The research on sensor and data fusion in virtual environment

With the development of virtual reality and virtual prototype technology, the vehicle virtual test system is put forward and possesses wide application perspective. This work presents an overview of the vehicle virtual test system and focuses on sensor and data fusion model and algorithm, which plays an important role in vehicle virtual test. Three modes of fusion function are proposed to fuse and process various data effectively. The concept and structure of data soft-fusion, data hard-fusion and hybrid fusion are given, as well as an interactive data fusion model oriented to the feature level. Furthermore, as an observer for the nonlinear model, the extended Kalman filter (EKF) is adopted to evaluate and monitor the virtual vehicle state (speed, acceleration, position, etc) and virtual test information. And also, we modified the degree of membership algorithm in fuzzy clustering as data association algorithm. By theoretic analysis, the proposed fusion model and algorithms in VE are reasonable. At the end of This work, we conclude with discussion of the future studies on the sensor fusion model and algorithm in VE.