of this method. In the last,by analyzing and discussing, a new strategy of distributed information fusion is proposed based on the evaluation of the fusion algorithm capability.