In order to monitor current distribution state of parallel silicon carbide（SiC）devices in photovoltaic inverter，a contactless monitoring solution based on anisotropic magnetoresistive sensor was proposed. According to practical position of SiC devices on PCB board，multiphysics simulation tool COMSOL was employed to analyze magnetic distribution around SiC devices. Then，the best location of magnetoresistive sensor was determined. Current monitoring unit consisting of magnetoresistive sensor，signal conditioning circuit and data communication interface was formed. Testing of a 1.5 kW photovoltaic inverter prototype was implemented. The experimental results demonstrate the contactless current mismatch monitoring solution features high bandwidth，high sensitivity，good linearity and simple circuit structure.

To solve the chaos problem of inverter in new energy generation system，a regulation scheme of H-bridge inverter based on active disturbance rejection control（ADRC） was proposed. Aiming at the phenomenon of poor steady-state and dynamic characteristics of inverters in chaotic state，a discrete model of the inverter was developed to control the chaos by establishing tracking differentiators，extended state observers，and state error feedback control laws. Through bifurcation diagram，folding diagram and simulation，the results show that the stability range of the H-bridge inverter based on ADRC control is expanded by 45% compared with the H-bridge inverter based on proportional control，which effectively suppresses the occurrence of chaos. It can be concluded that the control strategy can restrain the chaotic behavior of the system and broaden the stable working range of the system.

In recent years，with the increasing shortage of energy，the ship power system is transitioning towards new energy upgrading. However，the uncertainty of new energy output has also brought new challenges to the economic and safe operation of the system. Therefore，traditional ship energy management is no longer applicable，and there is an urgent need for a comprehensive energy management system suitable for modern ships. In response to the above situation，a comprehensive energy management strategy using energy optimization scheduling was proposed，coordinated control at the upper level，and a combination of intelligent algorithms. A new energy ship microgrid system model was constructed，and four different operating conditions of the ship on the corresponding simulation platform were simulated，including accelerating navigation，normal navigation，decelerating navigation，and berthing. Finally， simulation models of various parts of the system on the Matlab/Simulink platform were built，and the simulation results verify that the strategy proposed can achieve an efficient balance of power supply and demand on both sides while maintaining the DC side bus voltage and system stability.

In order to address the issue of large output voltage variation from self-powered current transformer in smart measurement switch，ultra wide input voltage range post-stage DC-DC module is required to generate stable system voltage rails. A ultra wide input DC-DC module based on constant on-time control and asynchronous Buck circuit operating in discontinuous mode was proposed. Accroding to the gap between intput voltage and output voltage，variable frequency control and quasi-fixed freqency control were combined. With large voltage gap conversion realized，this DC-DC module also features super fast dynamic response. PSIM software was used to analyze and verify the characteristics of the propsed DC-DC module. A 5 W hardware prototype was fabricated and tested to produce key experimental waveforms. The experimental results demonstrate that the proposed DC-DC module has features of ultra wide input range，super fast dynamic response and high stability.

An adaptive active arc suppression method based on S-transform correlation degree for distribution network unidirectional grounding was proposed to effectively eliminate single-phase grounding fault arc in distribution network and improve power system stability. The time series data of zero-sequence current of distribution network lines were collected，the time-frequency information matrix of zero-sequence current through S-transformation was obtained，the correlation matrix of time-frequency information matrix between each line was calculated and obtained. Input the convolution neural network model，and output the single-phase grounding fault line selection results；the double closed-loop adaptive control method based on fuzzy quasi-PR and PI control was adopted in the fault line to inject zero-sequence current into the neutral point of the fault line of the distribution network，effectively control the phase and amplitude of the injected current，force the voltage value of the phase of the fault line of the distribution network to be 0，and complete the adaptive active arc suppression. The experimental results show that this method can effectively identify the single-phase grounding fault lines in the distribution network，realize the single-phase grounding adaptive active arc suppression，and complete the single-phase grounding adaptive active arc suppression work as soon as possible to ensure the safe and stable operation of the distribution network.

