Multi-sensor Data Fusion System Data Pretreatment

This topic comes from the Northwestern Polytechnical University' s subject construction - modern airborne radar simulation system. This study mainly carry on the pretreatment research of the data from airborne multi-sensor system. The content includes: time alignment, coordinate alignment and data association.In this thesis. First, we provide two commonly used time alignment method and establish the model. When we study the coordinate alignment, We divide it into two part, platform alignment and system alignment. In each part,we divided the alignment process into two process:coordinate conversion and the correction of system error.We used the airborne multi-sensor multiplatform data fusion as a model to simulate. In each airborne system, We select PD radar,infrared sensor(IR) and electronic support measurement(ESM) to form a typical airborne system. Because the merit of least square measure, we select this method to estimate the system error and amend it. At the usage,we take the random error into consideration and use the technique of slip window to overcome the defect of traditional least square method. The result show this method is very effective in airborne system.At the last chapter of this thesis, we carry a brief research on probabilistic data association, and introduce a more effective improved PDA method. Simulation result show the validity of this method.