Front-End Parameter Identification Method Based on Adam-W Optimization Algorithm for Underwater Wireless Power Transfer System

Monitoring coupler parameters in underwater wireless power transfer (UWPT) systems is crucial for improving the system transmission characteristics. Due to the eddy current effect, the equivalent circuits of the magnetic coupler are more complex and contain more parameters than those in air-based circuits. Traditional parameter identification methods, which rely on solving circuit matrix equations, often struggle with the complex UWPT systems or require significant computational resources to solve high-order multivariate equations. This article proposes a coupler parameter identification method based on the adaptive moment estimation with a weighted adjustment (Adam-W) optimization algorithm to address the multiparameter identification problems in the UWPT system. This method transforms the problem of solving high-order matrix equations into an optimization problem to address multiparameter identification problem caused by seawater eddy effect, which is challenging for traditional methods. In addition, it facilitates online monitoring of the characteristic parameters of the system's coupler by detecting the system's front-end input current and voltage without wireless communication and additional sensor modules, and exhibits better reliability in underwater environments. The experimental results show that the proposed Adam-W algorithm achieves a high parameter identification accuracy within 1 s and 94 iterations with an average error of 2.78%. Comparably, an average error of 56.74% is achieved by gradient descent and 41.49% by adaptive moment estimation, and 37.42% by particle swarm optimization in 150 iterations. The proposed Adam-W algorithm achieves higher accuracy in parameter identification of the UWPT system within a shorter time and demonstrates better applicability in seawater environments.