Aiming at the low voltage ride through（LVRT）problem of the combined grid-connected system with new energy-energy storage under the control of the grid-forming converter，an additional voltage limiter control scheme based on the traditional LVRT was proposed. Firstly，the shortcomings of the traditional LVRT strategy based on current limitation was analyzed. When the fault duration is long，only setting current limitation would easily lead to synchronization instability. Then，an improved LVRT strategy with additional voltage limiter was proposed to suppress the fault current under a longer time scale and improve the synchronization stability under the fault. Finally，a simulation model of new energy-energy storage combined grid system was built in Matlab/Simulink to validate the proposed scheme. The experimental results show that the proposed scheme can ensure that the new energy-energy storage combined grid-connected system not be taken off-grid during the three-phase symmetrical sag fault，and effectively improved the synchronization stability of the system during the fault.

Permanent magnet synchronous machines（PMSM）has serious chattering problem when traditional sliding mode observer（SMO）is applied to low-speed operation with fixed sliding mode gain. To solve this problem，a full-order sliding mode observer（FSMO）with fuzzy adaptive adjustment of sliding mode gain was proposed，which can improve the chattering suppression performance in a wide speed range. First，a FSMO was constructed with stator current and extended back-EMF as observation objects. In addition，the fuzzy control rules were constructed by using the error and error change-rate of the actual and observed values of the stator current as the input，and the adaptive adjustment of the sliding mode gain was realized according to the different speed of the motor. Then，the soft switching continuous sliding mode control was realized by using hyperbolic tangent function instead of traditional sign function. Moreover，in order to eliminate the influence of variable speed，the normalized orthogonal phase-locked loop was used to obtain the accurate rotor position from the extended back-EMF. Finally，the PMSM drive test platform was built，and the performance of low speed and medium-high speed was tested. The experimental results show that the proposed algorithm has significant advantages in wide speed range over traditional algorithms.

Accurately and quickly identifying the fault types of traction transformers is a key technology for intelligent operation and maintenance. Aiming at the problems of single model deviation in the current traditional algorithm and the constraints between the iteration rate of complex models and the deployment of computing resources，a traction transformer fault diagnosis model based on the Stacking ensemble learning framework was proposed，and incorporated knowledge distillation technology to compress model iteration time to improve the computational performance of the model. First，an evaluation feature vector composed of gas indicators in transformer oil was constructed，and then the single Bagging and Boosting framework algorithm were combined based on the Stacking integrated learning framework，and knowledge distillation technology was incorporated to realize the effective mapping of feature vectors and fault types. The actual generalization effect in the DGA data sample shows that this method solves the problem of bias and variance in the traditional integrated model，accelerates the iteration speed of the integrated model，and proves the engineering application value of the model.

To cope with the impact of electric vehicle（EV）charging loads on the power grid and achieve a balance of interests between the power grid，charging station and users，a customized charging strategy for EVs that takes dynamic electricity prices into consideration was proposed. Initially，the dynamic pricing mechanism was designed for charging station based on the needs of tripartite interests. Then，the optimization model with the lowest user charging cost and the lowest load fluctuation rate of power grid had been established to perfect the charging process of EV. Moreover，on the basis of the traditional artificial bee colony（ABC）algorithm，the adaptive normal attenuation coefficient was introduced to form the adaptive ABC algorithm and it was applied to solve the optimization model in order to obtain the customized ordered charging scheme. Furthermore，the charging process of each EV was optimized by combining the dynamic electricity pricing mechanism and EV customized charging method. Finally，based on Monte Carlo method，the charging situation of different number of electric vehicles under different charging modes was simulated. The simulation results show that the proposed method can significantly improve the load index of the power grid，ensure the revenue of charging station，reduce the charging costs of users，and achieve an all-win benefits between power grid，charging station and users.

With the increased penetration rate of new energy year by year，it is difficult to accurately predict the randomness and fluctuation characteristics of its output，causing a severe challenge to the operation，planning and scheduling of electrical power system. Therefore，modeling for the uncertainty of new energy has attracted more and more attention. To obtain the time sequence characteristics of new energy output scenario more effectively，a new energy scenario generation method was proposed based on data drive，and combined self-attention mechanism with generative adversarial network discriminator with gradient penalty through applying the SA/WGAN model. Through building a deep learning model based on the combination of two models，effectively highlight the timing sequence characteristics of new energy output scenario and enhancing the nonlinear fitting capability in scenario generation. The example results show that，compared with the scenario generation results of original WGAN and WGAN-LSTM，the new energy generation scenario of proposed model can not only effectively improve the accuracy，but also possess the advantages of stable WGAN-GP training results and quick SA calculation speed，which can achieve a more efficient generation of scenarios that is close to the distribution of real new energy scenario.