To analyze the harmonic characteristics and interaction mechanism of modular multilevel converter based multi-terminal HVDC system (MMC-MTDC), firstly, the concept of average switching function is introduced, and the detailed time domain model of MMC is established. The simplified model of MMC is obtained by neglecting the internal dynamics, thus the equivalent circuit of MMC-MTDC is obtained, and the transfer function of DC side harmonic current to DC side harmonic voltage on each converter is deduced. Secondly, by defining the dynamic phasors of each order of AC side current and DC side voltage as state variables, the time domain model of MMC is transformed into harmonic analysis model with dynamic phasor sequence components, which is used to analyze and obtain the harmonic distribution characteristics and harmonic transmission mechanism at AC/DC side of MMC. By describing the frequency characteristics of the transfer function of DC side harmonic current to DC side harmonic voltage on each converters, the resonant characteristics of MMC-MTDC and the interaction mechanism of harmonics between different converters are analyzed. Finally, the correctness of the analysis of harmonic characteristics and interaction mechanism are verified by simulation on MATLAB/Simulink. The analysis results can provide theoretical support for proposing harmonic suppression strategy.

With the large-scale application of new power electronics technologies such as renewable energy generation, flexible HVDC and STATCOM, the harmonic problem of AC/DC hybrid power grid is becoming more and more prominent. This paper combs the basic concept of harmonic instability of AC/DC hybrid system. The mechanism of complementary resonance and core saturation instability is emphasized, and the solution of the harmonic instability problem is proposed. In addition, the harmonic instability events are investigated comprehensively, and the mechanism and suppression measures of harmonic instability in Niuzhai-Congxi HVDC, Luxi Back to Back HVDC and Yongren-Funing HVDC are analyzed.

A fast protection scheme considering high-resistance fault is proposed for multi-terminal DC grid with current limiting inductance. The rate of change of voltage measured at the front-end of current limiting inductance is used as the main protection criterion in the new algorithm, and an auxiliary criterion, which uses the wavelet-based high frequency energy, is added to improve the performance of protection under high resistance faults. The auxiliary criterion extracts the high-frequency characteristics of the initial and reflected traveling waves by using stationary wavelet transform (SWT), and can improve the sensitivity of the protection significantly under high resistance faults. The main and auxiliary criterion of the protection scheme complement each other in terms of speed and tolerance to fault resistance. In addition, the protection scheme only utilizes the local signals, and thus has no signal communication delay, which can satisfy the requirement of protection speed of multi-terminal DC grid. At last, a four-terminal DC grid is modeled based on PSCAD/EMTDC simulation platform, and substantial simulation results verify the effectiveness of the proposed method.

At the present live-line measurement practice of non-gap metal-oxide surge arresters, there are problems of inaccurate measurement results. Especially the dominant capacitive current compensation method which brings large principal deviation without extracting potential transformer secondary signal or based on station transformer reference voltage signal, would cause false or missing judgment. Combined with specific test cases, the application and problems of the prevailing live-line measurement methods of non-gap metal-oxide surge arresters are summarized, such as capacitive current compensation method, triple frequency harmonic method and the infrared thermography method. And the corresponding solution for the above problems is discussed. An inversion method of both resistive current and harmonic current component based on the total-current harmonic waves processing is also presented, and the simulating and comparing experiments’ results show that the method accords with the safe, simple and accurate tendency of live-line measurement of arresters.

For the harmonic instability problem under DC bias caused by converter transformers in HVDC, firstly, from the view of circuit principle, and based on the impedance analysis method, the nonlinear process is linearized by using switching function and two-stage equivalent magnetization curve of the transformer, and the system loop equation is listed. Secondly, based on Nyquist criterion, an engineering criterion for judging whether the harmonic instability is caused by DC bias is derived. According to the criterion, the main factors affecting the harmonic instability are analyzed comprehensively. Finally, based on the Cigre model, the criterion is verified by simulation, and the correctness of the criterion is verified by simulation results.

To solve the problem of numerical oscillation in electromagnetic transient calculation and improve the computational efficiency of electromagnetic transient simulation, this paper proposes a new electromagnetic transient parallel computing method, which uses 2-stage 3-order Radau ⅡA method to discrete continuously at multiple time points, and then solve the problem by applying Newton method. Since the 2-stage 3-order Radau ⅡA method has good calculation accuracy and stability, the proposed method can not only perform parallel calculations at a larger step, but also eliminate numerical oscillation. Case calculations prove that this method not only eliminates the numerical oscillation, but also effectively improves the accuracy, efficiency and real-time performance of the electromagnetic transient simulation of the power system.

To resolve the voltage violation and fluctuation in active distribution network (ADN), this paper proposes a two-level coordinated control method of the traditional voltage regulation devices (TVRD) and the distributed generators (DGs) based on multi-timescales. The outer-level is an intraday rolling optimization, using the TVRD and DGs to sustain normal voltages of all nodes with a relatively long time interval. In this level, the minimum network loss is taken as the objective function to set up the optimization model, and it is solved by applying the second-order cone programing algorithm. In the inner-level, only DGs are controlled to regulate online the critical nodes’ voltage within the time step of the outer-level. The online optimization model is built based on voltage sensibility, considering the minimum of the difference between critical nodes’ voltage and its reference value and control cost as the objective. The proposed scheme is verified on a modified IEEE 37-bus distribution network. The results show that DGs and TVRD can work together to restrain the frequent fluctuation of the critical nodes’ voltage, keep the network voltages within the limits and relax the burden on the TVRD operation.

The insulator icing is one of the major threats to power network. To satisfy the need of simulation of insulator icing, it’s necessary to obtain the local collision efficiency of water droplets on insulator efficiently.Taking advantage of Eulerian-Lagrangian method and Eulerian-Eulerian method, the properties of local collision efficiency β on insulator surface are investigated. Insulator LXY-120 is used as the test sample, with layered approach, the β at different places (surface, steel cap and insulator edge) are analyzed under different environmental paratmeters (wind speed U and medial volume diameter MVD). It is found that the accuracy of the two methods are similar, but the latter Eulerian-Eulerian method has a better calculation efficiency. The results show that β decreases from insulator centerline, the value of β on insulator edge and steel cap are much greater than that on surface. Meanwhile, the β at insulator edge and steel cap increase with the increasing of U and MVD,there most likely is a saturation of β when U and MVD are high.

Frequent occurrence of ice disaster is one of the most threats to the power system. The intelligent technologies such as cloud computing, big data, internet of things, mobile internet and artificial intelligence continue to integrate, innovate and develop rapidly, and will affect profoundly the development of anti-icing technology in power system. Firstly, an overview of the connotation of intelligent technologies and their applications in power system is summarized,its application scenes in pre-disaster prevention and early warning, ice monitoring, melting and de-icing technology and disaster relief are discussed. Secondly, the application route scheme of intelligent technologies in anti-icing work is highlighted, and the solution path of related engineering and technical problems with specific intelligent technology is analyzed. Finally, it is verified that the intelligent technologies have a wide range of application scenarios in anti-icing work through analysis, furthermore, the development direction of anti-icing intelligence is pointed out systematically.