| C~4ISR means "Command, Control, Communication, Computer, Information, Surveillance and Reconnaissance". It could help a commander to control information and the weapon in modern high technology war so as to obtain higher battle effectiveness. Therefore, it's significant to develop C~4ISR. Data fusion was mainly used to confuse the data from many information sources. Compared with other datum received from the single information source, it could offer more accurate and certainty data. Information fusion is one of the key technologies in C~4ISR system, its function model includes signal testing, position estimation and status estimation in low layer, as well as situation estimation and menace estimation in high layer. It also is a new technology based on traditional sciences and modern engineering domain.This thesis discussed the develop direction of C~4ISR information fusion system, then put forward a viewpoint that the fuzzy neural network method that put forward in this paper is the basis of C~4ISR. The development on intelligent aspect will bring intelligent of C~4ISR system certainly. And it's more and more necessary of intelligent information fusion technology in future battlefield, especially the C~4ISR system. On the foundation of analyzing the characteristic of C~4ISR and requirement of command control system in information battlefield, this paper configured information fusion frame of C ISR in order to solve the problem of real time, fault-tolerant and illation of uncertainty information. Then the information fusion was discussed in application of target tracking and identification in C~4ISR system.Target tracking is one important aspect of low layer in C~4ISR information fusion. For a tracking problem of mobile target, using current statistical mathematics model, we discussed the neural network self-truing target tracking algorithm in detail. Aiming at the contradiction between speed and accuracy in traditional target tracking technology and system variance was adjusted by direct ratio relationship between Kalrnan filtering tracking system variance and acceleration variance, this thesis put forward a self-truing target tracking algorithm used improve BP network, this algorithm adjusted system variance by using the relationship of speed residual and acceleration, in order to tracking target accurately. The simulation results showed that this improve algorithm was better than traditional algorithm.The other aspect of information fusion is target identification, which is the basis of situation estimation and menace estimation. We abstracted information property according to datum fusion, and solved the problem of uncertainty and so on by using fuzzy neural network. The simulation results showed that adopting fuzzy neural network to identify target is feasible. |
