Abstract:At present, voltage sag has become one of the most prominent power quality problems.In order to effectively analyze the impact of lightning stroke on power grid sag, the situation of voltage sag caused by lightning stroke is analyzed in detail.Four types of temporary relief, including lightning induced temporary relief, are identified accurately, which provides an important basis for the rational division of temporary relief liability.Firstly, the difference between the effective waveform of temporary sag caused by lightning stroke fault and the common short circuit fault is analyzed.The characteristics of RMS waveforms of four kinds of sag types, short circuit fault, lightning stroke, transformer switching and induction motor starting, are summarized.Five characteristic indices of sag voltage are introduced and the characteristic matrix of sag type identification is established.Then four types of sags are identified by using decision tree support vector machine (PDT-SVM) classifier based on particle swarm optimization.The training and testing data of the classifier come from the measured sag voltage data of the power grid, which is closely in line with the engineering practice.Finally, the validity and accuracy of the algorithm are verified by the analysis results of an example.

Abstract:With the high penetration of distributed generators, the radial structure of conventional distribution network system will change to a multi-terminal type, the traditional fault location method will be invalid. In this paper, a fault location method based on deep learning for active distribution network is proposed. This method firstly collects the current and voltage data through the feeder terminal unit. Combining the power output data, a fault data vector is formed; secondly, it uses tensorflow framework to build a deep neural network model based on fully connected network to mine the mapping relations between fault data vectors and fault sections and form the final fault location model through training. Finally, the fault location results demonstrate the effectiveness of the proposed method. Case studies show that compared with the traditional BP and learning vector quantification neural network model, the deep learning model has faster convergence speed and higher fault location accuracy. The final model has high fault tolerance to information distortion and loss.

Abstract:Substation equipment inspection plays a vital role in the reliability of power supply. The traditional manual inspection method is inefficient and heavy workload, and it is difficult to achieve ideal results. Robot inspection has become the future trend of smart grid development. The real-time localization and map creation of substation inspection robot are studied. A SLAM method of inspection robot based on adaptive Monte Carlo algorithm is proposed. Considering the dynamic environment of substation and the complexity of inspection robot′s characteristics, the two-dimensional map creation and real-time localization of robot are realized. The simulation results show that the SLAM positioning accuracy of inspection robot based on AMC algorithm is higher, and the anti-jamming ability for different process noise is stronger. Therefore, it is more suitable for SLAM problem solving and practical application of substation inspection robot.

Abstract:In order to improve the prediction accuracy of wind farm output power, wavelet analysis(WD) and cuckoo optimization support vector machine(CS-SVM) algorithm are used to predict wind power in ultra-short term, which is more direct and accurate than indirect wind power obtained by predicting wind speed. Firstly, the wind power model is decomposed into approximate sequence and detail sequence by using wavelet decomposition and reconstruction. Then, the support vector machine optimized by cuckoo algorithm is used to predict each sequence, and the prediction results of each sequence are obtained. Finally, the prediction results of each sequence are superimposed to form the final prediction value of wind power. The results of numerical examples show that the prediction results have high accuracy, and the method used in this paper is more accurate, superior and practical than the support vector machine and other methods optimized support vector machine prediction results.

Abstract:In view of the problems existing in the process of insulator operation and maintenance, such as too complicated regulations and too dependent on manual identification of operators, this paper presents an insulator condition evaluation method, which uses historical insulator defect images as training samples and realizes the basis of excellent performance of small sample data processing through migration learning. Based on the training of defect recognition model of deep convolution neural network and the feature extraction ability of convolution neural network, the quantification score of insulator defect can be achieved, and the comprehensive state evaluation of insulator can be realized by considering the operation life and external environment with the help of historical samples and expert experience. An example shows that the recognition accuracy of the proposed transfer learning model can reach more than 90% after training, while the recognition accuracy of the new learning model is only 70% under the same sample conditions, and the evaluation model established in this paper can more sensitively reflect the defect status of insulators in daily operation and maintenance. It shows that the evaluation method established in this paper is quite reliable and can provide experience for the daily maintenance arrangement of operation and maintenance personnel.

Abstract:Transformer oil is one of the main insulating material in power transformers. The density index of oil is closely related to the safe operation of transformers. On the basis of the principle of multi-frequency ultrasound, genetic algorithm GA and back propagation neural network BPNN, a prediction study of density of transformer oil is proposed. Taking 110 sets of transformer oil belonged to China southern power grid as an example, 100 of which are training sets and 10 are forecast sets, a prediction model of density of transformer oil is established based on BPNN, with the 242 dimensional multi-frequency ultrasonic data of oil sample as the input and density as the output, through the experimental method to determine the BP neural network hidden layer neurons number, the nonlinear mapping relationship, and use genetic algorithm GA to optimize the BP neural network connection weights and threshold of every layer. By adjusting the number of hidden layer neurons, the network is trained. Moreover, the genetic algorithm GA is introduced to optimize the network parameters. All results show that compared with the traditional standard BPNN model, the output value of density of transformer oil with the GA-BPNN model is much close to the real value with small errors, which lays a solid foundation to test transformer oil other parameters with tell multi-frequency ultrasonic technology.

