Aimed at uncertainties in the output from grid-connected wind turbine, the scenario analysis method based on probability occurrence is adopted to transform the uncertainty model into a multi-scenario problem with different occurrence probabilities, and a reactive power optimization model with the goal of minimizing the active power loss and voltage deviation is established. In view of the poor diversity of Pareto frontiers obtained using the traditional methods, an adaptive grid multi-objective particle swarm optimization (AG-MOPSO) algorithm is proposed, which uses adaptive grids to obtain the density of particles in external archives, selects the global optimal particles and maintains the scale of the external storage library according to the density information as well as the betting mechanism, thus effectively ensuring the uniformity and diversity of the Pareto frontier distribution. This algorithm is used to perform reactive power optimization calculations on an IEEE 33-bus system with wind power, and it is also compared with the existing NSGA-II algorithm. Results show that the Pareto frontier obtained using this algorithm is better, which verifies the feasibility of the proposed model and algorithm.

The traditional phase-locked loops (PLLs) in a VSC-HVDC system such as a synchronous reference frame PLL (SRF-PLL) and a dual second-order generalized integrator PLL (DSOGI-PLL) cannot accurately track the positive-sequence voltage phase of grid fundamental wave under non-ideal power grid, which will cause different degrees of phase-locked error, affect the control performance of VSC and reduce the system stability. To solve this problem, an improved DSOGI-PLL scheme is proposed. First, the attenuation characteristics of SOGI-QSG at different frequencies are analyzed according to the Bode diagram, and the limitations of DSOGI-PLL applications are obtained. Then, based on the voltage of harmonic grid, the internal model of repetitive control is introduced on the basis of DSOGI to realize the real-time tracking and regulation of harmonic signals, thus suppressing the harmonic voltage interference. At the same time, considering the DC bias and voltage frequency fluctuation of power grid, a method of DC bias elimination and frequency adaptation is proposed to realize the adaptive phase tracking of power grid. Finally, the superiority of the proposed strategy was verified through a comparison between the simulation and experimental results.

To improve the metal object detection performance for an electric vehicle wireless charging system, a sensitivity optimization method for the detection coil was proposed. The equivalent electromagnetic model of a rectangular detection coil with a metal object approaching was established. The changing mechanism of the physical parameters of the coil was studied, and a theoretical formula for changes in the parameters of the detection coil which were caused by metal objects of different sizes was obtained. The influence of several configuration parameters on the sensitivity of the detection coil was analyzed, and the configuration of the detection coil was optimized by combining the actual operating conditions of wireless charging. Through the electromagnetic field simulation and experimental verification, the accuracy of the theoretical model was proved. Finally, a metal object detection experiment was carried out in a 3 kW wireless charging system, which verified that the metal object detection system built on the basis of the optimized detection coil can detect metal objects with a size of 25 mm and above. In addition, it had anti-interference capability.

Compared with the traditional plug-in charging method, it is safer and more convenient to employ an inductive power transfer (IPT) system to charge autonomous underwater vehicles (AUVs). To alleviate the strong magnetic field inside the AUV hull and the dramatic power fluctuation caused by the rotation misalignment of the AUV under the turbulent water, a three-phase IPT system with a novel coupling structure is proposed. The coupler is composed of three transmitting coils and four receiving coils connected in alternating reverse series, which can suppress the central magnetic field and improve the anti-rotation misalignment performance simultaneously. The Maxwell simulation results show that when the AUV hull rotates, the equivalent mutual inductance Meq fluctuation is less than 2%, and the magnetic field of the AUV center always maintains a low level. In addition, to simplify the system analysis, a decoupling method based on a passive component is adopted to decouple the three transmitting coils. A laboratory-scale prototype based on an LCC-S compensation topology was built to verify the feasibility of the system. Experimental results show that when AUV rotated, the output power varied from 536 W to 595 W with a maximum fluctuation of 9.91%. The maximum DC-DC efficiency of the system was 86.28%.

At present, the transmitting coil in a dynamic wireless charging system for electric vehicles usually adopts a segmented guide rail structure to realize the relay dynamic wireless charging. However, the problem of mutual inductance drop will occur at the switching of the guide rail and result in the reduction of the system transmission efficiency, and this is more prominent at the corner. A corner dynamic wireless charging model was established, and the relationship between the mutual inductance of primary and secondary coils and the deflection angle was deduced through theoretical analysis. An improved structure of the guide rail transmitting coil at the corner was proposed and simulated, and a corner dynamic wireless charging platform based on resonance magnetic coupling was built. Simulation and experimental results show that by using the improved guide rail transmitting coil structure, the mutual inductance drop of the wireless charging system was significantly reduced, and the system transmission efficiency was improved by 8.66% at the maximum deflection angle, thereby verifying the effectiveness of the improved coil structure.

