Proceedings of 2021 International Top-Level Forum on Engineering Science and Technology Development Strategy

Inhaltsverzeichnis

1. A Frequency-Range Selection Method of Fast Traveling-Wave Protection Suitable for VSC-Based DC Grids

   Yingang Tu, Wei Dong, Yijia Yuan, Zongjun Li, Chongshan Wan, Hao Li

   Abstract

   Accurate fault feature extraction is the premise of fast and reliable traveling-wave (TW) protection for VSC-based DC grids. The frequency-dependent transmission line parameters make wave impedance and propagation function perform different characteristics in different frequency ranges. In this paper, a frequency range selection method of fast TW protection suitable for VSC-based DC grids is proposed. Firstly, the features of fault initial voltage TW are given. Secondly, the frequency-dependent characteristics of wave impedance and the propagation function of the transmission line are analyzed. Next, refraction and reflection characteristics of fault initial voltage TW at the DC reactor are researched. Then, the frequency range selection method is proposed by comparing the wavelet transform modulus maxima (WTMM) features of fault initial reverse voltage TW in different scales. Finally, the effectiveness and applicability are validated by extensive simulations.

2. A Short-Term Wind Power Forecasting Method Based on High-Level Meteorological Data

   Wenwen Ma, Xu Cheng, Ji Wu, Hai Zhou, Rui Fan

   Abstract

   Compared with the surface meteorological data, the high-altitude meteorological data is less affected by the surface environment. A short-term power prediction method for wind power based on high-level meteorological data is proposed. Firstly, a new idea of feature selection is proposed, which is reduced dimensionality by recursive feature elimination method and combined with greedy search algorithm. Secondly, a random forest model combined with Savitzky-Golay filter is established to correct the wind speed at the top of the wind turbine, and then the wind speed is converted into predicted power. Finally, an experiment is carried out with the high-level meteorological data and surface data generated by NWP of a wind farm, and the prediction accuracy is calculated with the local power assessment standard. It is proved that compared with the surface meteorological data, the high-level meteorological data converted by the proposed method has a better prediction precision.

3. High-Resistance Fault Detection Method of DC Microgrid Based on High-Frequency Component Energy

   Liuming Jing, Yibo Wang, Tong Zhao, Jinghua Zhou, Chunxue Wen

   Abstract

   Direct Current (DC) microgrids have been widely used. It protection methods have been improved gradually, but when high-resistance faults occur, the countermeasures are still lacking. There is less information about high-impedance faults, which makes many protection methods unsuitable. This paper takes the high-frequency component of high-resistance faults as the research point, and proposes a method to decompose the transient zero-mode current through the Empirical Mode Decomposition (EMD) algorithm, and calculates the ratio of its high-frequency and residual energy, and compares it with the threshold. So as to carry out the method of high resistance fault identification. After inspection, the identification of high resistance can be realized, and the change law of high-frequency component and low-frequency component are found. Using Power Systems Computer Aided Design/Electromagnetic Transients including DC (PSCAD/EMTDC) as the platform, a simple two-port network is built to verify the proposed method. After that analyzed the shortcomings of the method, has looked forward to the future direction of improvement.

4. Analysis and Test of RSA Algorithm in Smart Electricity Meter Identity Authentication

   Zhiling Yin, Feng Xue, Piao Liu, Jinman Luo, Xianyu Zeng, Ruijue Feng, Xiaonan Zhao

   Abstract

   The wide application of smart electricity meters is conducive to fine management of electricity consumption and better experience of electricity consumption, but the possibility of users’ privacy being snooped is also gradually increasing. The data security of smart electricity meters has become a priority problem to be solved. Smart meter identity authentication protection is one of the key technologies of smart meter data protection. Identity authentication and digital signature technology based on RSA public key technology provide protection for smart meter identity authentication. Through the transplantation of RSA algorithm to Linux system and performance test, the test analysis shows that RSA algorithm has certain security for data addition. There is no way to be cracked in a short time, and signature verification is faster than signature.

5. Evaluation of Operation Effect for Intelligent Distribution Terminal Based on VIKOR Model

   Wenxuan Liu, Liu Han, Yi Liang

   Abstract

   In recent years, intelligent distribution terminals have been widely used in distribution network, and the number of terminal has increased significantly. However, the investment and benefits of distribution terminals have always been the focus of attention. Disorderly investment is not conducive to the development of the distribution network. A multi-objective optimization model for the benefit evaluation of the intelligent distribution terminal is established in this paper. The highest reliability of the distribution network, the minimum total investment of the intelligent distribution terminal and power loss of black out are regarded as the objective functions. Various constraints of the distribution network are considered at the same time. Then the genetic algorithm is applied to solve the optimization model. The final scheme is selected by combining the comprehensive weight and VIKOR method. This method takes the benefits of intelligent distribution terminals into account. It greatly improves the reliability of distribution network. Finally, the effectiveness of the proposed method is validated by the optimization results on regional distribution network in North China.

