考虑水下跨接电容效应的水下磁耦合谐振式无线电能传输系统多参数辨识方法OA

Due to significant differences in physical parameters between marine and aerial environments,including electrical conductivity and relative permittivity,the underwater magnetic coupled resonant wireless power transfer(UMCR-WPT)systems require precise modeling and parameter identification to address dynamic operational challenges.According to the impact of the cross-connected capacitance between the two coils,this paper develops an improved mutual inductance model for UMCR-WPT systems to address variations in mutual inductance and cross-connected capacitance caused by relative coil displacement,as well asreal-time load fluctuations and abrupt resistance changes due to load disturbances during charging. It is essential to acquire the system's operational parameters.The common practice utilizes communication modules to measure parameters through wireless communication between the primary and secondary sides of systems.However,for UMCR-WPT systems,signal transmission conditions in underwater environments are significantly harsher than in air,making them prone to transmission delays,electromagnetic interference,communication security risks,and signal distortion.Current parameter identification methods have the following limitations.(1)Complex circuitry and control mechanisms increase receiver-side volume,raise costs,reduce flexibility,and lower reliability.(2)Relianceon communication methods contradicts the high difficulty of underwater communication.(3)Predominant applicability to aerial wireless power transfer systems,rendering them unsuitable for complex underwater physical environments. Therefore,this study proposes a multi-parameter joint identification method based on the cross-connected capacitance effect in the UMCR-WPT systems.The presented method only requires a single measurement of the transmitter-side voltage and current.It can achieve online joint identification of mutual inductance,cross-connected capacitance,and load resistance through a hybrid genetic algorithm-differential evolution(GA-DE)algorithm.Experimental results show that the maximum identification errors are 2.62％,4.10％,and 4.38％for mutual inductance,cross-connected capacitance,and load resistance under varying coil coupling and load conditions.The maximum deviations in output power and efficiency are approximately 4.451％ and 5.15％,respectively,demonstrating the effectiveness of the proposed multi-parameter identification method. The following conclusions can be drawn.(1)The proposed parameter identification strategy can be effectively implemented on the established experimental platform.(2)The mathematical model of the UMCR-WPT system based on the cross-connected capacitance effect exhibits goodagreement with the efficiency characteristics of the experimental platform.(3)The efficiency improvement of UMCR-WPT systems should focus on reducing inverter and coil losses.