Abstract:The application of multi-core processors in the embedded power terminal system can significantly improve the system performance, whereas will also bring about the increasing of the reaction time and the energy consumption. Aiming at the operating characteristics of the power industrial control terminal system, this paper proposes a time-trigger-based energy-saving scheduling method for multi-core systems. An intelligent scheduling scheme is established to ensure the requirements of real-time performance for multi-core systems. Also, two common system-level energy management technologies, such as the dynamic voltage and frequency regulation (DVFS) and dynamic power management (DPM), are adopted to reduce the power consumption of multi-core processors.

Abstract:The wide application of multi-threading technology in power terminal multi-core chips has significantly increased the system communication time of the power terminal system. The number of threads in power applications has also increased, and the communication between threads has become more frequent. Aiming at the problem of system communication time overhead and inter-thread overhead, this article first introduces communication pipeline technology. The communication pipeline technology allows communication and calculations in the same thread to work at the same time, while also reducing communication transmission time; In addition, this article also introduces message aggregation technology, the communication channels are gathered to increase the amount of data transmission per unit time and reduce the number of communications. In order to reduce the number of thread switching and reduce system synchronization time, the communication queue technology is also introduced; however, the dependency between threads will affect the use of communication queue technology. Finally, an optimization method is proposed for the dependency loop, which can effective solve the limitations caused by the dependency ring, and improve the utilization of communication queues and the efficiency of communication between threads.

Abstract:The protection installation mode and detection requirements in intelligent substation is enhanced, and the low reliability problem of previous protection system due to the slow transmission speed and the overly complex architecture need to be handled. In this paper, a k-means algorithm-based priority management method is proposed to divide and exchange the data, and to ensure the real-time performance of the data. The FPGA is used as the parallel communication coprocessor to enhance the transmission efficiency to the processor. Thus, a SoC chip protection device is developed. The operation time of the protection device is determined by the fault oscillogram of simulation by the RTDS, which can prove the reliability and practicability of the devices. Finally, the protection device is applied in some substations of Southern Power Grid for three years. The results show that the operation of the device is reliable and stable, the protection can therefore be localized, miniaturized, and the protection and power consumption can also be ameliorated.

Abstract:Improving the working efficiency of power terminal chips while reducing their energy consumption is one of the research direction for optimizing smart grid systems. Aiming at the efficient management of cache data in MPSoC, the multi-processor shared cache partitioning technology is studied. The curve fitting technology is utilized to model the cache, and mathematical methods is incorporated to solve the CP problem. The mathematical expression of the energy consumption in subsystem can be obtained according to the mathematical relationship between the obtained missing rate curve of the shared cache and the energy consumption of the subsystem. Combined with the energy consumption model of the processor, the comprehensive optimal CP solution is generated. Experimental verification shows that the processor subsystem energy consumption can be reduced to 27.9% of the subsystem before optimization using this CP method.

Abstract:Due to its own vulnerabilities and the vulnerability of the communication network, the intelligent distribution transformer terminal deployed for the station area is vulnerable to network attacks. For solving the security problems existing in the intelligent distribution transformer terminal, this paper proposes an intelligent distribution transformer terminal network security situation awareness method based on RBF-SVM. Firstly, the potential network attack that the terminal may suffer is analyzed, the corresponding security detection indicators areextracted and normalized. Then, a nonlinear support vector machine (SVM) classifier based on the Gaussian (RBF) kernel function is conducted. The k-fold crossvalidation and grid search method is applied for determining the optimal parameters of C and g for the classifier, and the Security Situation Awareness model of the intelligent distribution transformer terminal is established. Finally, the test index data are substituted into the model for training and testing. The results show that compared with s random forest and logistic regression methods, the proposed method has a higher accuracy rate, can realize terminal security situation awareness, and canbe used for practical power terminal security protection.

Abstract:In order to solve the poor economy, long power supply radius, and low voltage quality problems in the operation of the distribution network in sparse and remote areas, a compensation method of active/reactive power and voltage based on power electronic voltage regulator is adopted to achieve the extension of power supply radius and adjust the voltage accurately and quickly. Then, in order to reasonably configure the role of power electronic voltage regulator in the distribution network, a bi-level optimization model for its location and capacity is designed. The upper layer of the optimization selects the best installation positions by considering the power flow sensitivity factor as the main factor and the voltage stability index as a supplement. The lower layer considers the voltage stability and economy to determine the optimal installation capacity. Finally, an example is given based on the improved IEEE33 node of the distribution network in sparse and remote areas. The optimal value of the model is solved by utilizing the dynamic inertia weight particle swarm optimization algorithm. It is proven that the feasibility and effectiveness of the configuration method.

