The detection system has important functions such as surveillance, search, capture, tracking and identification, and plays a pivotal role in the military and civilian undertakings of all countries in the world. The purpose of this article is to study the research and data processing of multi-source exploration data modeling technology. The multi-source data processing methods are sorted out, the existing single non-acoustic detection data modeling technology is analyzed and summarized, the development status of domestic and foreign underwater detection technology, mine anti-submarine technology and the basic knowledge of underwater detection are summarized. The significance and main content of the research work of the thesis: A multi-source non-acoustic detection data fusion modeling technology based on the federated Kalman filter is proposed. Launched the research on the modeling and simulation of lidar detection and magnetic anomaly detection, and carried out the comparison experiment of multi-source non-acoustic detection data and single-feature non-acoustic detection based on the federated Kalman filter. The results It shows that the detection results of lidar detection and magnetic anomaly detection are fused by the federated Kalman filter to obtain the global optimal detection, in which the position of the navigation trajectory in the x-direction 4000m and the y-direction 400m is accurately located.