A Dynamic Evolution Value Assessment Scheme for Maritime Target Groups Based on Attention Mechanism and Hybrid Quantification Methods

The dynamic evolution value assessment of maritime target groups is a complex problem. A key challenge lies in converting physical data into quantifiable indicators and dynamically adjusting weights based on situational changes. These capabilities critically influence assessment accuracy and subsequent decision-making. To accurately assess the dynamic evolution values, an assessment scheme for maritime target groups based on the attention mechanism and hybrid quantification methods is proposed. First, based on analyzing the equipment and characteristics of various types of targets, an indicator system with self-generating capabilities is constructed. Secondly, based on real-time data, quantitative methods for various indicators are proposed. Finally, the attention mechanism is introduced into the fully connected neural network and an intelligent weight learning model with adaptive capability is proposed. The results of the simulation example show that the hybrid quantification methods can achieve accurate quantification of indicators based on real-time information, and the weight learning model can achieve accurate weight distribution under different situations. The proposed assessment scheme is feasible for the assessment of the dynamic evolution value of the maritime target group. It provides an intelligent scheme for the assessment of the dynamic value of the target group.