Abstract:AC/AC converter, as the key part of power conversion, plays an important role in the unique power quality conditioner, electric locomotive traction technology and frequency variation power supply. The modular multilevel AC/AC converter (AC/AC-MMC) is a promising topology for direct AC/AC converting applications. Under the background, according to the relationship between the arm electrical quantities and the input and output of the single AC/AC-MMC, the equivalent current decomposition model of decoupled input circuit, output circuit and circulating current circuit is derived in this paper. The independent control of the three circuits is realized successfully. Then, the performance indexes evaluation function of the current model and capacitor voltage balance is established. And the model predictive control (MPC) method based on arm level traversing is proposed and utilized to realize the accurate current tracking of input current, output current and circulating current as well as the balance control of submodule (SM) capacitor voltages. The method selects the optimal arm level with the goal of the optimal performance indexes in the evaluation function. Based on the proposed MPC method, the optimized method of SM internal loss equalization control is proposed by equally distributing the switching states of the left and right arm of each SM, realizing the balanced distribution of SM internal loss, and improving the operation reliability. Finally, the PSIM simulation model is established to validate the feasibility and effectiveness of the proposed control method.

Abstract:With theintegration of large-scale renewable energythe problem of the reduction of the regional voltage qualification rate becomes more significant than ever before because of the stability issues and the poor control performance of the renewable energies.In order to improve the accuracy and stability of the regional voltage, this paper investigates the reactive power limit of DFIGsby detailed calculation.The overcurrent factor is taken into account to further correct the reactive power lower limit of DFIGs.Thenthe control structure of DFIGs is improved by considering the reactive power capacity of GSCs in the DFIG reactive power compensation. Hence the reactive power compensation capability of the wind farm can be deeply exploited and the maximum active power would not be affected. In addition, a coordinated control strategy of wind farms and SVG is proposed, which can not only retain the flexible compensation capability of SVG, but also enable the wind farm to provide reactive power support and share the burden of SVG compensation.Finally, by analyzing parameters such as sensitivity and reactive power remaining, this paper designs a regional reactive power allocation strategy, which can further improve the voltage control performance. The control performance is verifiedthrough a two-area power systems with the integration of wind farms.

Abstract:Fast and random charging of electric vehicles (EVs) in the charging station can easily cause the overload and uncertain peak load of the grid. In order to improve the stability of the grid and reduce the power fluctuation caused by random charging, a charging load discrete model for energy storage configuration is proposed. The time-continuous charging load curve is discretized into charging sequences at equal intervals, the power supply cap of the grid and energy storage capacity are quantified to the number of available charging piles, and the load status is described by applying queuing theory within the set time. Then the system load state transition relation is established by applying two-dimensional discrete time Markov chain, thus a random charging model of a single energy storage fast charging station is established. According to the state space and probability distribution of the discretized load, the economics under different load statuses are analyzed. Finally, through the case study, the impact of the EV arrival rate on the service quality and economy is analyzed under different ratios of grid power and energy storage power, so that the feasibility of the model is verified.

Abstract:The uncertainty and randomness of the charging behavior of electric vehicles make a large number of charging loads connect to the grid in a short period of time, which will lead to large load fluctuations. At the same time, the disorderly charging behavior of electric vehicles can not guarantee the interests of charging users under the condition of time-of-use electricity prices. In order to alleviate the negative impact of these problems, the charging behavior of electric vehicles is firstly analyzed based on the Markov Decision Process (MDP) in the reinforcement learning. Secondly, an incentive function is constructed to guide the electric vehicle to make charging choice according to the power supply margin of the grid. Then an orderly charging strategy that meets the minimum load fluctuation and the minimum user cost at the same time is produced. Finally, the Monte Carlo method is utilized to simulate the charging status of electric vehicles. Results of orderly charging simulation show that the strategy can effectively improve the load superposition curve, play the role of peak shaving and valley filling and reduce the user charging cost.

