Objective  This study aims to improve the prediction accuracy of acoustic scattering characteristics for underwater vehicles while maintaining computational efficiency, offering a novel technical approach for acoustic stealth design and rapid prediction of complex underwater targets. To achieve this, a modified highlight model method based on equivalent reflection coefficients is proposed.

Method The proposed method introduces equivalent reflection coefficients for each component of the underwater vehicle and utilizes a genetic algorithm (GA) to invert these coefficients. By segmenting the target into manageable components and using known data to invert the reflection coefficients, the modified highlight model is established.

Results Compared to the conventional planar element method, the modified highlight model method achieves an average error of less than 4 dB across typical frequencies, with a computational efficiency improvement of approximately 1000 times. The method demonstrates strong applicability and accuracy across different underwater vehicle models with varying complexities, maintaining second-level computational time while showcasing excellent real-time capabilities.

Conclusion The study concludes that the modified highlight model method significantly enhances the prediction accuracy of acoustic scattering characteristics for complex underwater targets while maintaining computational efficiency. Future work could focus on developing a comprehensive database of equivalent reflection coefficients for various underwater acoustic targets across multiple frequency bands, which would enable broader and faster prediction of scattering characteristics, offering considerable engineering application value.