Surrogate-based optimization of location hole for contactless power transmission system

This paper designs and optimizes a location hole for contactless power transmission (CPT) system on autonomous underwater vehicles (AUV) hull. Through mechanic coordination, AUV could keep stable and high efficiency for charging and data transmission. Optimization target is to minimize drag coefficient and satisfy volume constraint. A simulation-based computing framework is established for drag coefficient calculation, including geometric parameterization, mesh auto-generation and solver. In the framework, 3D structured mesh is adapted for accurate and efficient calculation, which is a major difference from previous researches. Optimization algorithm is based on surrogate model with parallel sampling strategy. Three sampling methods pick targeted samples and gain responses through parallel computing at each iteration. Results show that initial design doesn't increase much resistance, nearly 3%. Drag coefficient of optimized hull is 2.26% larger than that without hole, a little smaller than initial design. However, on the basis of initial increment of drag, optimized design decreases nearly 20%. This paper also discusses effect of depth and length of location hole on drag and results show greater depth and smaller length contribute to drag reduction. The framework and optimization algorithm can also be employed to optimize location and shape of other appendages on AUV.