Application And Research On Humandoid Method Of Information Fusion

The information fusion is the sublimation of the data processing and information processing. Its basic thought is to comprehensively make use of the data information in each aspect of the system and to extract effective information of the concerning object or the environment to the most limit, so as to attain more accurate , more complete understanding the observed object or environment. Therefore, today in the fast development of the intelligent science and technology, researching and applying information fusion is an important research topic which can effectively obtain information and provide system decision.The humanoid information fusion thoughts that this topic put forward is the simulation,extension and prolongation of intellectual characteristics about human perception and the multi-layer information fusion .Humanoid information fusion starts with the angle of artificial intelligence and artificial life and inquiries into information fusion method based on intelligence calculation to acquire a better information fusion effect. This paper first elaborates the information fusion technique, the basic principle about the artificial neural network and rough set and its characteristics used for information fusion. Then it adopts methods such as characteristic analysis, contrast and fusion etc. to study and realize the calculation .At last, it applies a system design and soft hardware experiment to put the principle into the information fusion of intelligent mobile robots and the sample of the coal mine safety.In terms of the calculation research, this paper explores the calculation of information fusion which based on neural network and rough set as well as the basic principle, structure, function and calculation of the data fusion. The knowledge system based on logic rule can be thought as the logic thinking which concentrates on imitating a person, while the artificial neural network can be thought as the image thinking which investigates a person. In order to bring the advantages of humanoid reduction intelligence and imaginable thinking intelligence into play. The paper researches information fusion algorithms about applying rough set theory and neural network. They obtained more ideal fusion effect respectively.As for the realization and application of the information algorithms, the application of humanoid information fusion to the mobile robot is the first applied example. It mainly includes the rule of multi-sensor information fusion when the robot is in the actual environment, the establishment of neural network model and the institution of mobile robots' moving rules. The coal mine safety information fusion is the second applied example. Data reduction based on rough set theories can be done. The paper combines artificial natural network which is suitable for mapping nonlinearity, randomness and complicacy functional relation characteristics. It deals with the coal mine safety information and made decisions according to the result of the information fusion.