Optimization Model of Multi-Source Input Target Tracking Strategy Based on LSTM Network

In recent years, unmanned aerial vehicle (USV) has been widely used in Marine resource exploration, environmental monitoring and maritime search and rescue. However, most of the existing tracking systems rely on traditional technologies, which are not flexible enough to deal with complex or variable data. In this study, we propose a multi-source input target tracking strategy optimization model based on Long Short-Term Memory (LSTM) network, which significantly improves the autonomous tracking ability of USV in dynamic ocean environments by dynamically selecting the optimal algorithm strategy. The experimental results show that compared with the traditional single scheme, the proposed model can reduce the average error by up to 16%, and the decision-making process is more intelligent, which provides a theoretical and practical basis for the development of unmanned maritime systems.