作者: Martina Daun , Frank Ehlers
DOI: 10.1155/2010/461538
关键词: Tracking system 、 Artificial intelligence 、 Clutter 、 Algorithm 、 Sensor fusion 、 Computer science 、 Tracking (particle physics) 、 Sonar 、 Sensitivity (control systems) 、 Computer vision
摘要: Activated reconnaissance systems based on target illumination are of high importance for surveillance tasks where targets nonemitting. Multistatic configurations, wheremultiple illuminators and multiple receivers located separately, particular interest. The fusion measurements is a prerequisite extracting maintaining tracks. inherent ambiguity the data makes use adequate algorithms, such as hypothesis tracking, inevitable. For their design, understanding residual clutter, sensor resolution characteristic impact propagation medium important. This leads to precise models, which able determine performance team. Incorporating these models in multihypothesis tracking situationally aware algorithm. Various implementations this algorithm evaluated with help simulated measured sets. model knowledge increased performance, but only if line physical reality: we need find compromise between refined robust models. Furthermore, implement model, inherently nonlinear multistatic sonar, approximations have be made. When engineering system, sensitivity studies tune assumptions approximations.