Abstract:Due to the multi-agent problem in the investment construction of a newly added distribution network and the unclear cooperation mode between the agents, the paper studies the cooperation mode between the incremental distribution network and the power supply company. Firstly, the paper constructs an evaluation index system based on saliency-redundancy analysis. Then, in order to ensure the weighting accuracy, an objective weighting method and a subjective method are utilized to construct an index weighting model based on the analytic hierarchy process method modified by the entropy weighting method. After that, on the basis of the changing speed state model and the changing speed trend prediction model, a dynamic dynamic-comprehensive fuzzy evaluation model for the Cooperation Mode of the incremental distribution network is constructed. Finally, four scenarios are set up:absolute control of power supply company, relative control of power supply company, the shareholding of power supply company, and non-shareholding of power supply company to perform example analysis. The calculation examples show that the dynamic-comprehensive evaluation model can reflect the actual level of effectiveness better than the static evaluation model.

Abstract:In order to improve the rationality and economy of distribution network maintenance plan, an optimization method of distribution network maintenance plan based on improved particle swarm optimization is proposed. Firstly, with the goal of minimizing maintenance cost, power supply loss, and failure loss, and the constraints of maintenance resource limitations, maintenance sequence, and safe and stable operation in the actual maintenance process, a maintenance plan optimization model that conforms to the actual distribution network maintenance process is established. Secondly, corresponding preprocessing methods are proposed for different types of constraintsto reduce the complexity of solving the problem. Finally, by incorporating the idea of natural selection into the iterative updating of population particles, the overall quality of population particles is improved to overcome the problems of premature convergence and easy to fall into local optimal solution of standard particle swarm optimization algorithm. The improved particle swarm optimization (PSO) algorithm is applied to solve a specific example. Results show that the proposed model and algorithm have good feasibility and rationality.

Abstract:The reactive power compensation can optimize the voltage distribution in AC-DC hybrid distribution network, reduce the network loss, and increase the safety margin of the network then improve the power supply reliability. In order to realize the directional compensation of the reactive power in AC-DC hybrid distribution network and describe the weak point of the voltage in the system directly, a reactive power compensation method based on sensitivity analysis is proposed. The alternating iteration method is adopted, and the power flow of AC-DC hybrid distribution network is calculated to obtain the Jacobian matrix in which the power flow has converged. Then, the sensitivity of node voltage change on reactive power injection is calculated, and the point with the maximum sensitivity is selected as the reactive power compensation access point. In addition, in the aim of the low system power factor, the reactive power compensation capacity of access point is determined. An example analysis conducted on 11-node system shows that this method takes less time to calculate and has better compensation effect compared with the harmony algorithm. The analysis results show that the proposed method can raise the system voltage level with the minimum compensation capacity at the greatest extent and effectively reduce the system network loss.

Abstract:In order to improve the economical efficiency of micro-grid, it is necessary to conduct the day-ahead optimal scheduling. In this paper, a micro-grid optimal scheduling model is established for minimizing the operation cost of the micro-grid. The optimization variables of this model includes the battery charge/discharge power and fuel cells output power. Based on the day-ahead predicted data, one day is divided into 24 hours and the the scheduling period is set as 1 h. Then, taking 15 minutes as a period and the day is divided into 96 scheduling periods. The results of the one hour optimal scheduling are taken as the input of the 15 minutes optimal scheduling for re-optimization. The several heuristic algorithms are used to solve the models, respectively. The results show that the improved black hole algorithm converges faster than the others. The optimal results can be utilized as a reference for the battery charge/discharge power and the fuel cells output power for the next day. This method can effectively reduce the operating costs of micro-grid.

Abstract:Considering the influence of heterogeneous characteristics on the demand side load resource regulation, a joint control strategy based on generalization index is proposed for heterogeneous thermostatically controlled loads (TCLs). The error correction and the behavior analysis is taken into account firstly, and the degree of refinement and applicability of the equivalent thermoelectric model for TCLs is further improved. Consequently, a generalization index is proposed with the consideration of the controllable energy space and the transferable time, which realizes the standard evaluation of the operating state of heterogeneous TCLs, and can be further utilized to guide the combinatorial optimization of heterogeneous load groups. Finally, the experimental simulation indicates that the proposed method can perfectly match the demand for the control of heterogeneous TCLs, and can maintain the control accuracy and improve the temperature distribution at the same time.

