Multiple Radar Track Fusion Algorithm And Research

Multi-radar track fusion is a new developed multidisciplinary emerging technique in recent years. Compared with traditional single radar track, it can increase the credibility of the radar data, and enhance the robustness and reliability of the entire system, and can improve track accuracy, enlarge the system time and radar coverage, improve system utilization and so on.This article describes the basic principles and methods of multi-radar track fusion, which including radar ASTERIX format resolution, tracking filter, the space aligning, and a detailed description of the neural network of the multi-radar track fusion method.According to the characteristics of multi-radar track, this article presents a neural network-based multi-radar track fusion method. In the method, a non-linear model based on neural network was established to solve the multi-radar track fusion problem, which method is composed of the procedures, namely networks architecture, initialization, and training.The experiments prove that multi-radar track fusion can reach high correct percentage in this way. And with the flying target was scanned by more than one radar, the neural network accuracy of fusion can be increased and the fusion accuracy is tending to stability.