Source signal recovery method based on population entropy quantum Remora fish optimization algorithm

Within the domain of underdetermined blind source separation, the efficacy of existing source signal recovery methodologies is critically contingent upon the precision of initial signal estimation. Suboptimal performance frequently arises from this dependency, stemming from the insufficiency of initial parameter configurations. This study introduces an innovative and efficacious technique for the recovery of source signals, specifically designed to overcome the encountered engineering obstacles. Thus, we introduce an innovative Remora fish optimization algorithm with population entropy and quantum-behaved characteristics, designed to enhance the optimization of the objective function through the integration of a dynamic penalty function. The preliminary estimated signal configuration for source recovery, utilizing the radial basis function network, is transformed into an initial candidate solution within the population. The solution characterized by optimal parameters, derived from the application of the population entropy quantum Remora fish optimization algorithm, is employed as an updated initial estimation signal to facilitate effective source recovery. The extensive simulation outcomes indicate that the method proposed in this study possesses enhanced robustness and broader applicability, thereby rendering the initial parameter configuration more rational.