Abstract:With the extensive application of information and communication technology in power system, power system relies more heavily on the stable operation of communication system. When a communication node is attacked and fails, the propagation of communication node failures across the powercommunication network will eventually lead to a cascading failure of the power information-physical system. Therefore, it is the necessary to to establish practical interdependent networks model, evaluating the importance of the communication network nodes and protecting the key nodes. Firstly, a weighted network model considering the reactance of the power line and the used rate of the information link is established for the unilaterlal networks in the interdependent networks. Then, the importance of two unilateral network nodes is evaluated according to the weighted network node importance evaluation method. Finally, considering the importance of communication nodes, the importance of the inter-network nodes is evaluated based on the internetwork coupling incidence matrix. An IEEE14 bus system is taken as an example to construct a dependent network and evaluate the importance of the communication nodes. The results show that the proposed method r is feasible and can be utilized in practical engineering.

Abstract:With the eventually improvement of power consumption information collection, the accurate analysis for user power consumption mode will provide an important basis for power intelligent construction. When analyzing power consumption modes, the load is the only clustering feature to be taken into account. Therefore, a weighted clustering analysis method considering multi-type feature selection is proposed. Firstly, the load and meteorological features are normalized to establish a feature set to be selected. Then, the clustering feature set is selected by combining mutual information and grey correlation degree. Finally, the weighted k-means algorithm is utilized to cluster the selected feature sets, and the typical behavior of each power consumption mode is analyzed with the load curve. Through the analysis of the electrical load data of a shopping mall in Shanghai, it is proved that this method can reduce the interference of redundancy information and improve the clustering quality.

Abstract:The integration of operation and distribution has realized the effective concentration of multi-level work, such as the marketing, production and dispatching. However, the risk level of the power supply system also increases. Under the background, a new risk evaluation model of operation and deployment is proposed in order to enhance the risk early warning capability of the system. Firstly, a power supply risk evaluation index system is constructed based on the characteristics of the integrated system. Secondly, an index weighting method based on AHP and DEA is proposed to reduce the error caused by subjective factors in the weighting process of traditional matter-element extension model. Then, a comprehensive risk assessment model with integrated operation and distribution is established by constructing the comprehensive risk element for power supply and demarcating a power supply risk rating. Finally, a typical area is taken as an example to analyze its risk level and influencing factors, and the effectiveness of the proposed model is verified.

Abstract:Constructing a multi-energy supplement system that combines heat, gas and other forms of energy, with electric power as its main energy, and has complementary advantages in multiple parts of production and consumption, is the key for improving energy efficiency, promoting clean energy development, and building energy Internet. Firstly, according to the distributed energy (such as wind, light, etc.) stability model, the smart city multi-energy supplement system architecture is constructed. Secondly, by comprehensively taking the investors, power companies, electricity customers and other energy interest bodies into consideration, the smart city multi-energy supplement system planning model on the conditions of power flow, node voltage with complementary energy, and installed capacity, is built. Thirdly, the improved genetic algorithm is introduced to determine the optimization flow of the model, and the node location and capacity of the energy network of multi-energy supplement system are optimized. The results show that the proper planning of multi-energy supplement system is helpful to the development of energy supply in smart city to be cleaner and more sustainable.

Abstract:The scale of the power grid is expanding day by day, and the short-circuit short circuit level of the power grid is continuously rising. Installing the current limiter is an economically feasible current limiting measure. Under the background, this paper proposes a new fault current limiter location optimization and capacity planning method. Based on the distribution characteristics of the grid nodes and branches, the vector of short-circuit current level and breaking margin and the vector of the cost and power loss of the whole network are defined. By studying the normative space characteristics of the defined vectors, a class of hyperspace norms with specific physical meanings is defined as the measure of vector distance, the cost and gain indicators of the fault-limiting device configuration scheme of the whole network are derived, and the objective function of the current limiter location and capacity configuration is constructed. Considering the safe operation constraint of power grid, the paper uses a genetic algorithm with variable chromosome length to solve the problem. The example analysis is based on the IEEE 14 bus standard test system to verify the effectiveness and practicability of the proposed method.