Abstract:Partial discharge is the main form of early insulation failure of electrical equipment. Pattern recognition of discharge type is of great significance for the estimation of equipment insulation performance. Considering that the extreme learning machine (ELM) method has the advantages of simple structure and fast training speed, yet the initial parameter selection is random and the algorithm is unstable. A pattern recognition method based on fusion ELM algorithm for partial discharge is proposed. Considering the different judgement precisions based on variable features, the adaptive weight assignment is used to achieve the decision-level fusion of the output. In this paper, four physic discharge models are designed to simulate typical partial discharge defects. Discharge signal waveform and phase-amplitude spectrum is collected by high-frequency current transformer method, sufficient samples of experiment data are obtained to extract time-frequency domain and statistical features for classification. The result shows that the fusion ELM algorithm is superior to the traditional ELM algorithm and BP neural network in the recognition accuracy and stability without sacrificing training speed.

Abstract:Accurate prediction of sensitive power customer groups can perceive customer demand and improve customer satisfaction with electricity consumption and the level of power service effectively. A power customer outage sensitivity prediction model based on Bayesian network is proposed to predict the power customer outage complaints, which defines customer power outage sensitivity data labels by customer basic information, power consumption information, smart meter energy measurement information, and user power interaction behavior, coming from 95598 customer service platform, marketing system and power information collection system.It experimentally verifies the power outage sensitivity analysis model using K-folding cross validation method, shows that the power outage sensitivity prediction model based on bayesian network has high precision in the application of power outage complaint analysis, and the experimental results demonstrate the effectiveness of the prediction model.

Abstract:As the transmission voltage level increases, the tower is also getting higher and higher, and its spatial structure changes from bottom to top greatly. The modeling of the Ultra-High Voltage(UHV)tower plays an important role in lightning protection analysis of transmission lines. If the UHV tower still adopts the multi-surge impedance model, the effects of changes in the spatial structure of its high altitude cannot be reflected. Therefore, the non-uniform transmission line model is developed based on the physical structure of the tower. To study the influence of the tower′s spatial structure changes on its electromagnetic transient characteristics, the nominal height is modeled by several surge impendences. Finally, the lightning electromagnetic transient responses of the tower with the non-uniform transmission line model and with the multi-segment multi-surge impedance model are compared and analyzed. It shows that compared with the multi-surge impendence model, the non-uniform transmission line model is more suitable for lightning transient analysis of UHV tower.

Abstract:In order to deal with the new features of complex AC/DC power grids, China is developing the construction for system protection. The broad control resources, complex control strategy and the great action response require real-time dispatching operation to fully grasp the operating state of the system and dynamically optimize the action strategy. This paper developed the overall frame of online monitoring and strategy optimization for large power grid system protection at first, including state monitoring, modeling and simulation, real-time analysis, strategy optimization, etc. Then, this paper introduced key technologies of the functions mentioned above. The specific method for strategy optimization is mainly discussed. The proposed online monitoring technology is applied to the frequency coordinating control system in East China Power Grid and the coordinating control system for the receiving end of Tianzhong UHVDC project in Central China Power Grid, which verified the effectiveness of the proposed method.

Abstract:Based on the active charging response of electric vehicles, this paper proposes a collaborative planning model of “electric vehicle charging station-distribution network”. As a typical DC load, the electric vehicle has the characteristics of active response which can be used as a planning consideration factor to adjust the electricity price of the electric vehicle in a time-sharing manner. Guiding the user to actively respond to the change of the charging price, and shifting charging demand from peak to off-peak periods, which can reduce substation capacity. In addition, this paper uses DC feeders to replace part of AC feeders to meet the increased DC load in the grid, reducing feeder investment in the grid, as well as the total planning cost. In the end, the mixed integer linear programming method is used to solve the optimization problem through CPLEX, the 13-node distribution network system is used to compare and analyze the AC distribution network and AC-DC distribution network respectively, and the proposed method is verified.