6. Techno-economic Feasibility Analysis of a Hybrid Energy System Under Different Load Forecasting Errors

   Hongxia Zhang, Chen Zhang, Zonglei Xu, Nian Peng, Xiongwen Zhang

   Abstract

   As the distributed power generation systems, hybrid energy systems (HESs) are widely used in building electrification construction and have become a research direction of the energy field. This study aims to investigate the influences of different load forecasting errors on system optimization and planning. A time series forecasting method based on long short-term memory is used to forecast the short-term load of the buildings, and the regional contraction algorithm is applied to optimize the component sizes. The feasibility of HESs under a certain load forecasting errors is analyzed from the aspects of operation and economy. The results show that the different load forecasting errors can cause considerably large change in grid capacity optimization results. Within the error accuracy range, the dispatch schedules are not affected by the errors. The battery should be reserved enough energy storage space to store or release energy that is not calculated in the estimation in the actual system.

7. Research on AC/DC Flexible Coordinated Control Method for Large Scale DC Infeed

   Changxiang Huang, Haijun Chang, Xuemao Zhao, Hongli Zhang, Chao Wang, Chutao He

   Abstract

   With the rapid development of long-distance and large capacity UHVDC, the coupling characteristics of AC and DC become closer and closer, especially the risk of simultaneous commutation failure of DC with same sending and receiving end is increasing, and the security and stability problems of power grid are becoming increasingly prominent. It is urgent to study the security and stability control technology suitable for AC/DC hybrid power grid. AC/DC flexible coordinated control technology is based on multi DC power emergency control to reduce the risk caused by the impact of DC fault with large transmission power, to improve the system oscillation damping level, to increase the transient voltage stability, and to replace or reduce the power grid load shedding and generator tripping after AC/DC fault. In view of the practical problems existing in the operation of multi-infeed of large-scale HVDC feed in receiving end, combined with the UHVDC projects to be put into operation in the next two years, and based on the theory of multi DC coordinated control and time-domain simulation technology, the effectiveness of AC/DC coordinated control technology in improving power transmission capacity and safe and stable operation of power grid is verified.

8. Development of Fault Diagnosis Software for the Center Station of High-Voltage Transmission Line

   Yingrui Wang, Lin An, Hualiang Zhou, Tonghua Wu, Guipeng Shen, Dan Rao

   Abstract

   The software of distributed fault diagnosis center station of high-voltage transmission line is a key link in fault diagnosis and location system of high-voltage transmission line. Based on the development and application of practical software system, this paper presents a software system implementation scheme of fault diagnosis center station. The paper first elaborates the overall scheme for system, analyzes the deployment ways of center station and its advantages and disadvantages, introduces the logical relationship between various software modules, expounds the front-end communication, key data processing technology and the range unit selection strategy for the central data station, and presents several key technologies such as three-point-method ranging algorithm independent of wave velocity, and identification of lightning strike and non-lightning strike, shielding failure lightning and back flashover lightning. The developed center station software stated in this paper has been operating stably and reliably in many places, with accurate fault location and accurate fault type judgment.

9. A Dual-Objective Optimization Control Strategy with Markov Forecast for Battery Energy Storage System to Mitigating Wind Farm Fluctuations

   Haotian Gu, Kai Hou, Yingwei Jiang, Hang Yin, Jing Liu, Xu Zhong

   Abstract

   In a renewable energy power system dominated by a high proportion of wind power, the configuration of energy storage system is an effective solution to mitigate wind power fluctuations, ensure the safe and stable operation of the grid, and improve the dispatch ability of wind power. This paper proposes a battery energy storage system (BESS) control strategy with the consideration of the impact of future wind power. To achieve maximum bidirectional output capability and minimum output of BESS, multi-objective hybrid optimization model is built on base of first-order low-pass filter’s variable time constant T_f, which contains two weight coefficients c₁ and c₂ to flexibly adjust BESS output. Markov chain prediction model is introduced to forecast the future multi-step wind power and PSO optimization algorithm is used to obtain the optimal BESS output power in real-time rolling calculation. Finally, the effectiveness of the proposed control strategy is verified through case studies.