The state-of-charge(SOC) of lithium-ion battery is an important parameter for the operation and maintenance of a battery management system(BMS), and its accurate estimation is related to the real-time monitoring and safety control of lithium-ion battery. The traditional unscented Kalman filter(UKF) algorithm has the risk of making the covariance matrix negative when estimating the SOC of lithium battery, and the estimation accuracy is not optimal. To solve the shortcomings of this algorithm, a ternary lithium-ion battery is taken as the research object, and a second-order RC equivalent circuit model is established to describe the working characteristics of the battery. Based on the traditional UKF algorithm, a square-root double unscented Kalman filter(SR-DUKF) algorithm with double unscented transformation is proposed, and it is verified under multiple working conditions. Experimental results show that the improved SR-DUKF algorithm can better estimate the SOC of lithium-ion battery based on the second-order RC equivalent circuit. The average errors under HPPC and BBDST conditions are 0.59% and 0.52%, respectively, and the convergence times are 60 s and 110s, respectively, which verifies that the improved SR-DUKF algorithm has a higher estimation accuracy, better convergence and better robustness.

To obtain the state-of-charge (SOC) estimation value well, a second-order equivalent circuit model is selected as the research object. Aimed at the disadvantage that the recursive least squares method with a forgetting factor is easy to be disturbed by environmental factors such as noises in the parameter identification, a bias compensation recursive least squares method is proposed to realize the accurate identification of model parameters, and the SOC is estimated combined with the unscented Kalman filter algorithm. In view of the disadvantages of the unscented Kalman filter algorithm such as poor stability, the weight vectors are used to update the Kalman filter gain in the filter algorithm. Experimental results show that the total error of the proposed algorithm in estimating SOC was controlled within 2.7%, which verified the robustness and effectiveness of the algorithm.

Simulation substation batteries often work under discontinuous operation conditions, which will result in capacity regeneration of batteries during their performance degradation. The degradation of batteries shows nonstationary and random characteristics, leading to a low prediction accuracy for the remaining useful life(RUL). Aimed at the problem of RUL prediction of batteries with capacity regeneration, a prediction method is proposed based on variational mode decomposition(VMD) and bat optimized kernel extreme learning machine(Bat-KELM). First, VMD is employed to decompose the battery state-of-health(SOH) time series into overall degradation components and capacity regeneration components. Then, Bat-KELM is used to construct prediction models of each component, so that the prediction accuracy of component trend is improved. At last, the prediction results of all components are blended together to yield the accurate battery SOH prediction results as well as the RUL results. The proposed method is applied to the analysis of battery degradation instance data, and results show its superiority in terms of prediction accuracy compared with the KELM and VMD-KELM models.

The light-emitting diode (LED) driver usually needs a large electrolytic capacitor to reduce its low-frequency current ripple. Therefore, the short life of the electrolytic capacitor is an important factor restricting the life of LED driver, and eliminating the electrolytic capacitor is a key to the long-life LED driver. Under this background, a two-switch electrolytic capacitor-less LED drive circuit topology based on flyback converter is proposed. The auxiliary power balance circuit and flyback converter are integrated, and a balance between the instantaneous input power and output power is realized through two switches. The electrolytic capacitor is eliminated, the low-frequency ripple of output current is suppressed, and a high power factor is realized. The working principle of this topology is analyzed in detail, and a control strategy for its circuit topology is put forward. Finally, a 30 W prototype was built, and experimental results verified the feasibility of the proposed topology.

Aimed at the current imbalance problem in parallel connected multiple light-emitting diode (LED) strings, an efficient multi-output LED driver is proposed to achieve current sharing among different LED strings and improve the efficiency of the circuit. A Boost converter and a series-resonant converter are cascaded to achieve current sharing and a high voltage gain. The transformer leakage inductance suppresses the switch's current rising rate in the turn-on period, which provides the zero-current-switch (ZCS) condition. Moreover, a lossless snubber circuit is added to slow down the switch's voltage rising rate in the turn-off period to significantly reduce the switching loss, and a large magnetizing inductance of transformer is selected to reduce the transformer loss. Meanwhile, only one active switch is used, which reduces the circuit cost and simplifies the control. Therefore, current sharing, high efficiency and high voltage gain are obtained, and the cost is low. The circuit operational principle and operating characteristics together with the specific procedure of parameter design are analyzed and presented in detail. Finally, a 7.9 W two-output prototype with 12 V input voltage was built to verify the effectiveness of the proposed LED driver.