Optimization of Flood Detection Systems Using Hybrid BP-TRQN Neural Network Algorithms: An In-Depth Review

Abstract

accurate, stable, and efficient real-time detection system. The purpose of this study is to conduct a Systematic Literature Review of the performance of the Hybrid Back Propagation–Time Recurrent Quantum Networks (BP-TRQN) algorithm in optimizing accuracy, learning stability, and data processing efficiency in flood detection systems. This study employed the PRISMA-based Systematic Literature Review (SLR) method to identify, screen, and analyze relevant studies. A total of 44 articles published between 2015 and 2025 were selected from the Scopus, DOAJ, and Google Scholar databases based on predefined inclusion and exclusion criteria. The results show that Hybrid BP-TRQN consistently provides significant improvements over single models, particularly in reducing Root Mean Squared Error (RMSE), increasing the coefficient of determination (R²), and accelerating convergence. The integration of temporal and spatial components and optimization has been proven to strengthen the model's ability to process large-scale hydrometeorological data and satellite images more adaptively. In addition, the application of hybrid algorithms also improves computational efficiency, making it potentially applicable in real-time early warning systems. Thus, BP-TRQN can be considered a promising approach in supporting flood disaster mitigation through a more reliable and responsive detection system.