基于改进BP神经网络的无线电能传输系统接收线圈参数优化

Optimization on Receiver Parameters of Wireless Power Transfer System Based on Improved BP Neural Network

为提升无线电能传输(WPT)系统的传输性能,该文研究了磁谐振式无线电能传输(MRWPT)系统的磁通与接收线圈内半径的变化规律,提出了最优接收半径的概念,通过分析平面方形、平面盘式和空间螺旋形三种类型发射线圈WPT系统,揭示了最优接收半径的固有存在性,并探究了最优接收半径随WPT系统参数的变化规律,从而确定影响最优接收半径的参数。在此基础上通过采用遗传算法改进的BP神经网絡对最优接收半径随影响参数的变化规律进行学习,实现了在不同线圈参数下最优接收半径的精确预测。最后通过有限元仿真和实验验证了WPT系统最优接收半径的存在和改进BP神经网络预测结果的准确性。

In order to improve the transmission performance of the wireless power transfer(WPT)system,the concept of the optimal receiver radius is proposed in this paper by studying the variation law of magnetic flux and inner radius of receiver coil in magnetic resonant wireless power transfer(MRWPT)system.By analyzing the WPT system of planar square,planar round and spatial spiral transmitter coils,the inherent existence of the optimal receiver radius is revealed.The variation rule of the optimal receiver radius with the parameters of the WPT system is explored,so as to determine the parameters affecting the optimal receiver radius.On this basis,the BP neural network improved by genetic algorithm is used to learn the variation law of the optimal receiver radius with the influencing parameters,and the accurate prediction of the optimal receiver radius under different coil parameters is realized.Finally,the existence of the optimal receiver radius of the WPT system and accuracy of the prediction results of the improved BP neural network are verified by the finite element simulation and experiments.