Abstract:In order to realize the high-performance current tracking control and robustness of T-type three-level active power filter (APF), a no-beat optimization model predictive current control (FCS-MPC) algorithm is proposed and designed. Filter inductor online observer. Firstly, in order to reduce the computational complexity of the rolling optimization process, a simplified method of voltage vector control set based on harmonic current deadbeat prediction is proposed, which reduces the number of iterative calculations in each control cycle. Then, in order to reduce the influence of the inductance parameter mismatch on the performance of the FCS-MPC algorithm, an in-line correction of the parameters of the inductive observer is designed. Finally, the proposed algorithm is verified by simulation. The simulation results show that the proposed algorithm not only maintains the superior dynamic and static response performance of the traditional FCS-MPC, but also significantly reduces the computational complexity of the digital implementation and improves the robustness of the control system when the parameters are mismatched.

Abstract:In recent years, many factors lead to the frequency fluctuation accident of power system occurring often. The simulation results of multiple faults show that the estimation error of frequency modulation capability of power system is large at present, so it is necessary to use the power system frequency response (SFR) model to calculate dynamic frequency response. The shortcomings of the classical SFR model are analyzed, and the structure of the SFR model is improved accordingly, in which the dynamic characteristics of the polymer governor are taken into account. Then a parameter tuning method for the improved SFR model is proposed. In this method some parameters are obtained by direct calculation, the parameter equation for keeping steady-state compliance is added, and the other parameters are identified. The structure and parameters of the model are calculated in the simulation example, which verifies the validity of the improved SFR model and parameter tuning. The results indicate that the improved SFR model represent the major characteristics of SFR very well, and its accuracy is much higher than the classical SFR model.

Abstract:In order to reduce the impact of wind farm grid-connected on power grid and improve the robustness of grid-connected system, a sliding mode variable structure control strategy with feedback precise linearization is adopted to design the control of converter stations on both sides of wind farm grid-connected system via flexible direct current transmission. Firstly,a mathematical model of the voltage source converter in the dq coordinate system is established. Then the sliding mode variable structure control method based on precise linearization decoupling is used to design the converter stations on both sides of flexible HVDC transmission system, which solves the problems of poor regulation ability and difficult parameter tuning of traditional double closed loop control method, and further improves the anti-interference ability and dynamic stability of the state feedback linearization control system. Finally, the simulation model of wind farm grid-connected flexible HVDC system is established in Matlab/Simulink. The simulation waveforms before and after adding sliding mode control are compared to verify the performance of the control system.

Abstract:As the scale of the power grid continues to increase, short-circuit current exceeding the standard has become a major problem in operation. A comprehensive equivalent sensitivity method for limiting short-circuit current of power grid is proposed. The method is divided into two sub-problems: limiting short-circuit current and electrical safety check. In limiting short-circuit current sub-problem, unified short-circuit current limiting measure simulation method is proposed to simulate line breaking, generator breaking, line out of string, line installation series impedance, replace high impedance transformer, bus splitting, etc. Furthermore, using the branch impedance addition method, the node impedance parameters after the implementation of the above measures and the short-circuit current comprehensive equivalent sensitivity of the component breaking are calculated. In the electrical safety check sub-problem, the power flow and stability are checked after short-circuit current measure is applied to ensure the effectiveness of short-circuit current limiting measure. The practical examples of Qinghai power grid verify the effectiveness of the proposed method and can provide reference for the planning and operation of Qinghai power grid.

Abstract:In order to solve the problems such as variable redundancy and model complexity in ultra-short-term wind power prediction based on dynamic neural network (DNN), a novel method is proposed by combine the mean impact value (MIV) and principal component analysis (PCA) to optimize the predicting model constructed by DNN method. MIV method calculates the influencing degree from the input variables to the output and obtain the most important input variables to simplify the predicting model. However, its information utilization is low. PCA method extracts the principal components from the rest of the input variables. The information utilization can be greatly improved by adding a small number of principal components to make up for the deficiency of MIV method. It is verified by the data analysis and experiment that the optimized predicting model can assure the high contribution of the input variables and reduce the model complexity, which preserves the important information of the original system greatly, reduces the risk of introducing noise to the model, and makes the predicting result being improved significantly.

Abstract:There is a wealth of dynamic information hidden in system responses, which is subject to the random excitations. The analytical form of electromechanical response of power system under small environmental excitation are deduced, and the existence of electromechanical oscillation characteristics in the ambient response of power system is proved from the mathematical point of view. A method for oscillation parameters identification of power system on the basis of dynamic mode decomposition (DMD) algorithm is proposed. By comparing the results of the electromechanical oscillations parameters identification with theoretical characteristic parameters and analysis of the probability statistical results show that DMD algorithm has strong adaptability, meanwhile the correctness of oscillations parameters identification by using the ambient signals is verified by the simulation analysis of IEEE 4 generators 2 area system and a regional power grid disturbance recording data.

