Multi-sensor Data Fusion And Its Comprehensive Performance Assessment

It has been unable to meet the military needs just relying on the single senor to provide information. The multi-sensor data fusion of multi-target tracking has become important method to capture the integrated trends of the battlefield and achieve the automation of the military operational commands. In the respective areas of data fusion, there are many kinds of systems and architectures. Every kind of architecture and algorithms share the advantages of their own. How to choose a suitable method depends on its performance evaluation. According to this, the performance evaluation of multi-sensor data fusion has become a hot issue currently. And how to give a integrated assessment on the performance of data fusion is still a difficult issue, which hasn't been reached consensus.A data fusion system is built to implement electronic information fusion system by software. Then the performance indexes of this system are described in detail, and the assessment method is designed.First of all, based on the testing model, the algorithms of track initiation, track identification and data fusion are researched step by step. After that, a testing platform is built though software.Then, the first-born track and second-born track are evaluated respectively on all aspects of system performance requirements face to the combat objectives and based on the relevant research results in this field. Though comparing to the first-born track, this assessment will be more reasonable and more objectively to show the improvement abilities.Finally, the fuzzy evaluation method is used to measure the overall performance and give a score. The practicability of this method is proved by the experimental data though simulation. It will make a foundation for the future more in-depth comprehensive study in this field.