Research Of Data Fusion Technology Applied In The Multi-Sensors Optics Measuring System

Data fusion is an advanced technology that can deal with more than one source information. It synthesizes this source information coming from an object and leads to more precise, more credible and more considerate estimate and judgement than single source information.The concept of data fusion was put forward in the early years of 1970s and developed into a special technology in 1980s as a result of imitating human's ability of dealing with the information of itself. In recent years, with the development of transducer, computer and communication, especially military urgent requirement, data fusion technology is investigated widely. In military field, data fusion is mainly used to C3I task, C4ISR system, object detecting and tracing, object auto-identifying, all kinds of unmanned arm systems. At the same time it has important applied value in such fields as robots, environment monitoring, medical diagnosis and remote sensing etc. But the deep research of data fusion theory only carried out in recent years. And in fact, the theory and method concerning the field are still investigated further and deepen to meet the demand of practice. In the data fusion field, object tracking using more than one sensor is an important branch and absorbed wide attentions in home and oversea academes. The task of multi-transducer data fusion technology is to obtain more precise estimate of object stale and it's flight path and larger scope of object monitoring and tracking than single sensor through dealing with the information from different transducers.The idea of applying the data fusion theory into theodolite system is put forward first in this dissertation. The electric control system of theodolite and tracking sensors' characteristics is investigated deep. Based on these characteristics and data fusion structure theory, the data fusion structure in the theodolite system is advanced. The theory method of multi-sensor data fusion is studied deep. And more, the estimate theory of data fusion is applied concretely to the theodolite with the aim of prognosticating the object state. It is anticipated that the system can still track the object when it is lost instantaneously or shielded to improve acquiring and tracking ability. According to the characteristics of TV and infrared sensor, the adaptive weighted data fusion algorithm is advanced and discussed theorectically. These researches prepare for the theory and experiment foundation of improving the acquiring and tracking ability. reliability and automation.Finally, the estimate theory method of multi-sensor data fusion is analysed theorectically by using least square fit extrapolation and Kalman filter estimation theory. The analysis result proves that the method is feasible. Now the technology is being applied to the theodolite system. The theory analysis Tightness and the value of engineering application are being tested by practice.