A Research On The Strategy Of Sensor Management In Fusion Automatic Target Recognition System

This work does a deep research on the strategy of sensor management in fusion automatic target recognition system. We use the expression of sensor performance and the technology of uncertain information processing to overcome the shortage of the traditional methods which are difficult to be realized because of the need of forecasting sensor observations .First of all, we provide a sensor management method based on sensor performance. Considering the sensors' or their groups' performances and using for referenc of the management algorithm based on functions of efficiency and waste, we have rebuilt a new target function, restrictions and optimization rules, and then realized sensor management with programming theory.In allusion to the characteristic that there is much uncertain information during the sensor management process in target recognition system, we use uncertainty reasoning combining linear programming to build up a comparatively systemic sensor management framework. This method can be used both in the case that the sensor observations can be forecasted and in the case that only the sensors' performances are known. The simulation results at the end of the chapter proved the validity and rationality of this method.Then the application of fuzzy controlling in sensor management has been researched. Fuzzy controlling can avoid the problem of accurate sensors and environments modeling. This method fuzzes the sensor performance and the characteristic of targets and environments and then makes a decision according to particular fuzzy reasoning rules.