Research On Distance And Foreign Object Detection Algorithm Of Wireless Charging System

In the current mobile phone wireless charging industry,the receiving end of the mobile phone is mainly powered by a wireless charger,but this magnetic induction "tightly coupled" charging method limits the expansion of the charging distance and charging position,which has caused many People have researched on the magnetic induction "loosely coupled" charging system with greater latitude in charging distance and charging position.In the magnetic induction "loosely coupled" charging system,the effective charging distance between the transmitter and the receiver is changed.If the system cannot adjust the starting ping voltage according to the actual distance,starting the ping voltage too high at close range will damage the receiving The Ping start voltage is too low at the end and at a long distance,so that the two ends of the transceiver cannot communicate and cannot be charged.Therefore,the distance detection algorithm is very important for the commercial application of "loosely coupled" wireless charging systems.In recent years,machine learning has been used in many other fields.Machine learning can analyze the relationship between variables well.By applying machine learning to the distance detection in the "loosely coupled" wireless charging system,the safety problem of wireless charging can be effectively solved..When metal foreign objects enter the magnetic field of the wireless charger,the metal foreign objects will cause Joule heat loss due to the eddy current effect in the changing magnetic field,causing the surface temperature of the metal foreign objects to be higher,which may damage the receiving end and the wireless charger May cause harm to human body.Similarly,when the charging distance changes,the previous mainstream foreign object detection algorithms(Q value detection,power loss)will have certain defects,combined with the distance detection and the previous foreign object detection methods to optimize to solve the variable Foreign matter detection problem of "coupled" wireless charger.The research in this thesis is all directed at the safety of wireless charging systems,and its content mainly includes:Firstly,study the communication,system state transition process,system state analysis and system control requirements of the mobile phone wireless charging system,and design the three levels of the software architecture according to the system logic and requirements to study the distance detection algorithm and foreign objects for wireless charging The detection algorithm builds a software platform to facilitate data collection and realize distance and foreign object detection algorithms.Secondly,In view of the current gaps in research on distance detection in the field of wireless charging,the machine learning method should be applied to wireless charging systems.First,through tools such as J-Link and RTT＿VIEW,the data during wireless charging of the mobile phone is collected at different distances,and then Multi-Layer Perceptron(MLP)and Support Vector Regression(SVR))To model the distance detection algorithm,by analyzing various feature data,you can filter out irrelevant data to obtain the feature data used in the final training,and use MLP and SVR to train the distance detection algorithm model to compare the experimental results The comparison results show that the adopted research method has strong feasibility and superior detection effect.Finally,The foreign object detection method based on the coil Q value and the system power loss is studied,and it is found that these two algorithms have great defects in the foreign object detection of the wireless charging system with variable charging distance.Optimization,and finally verified the effect of the algorithm through experiments.Then further think about and try to use the system efficiency for foreign object detection,analyze the test data and propose a method to calculate the system efficiency by setting the compensation power according to the distance,which increases the efficiency difference by 8% with or without foreign objects And the system efficiency design algorithm flow,through analysis found that the designed foreign object detection algorithm has good feasibility.