10. Research on Comprehensive Evaluation Method of Transformer Health State Based on Collaborative Analysis of Multidimensional Indicators

    Lu Lu, Zhang Qingwei, Shan Changwang, Li Xiaomeng

    Abstract

    With the continuous development of smart grids, there are more and more transformer monitoring systems deployed, and the operation status of the transformer can be fully monitored. But there are new problems have emerged. Multi-dimensional data is difficult to correlate, and a single information source is easy to cause error judgment and warning delay. The ability of data extraction is insufficient, lacking data depth mining and advanced applications, cannot provide a substantial reference value for transformer fault prevention and maintenance. This paper presents a comprehensive evaluation method of transformer health based on multidimensional indicators collaborative analysis. Firstly, the large transformer monitoring database is preprocessed and screened. The multi-dimensional and multi-level transformer comprehensive evaluation index system is constructed. Then, combined with the historical data, find out the internal relationship between the parameters, and realize the multi-dimensional index collaborative analysis. Using the variable weight coefficient method, the corresponding multidimensional indicators deduction model and evaluation basis are established under different data types. Finally, the comprehensive evaluation system based on AHP and fuzzy comprehensive evaluation method is established, and the accurate scoring and grade judgment based on transformer data fusion is realized. This method can find the early defects in time, guide the operation and maintenance, reduce the time to deal with problems, and improve the safety and stability of transformer operation.

11. A Rapid Method to Configure Inspection Object for Substation Robot Based on Visual Equipment Detection and PTZ Control Model

    Qingwei Zhang, Han Sun, Jing Wang

    Abstract

    The configuration for substation robot inspection task includes the following steps: configuring inspection object, scanning map package, building wireless AP, establishing database, etc. Previous work mostly relies on the experience of staff to configure inspection object, which is a waste of manpower and material resources. Besides, configuration for each inspection object of substation robot manually is a time-consuming task, which leads to a decrease in the effectiveness and accuracy of robot because of artificial error. A rapid configuration method to configure inspection object for substation robot inspection task is proposed based on visual equipment detection and PTZ control model. It aims to grab the inspection object automatically without manual intervention. The efficacy of the proposed method is verified through experimental results of substation.

12. Searching Loops in Distribution Systems: A MILP-Based Method

    Chao Qin, Yongkang Zeng, Jiancun Liu

    Abstract

    In existing studies, additional integer variables and complex constraints are usually introduced to ensure the radial topology in distribution system planning and operation optimization. A straightforward method to deal with the radiality constraint is to find all loops in a distribution system and assure that for each loop, at least one branch is open. However, there is still a lack of effective methods to find all loops in a distribution system. Based on graph theory, a model is presented to check whether an edge-induced subgraph of a graph is a loop or loop group and a method to search all loops in a distribution system is further proposed. All loops in a distribution system are searched out in the order of the number of edges included in loops from small to large. The proposed method is formulated as a mixed integer linear programming (MILP) model. The simulation results validate the feasibility and effectiveness of the proposed method.

13. Design of Integrated Energy Market Service System Based on Blockchain Smart Contract

    Wang Lei, Ma Yichao, Zhu Liuzhu, Wang Xuli, Cong Hao, Shi Tiancheng

    Abstract

    There are some problems in the traditional integrated energy market, such as single transaction mode, difficulty in effective dispatch, and lack of targeted subject to provide integrated energy market services. Therefore, this paper designs an integrated energy market service system based on the blockchain smart contract. The participants in the integrated energy market are divided into integrated energy suppliers, integrated energy users and integrated energy service providers. The integrated energy service providers set up energy trading centres and energy dispatch centres according to the needs of integrated energy suppliers and integrated energy users, and provide energy trading and dispatch services for them by using the Digital Currency Electronic Payment (DCEP) of the central bank and the blockchain smart contract. “Antchain” platform is used to simulate the energy trading and dispatch process, and the verifiable simulation results are given. It is proved that the integrated energy market service system based on the blockchain intelligent contract designed in this paper can effectively realize the security and convenience of energy trading and the efficiency and intelligence of energy dispatch.

14. An Improved Single-Phase Fault Line Selection Method for Non-effectively Grounded System Based on Transient Energy

    Lei Xia, Guo Hu, Hai Wu, Song Hu

    Abstract

    The domestic distribution network is mainly non-effectively grounded system, single-phase grounding fault accounts for more than 80% of the total number of faults, the fault characteristics are not obvious and easy to be affected by noise signals, at the same time, the polarity of zero sequence voltage and current transformer is unknown or reverse connection often occurs, which makes the local fault identification and line selection more difficult, and directly affects the fault response speed and disposal efficiency. In this paper, an improved single-phase to ground fault line selection method based on transient energy is proposed. By using the energy matching degree of transient zero mode current in each interval and the polarity relationship between transient zero mode current and steady-state zero sequence current in this interval, the existing fault line selection methods are easy to be affected by the polarity of external mutual inductors, and the polarity of mutual inductors can’t be checked online in the fault process. The physical concept of the method is clear, and the criterion threshold is not affected by the grounding mode and fault location of the system. RTDS simulation experiment and field engineering application verify the effectiveness and universality of the proposed fault line selection method.

