Triboelectric nanogenerators (TENGs) are extensively applied in the energy harvesting field owing to their advantages of low cost, diversified structural design, and superior conversion efficiency from low-frequency mechanical energy. Among them, the grating/disk-structured TENG is one of the most promising types because it can continuously provide a high output. However, owing to its complex structure, modeling simulations and structural optimization of grating-structured TENGs in freestanding mode have not been performed previously. Moreover, the influence of nonideal structural parameters such as height and parasitic capacitance on the performance is unknown. Furthermore, more advanced numerical data processing methods are expected to be proposed. In this work, we obtained the optimal grating number n and optimal average power of cylindrical grating-structured TENGs using the support vector regression algorithm and other numerical analysis methods during simulation. Subsequently, we studied the influence of the gap h between the rotor and the stator and the parasitic capacitance C_p on the performance of the grating-structured TENGs. This work can complement the previous structural optimization and simulation works on grating-structured TENGs, providing a reference for the application of machine learning to the structural optimization of TENGs.

摩擦纳米发电机（TENG）具有成本低、结构设计多样化、低频机械能转换效率高等优点，在能量收集领域得到了广泛的应用。其中，光栅/盘结构TENG是最有前途的类型之一，因为它可以持续提供高输出。然而，由于其复杂的结构，光栅结构的 TENGs 在独立模式下的建模模拟和结构优化之前尚未进行过。此外，高度和寄生电容等非理想结构参数对性能的影响尚不清楚。此外，预计将提出更先进的数值数据处理方法。在这项工作中，我们获得了最佳光栅数n以及在仿真过程中使用支持向量回归算法和其他数值分析方法的圆柱形光栅结构TENG的最佳平均功率。随后，我们研究了转子和定子之间的间隙h和寄生电容C _p对光栅结构 TENG 性能的影响。这项工作可以补充之前光栅结构TENGs的结构优化和仿真工作，为机器学习在TENGs结构优化中的应用提供参考。