Abstract:Aiming at the problem of transformer DC bias caused by the urban rail transit, this paper proposes a DC bias state identification method based on vibration signal for transformers of urban power grid. Firstly, the time of duration and the frequency characteristics of the transformer vibration signal under the conditions of DC bias, short fault and harmonic interference is analyzed. It is found that compared with other faults, the transformer vibration intensifies under DC magnetic bias. The frequency component of vibration signal becomes complex. A series of high-order harmonic components, especially the 50Hz odd double frequency component, increase significantly. Based on these phenomena, the influence of short circuit fault on DC bias state identification is eliminated by using the sum of energy of 50 Hz octave component of vibration signal except 100 Hz. Then, the ratio of the sum of the energy of 50 Hz odd octave components except 100 Hz over the sum of energy of 50 Hz octave components except 100 Hz is used to eliminate the influence of power grid harmonic interference on DC magnetic bias state identification, so as to realize the state identification of transformer DC magnetic bias caused by urban rail stray current. Finally, the on-site measured data is analyzed and processed to further verify the accuracy of proposed method.

Abstract:The line loss rate is an important basis to reflect line loss management. Due to the complexity of its theoretical calculation, it has been widely concerned by power workers. Based on the current research status of line loss management at home and abroad and related theoretical calculation methods, a multiple linear regression model based on the K-Means clustering algorithm is proposed to predict the line loss rate of the station area. Firstly, the proposed K-Means clustering algorithm is utilized to cluster and analyze the sample data of the station area. Linear regression prediction models are established to calculate the line loss rate of the station area according to the clustering results. Then, through the comparison and analysis of the predicted line loss rate and the actual line loss rate, much attention is paid on the stations with large errors in line loss estimation. Finally, Finally, based on the sample data of some regions in Guizhou, the accuracy and rapidity of the proposed method are verified, which provides a theoretical basis for line loss management in Guizhou.

Abstract:The data quality defects in the emergency repair service platform of intelligent distribution network have negative impacts on the accuracy of early warning and emergency repair. Under the background, a data quality processing method for power-off information of power consumption information acquisition system is proposed. Firstly, the regression method is utilized to identify the overall anomaly rate of the data. Secondly, the defects of the shutdown and power on data are identified and processed according to the identification indicators of integrity, uniqueness, consistency and accuracy. Finally, the significant relationship between the latest data and historical data is analyzed to help identify data timeliness. This method identifies and processes the quality of source data, and provides reliable data support for the development and operation of information integration and fault diagnosis in the later platform construction.

Abstract:For the purpose of solving the endurance problem in the process of unmanned aerial vehicle(UAV) line patrol, an autonomous line patrol system of UAV based on leapfrog charging is designed and implemented. As the first part, the autonomous positioning of UAV based on Global Positioning System/Real-time Kinematic (GPS/RTK) is studied in this paper. Firstly, aiming at the goal of UAV autonomous line patrol proposed by China Southern Power Grid Corporation, a leapfrog charging platform is built to improve the endurance of UAV. Then, in order to ensure that the UAV can conduct autonomous navigation according to the preset track and accurately land on the ground platform for charging, the key technologies of autonomous line patrol in the UAV GPS / RTK integrated navigation module are studied. Finally, the three main links of attitude calculation, acceleration compensation and delay compensation in line patrol technology are experimentally simulated. Experimental results show the attitude calculation algorithm can accurately estimate the flight attitude of UAV in flight and can quickly compensate the motion acceleration in time. The delay compensation algorithm can effectively solve the problem of positioning sensor delay. When the barometer output altitude data is unhealthy, the system can identify the health status of the data and automatically fuse the GPS output altitude data. Therefore, the methods and research conclusions proposed in this paper have important engineering guiding significance for the autonomous line patrol of UAV.

Abstract:The convolutional neural network algorithm is widely applied in the automatic inspection of transmission lines. However, the generalization ability of traditional convolutional neural network power defect-recognition model is not ideal. Under the background, this paper proposes a cross-view relation region convolutional neural network (CVR-RCNN) detection algorithm that integrates dual-angle image information, which utilizes two-view visible light images to identify typical defects in transmission lines. The testing shows that the CVR-RCNN model has good robustness. The area under curve (AUC) value of the receiver operating characteristic (ROC) curve is as high as 0.927, and the defect detection accuracy is significantly improved compared with traditional algorithms. Therefore, CVR-RCNN can significantly improve power defect detection and improve the accuracy and stability of the algorithm architecture for the automatic inspection of transmission lines by UAVs.