Maximum efficiency tracking design of wireless power transmission system based on machine learning

Abstract

For the typical magnetically coupled wireless power transmission system, power consumption is often caused by the impedance effect of the primary and secondary sides, the loss and heat generation of the actual semiconductor devices and the change of the coupling coefficient caused by distance. The uncertainty of frequency tracking efficiency caused by high enough coil transmission with low load receiving efficiency and load impedance changes makes it difficult to stabilize the output of the optimal efficiency point of the system. In order to solve this problem, a wireless power transmission system is proposed based on the traditional wireless power transmission system a wireless power transmission system is proposed to predict and stabilize the variable impedance, so as to stably track the efficiency of the system in the resonant frequency range. It is not affected by load jitter, actual circuit parameters and environmental disturbance. The test results show that the transmission efficiency of tracking frequency and predicting impedance can be realized at a certain distance, and the difference between the prediction error and the actual efficiency is about 1%–3%.