15. Simulation and Verification of Sub-synchronous Oscillation from Impedance Perspective-Taking a Wind Power Base in China as an Example

    Guang Xu, Ling Zhu, Jianhua Li, Zhiguang Huang, Zhaowei Li, Yuqiang Hou

    Abstract

    Recently, numerous sub-synchronous oscillations (SSO) events have occurred in many wind farms in China, which seriously threatened the power system security and stability. It is generally believed that the LC resonance between the wind farm side and the grid side with series compensation (SC) devices is one of the conditions that cause the SSO. This paper takes a real power grid as an example, reappeared the SSO events through electromagnetic simulation, analyzed the key influence factors, and proposes a strategy from impedance perspective in the end based on reasonable assumptions, in order to suppress the SSO shortly after the oscillations happened. Firstly, make off-line impedance analysis based on the deterministic grid structure, and then give the sensitivity results of each wind farm to the SSO. Finally, cut off the wind farm based on the sensitivity orders during the SSO. The simulation results were performed to verify the validity.

16. Dual-Objective Optimization Scheduling Strategy Considering V2G Compensating Battery Capacity Under Photovoltaic Microgrid

    Hui Wang, Shuaishuai Lv, Xiu Ji, Xiangping Meng, Chunsong Wang

    Abstract

    In view of the large fluctuations in the output of photovoltaic microgrids, large energy storage capacity is required to solve the problem of stabilizing the load. In order to reduce energy storage investment costs, considering the energy storage characteristics of V2G, the range of decision variables, the constraints of power balance requirements and energy exchange strategies are given. Through the establishment of a coordinated scheduling model of battery and V2G in the photovoltaic microgrid with the dual goals of economy and grid-connected load fluctuation rate. There are four load modes of electric vehicles, namely none, disordered, transferring and V2G respectively. After realizing the optimal scheduling by using particle swarm optimization algorithm, the economic and safety benefits of power grid, micro-grid dispatching center and electric vehicle users are compared in the sub-mode. Finally, according to the analysis of the calculation examples, the two-level load curve and the economic income table under the four modes are solved. The comparative analysis shows that the introduction of V2G can replace some capacity batteries, so that the photovoltaic microgrid can be further optimized in terms of load peaks and valleys, tripartite economy and safety.

17. Research and Engineering Practice on Reactive Power Control Based on Deep Reinforcement Learning in Local Power Grid

    Minrui Jiang, Dawei Xu, Tianhua Chen, Huijun Li, Xiaochun Xu, Shuai Wang

    Abstract

    Due to the increasing scale of power grid with a large number of renewable energy stations connected to local power grid, the dynamic characteristics of power grid have been constantly enhanced and the structure of power grid has become more and more complex, which has bought huge challenges to reactive power optimization of the power system. However, the rapid development of artificial intelligence (AI) technology provides a new way to realize the reactive power optimization decision-making of high-dimensional, highly nonlinear, and highly time-varying power grid. Based on the deep reinforcement learning (DRL) technology in AI, this paper proposes a grid reactive power optimization control and decision method with online learning function. Based on Markov Decision Process (MDP), this method adopts deep solution strategy of reinforcement learning and continuously accumulates experience based on the historical data of the power grid within 3 years, which finally can generate a control strategy that meets the reactive power optimization needs within seconds. In order to verify the feasibility of the method, the method proposed in this paper has been tested in a closed-loop control on the actual local power grid. The results of engineering practice show that the reactive power control system based on deep reinforcement learning can quickly and accurately respond to grid reactive power optimization requirements, generate correct and usable control strategies, and have obvious effects on reactive power optimization, and thus proving the reliability and effectiveness of the method.

18. Study on Block Strategy Optimization of Passive Isolated Island Operation of DXB UHVDC Project

    Run Huang, Lifei Zhang, Chen Wu, Chengxiang Li, Lei Yang, Mingkang Wu

    Abstract

    Due to the lack of generators at sending end in the early operation stage of the DXB UHVDC project, if it passively entered into island operation mode, the system will encounter harmonic instability and will cause a great risk to the steady operation. In previous projects, the common solution is to disconnect AC filters and block bipolar DC with a certain time delay when receiving island operation signal. However HVDC system may block valves, cut generation capacity or shed power by error when receiving false island operation signal. Therefore, the blocking strategy at low power under passive island mode has been improved. In DXB UHVDC project, the power on AC tie lines is used as criteria for island operation to prevent false action. The logic and realization of the control strategy are illustrated, then RTDS simulations are carried out and it is verified that by judging the power on AC tie lines, the system can effectively prevent from entering island mode by error. Finally, the field test proved that the proposed improved strategy will not cause rejecting of the existing “island blocking logic”. The proposed strategy can be a reference to other DC projects.