Abstract:Electricity, gas, heat and other energy sources are coupled in the micro-energy network, and the coordination optimization of capacity and operation of micro-energy grid is the key to to promote the absorption of renewable energy and improve energy efficiency. Aiming at the problem of optimal allocation of micro-energy network with CCHP, the system architecture and energy flow of micro-energy network with electricity, electricity input and electricity and heat output coupling are firstly analyzed, operational strategies to promote renewable energy consumption and improve energy utilization are proposed, and heat supply reliability index based on energy flow is established, then, the micro-energy grid system capacity and the operation strategy system optimal allocation model is established, considering the comprehensive evaluation index of investment cost, primary energy consumption, carbon dioxide emission and energy supply reliability. The simulation results show that the model can promote the consumption of renewable energy and improve the economy of the system while ensuring the reliability of energy supply.

Abstract:With the development of electric vehicles (EVs), the performance of the chargers has a great influence on the feeling of EVs users. This paper proposes a method based on the combination weighting method and the TOPSIS (technique for order preference by similarity to an ideal solution) method in order to evaluate the operating performance of EV chargers scientifically and reasonably. Firstly, the static and dynamic indicators of the charger performance evaluation are established. Static indicators include the output performance, safety performance, electromagnetic compatibility and other performance of EV chargers. The dynamic indicators are defined as the influence of the chargers on the power quality of the grid. Then, the combined weights are formed with the subjective and objective weights based on Game theory and all the weights are normalized. Finally, the TOPSIS method is used to score the closeness of the operating state of the charger. Three chargers chosen from Tianjin market are used as case study and the evaluation results correspond with the sales. So the proposed method can evaluate the chargers in a comprehensive way and provide a reliable suggestion for the charger consumers.

Abstract:With the large-scale intergration of the distributed generators and the flexible load, the dynamic behavior and fault characteristics of the distribution network are more complex and variable. The fault diagnosis in distribution networks faces new technical challenges. The waveforms-based fault diagnosis method is reviewed, which can be divided into direct matching method based on waveform similarity and indirect method based on waveform feature identification. The popular algorithms are analyzed and compared with each other. Then the advantages and disadvantages of each kind of algorithm are analyzed. Diagnosis methods for different application scenarios are summarized. Then, the future development and technical route of synchronous waveforms-based diagnosis methods are analyzed. The advantages and disadvantages of the algorithm are analyzed, and the diagnostic methods under different application scenarios are summarized. Finally, the development trend and technical roadmap of the fault diagnosis based on synchronous waveform are discussed. This review provides a useful reference for the futher study on smart distribution network fault diagnosis technology.

Abstract:In recent years, insulating faults caused by surface discharge of insulators are more common. Impulse withstand voltage detection and partial discharge detection under impulse are effective means for insulation evaluation and diagnosis.In this work, PD diagnosis was performed under standard oscillating lightning impulse (OLI), oscillating switching impulse (OSI) and alternating current (AC) voltages on a 110 kV gas insulated switchgear (GIS) bus model with artificial particles on the insulator surface. The results demonstrated that PDs under OIVs occurred at the rising slopes of the oscillating periods. Most of the reverse polarity pulses were found under OLI voltage, few were found under OSI voltage, and none were found under AC voltage. PD sequences excited from the negative needle point were composed of small and compact pulses. In contrast, positive-point excited PD sequences were composed of larger pulses but fewer of them. These results prove that the defects of particles on the insulator surface are more sensitive to OIVs than to AC voltage.

Abstract:Differential Scanning Calorimeter (DSC) was used to test the thermal history temperature of the insulation of XLPE cables in service. It was found that there were significant differences in the DSC 1st heating curves of different XLPE samples. On the most of the DSC curve of XLPE samples, thermal history peaks were found, and there were differences in the position, size and shape of these peaks. However, no thermal history peak could be observed on the DSC curve of the other samples. In order to investigate the reasons for the above differences, we performed DSC tests on XLPE samples at different temperatures for different temperatures. It was found that when the thermal history temperature experienced by XLPE cable insulation is higher than the crystal melting temperature of XLPE, there is no thermal history peak on the DSC primary heating curve.At this time, the change of DSC curre is mainly caused by the destruction of XLPE crystal structure by thermal sxygen degradation. When the thermal history temperature experienced by XLPE cable insulation is lower than the crystal melting temperature of XLPE, a thermal history peak can be observed on the DSC primary heating curve. The longer the XLPE cable insulation is at this thermal history temperature, the larger the area of the thermal history peak and the higher the peak temperature of the thermal history peak, and may even be higher than the actual thermal history temperature.

Abstract:In practical operation, DC field strength and thermal environment can affect the accumulation of space charge in polyethylene cable. And excess space charge can threaten the insulation performance. Although nanocomposites can suppress space charge, the space charge accumulation characteristics after thermal aging still need to be further studied. In this paper, the space charge accumulation behavior in low-density polyethylene (LDPE)/ silica (SiO₂) nanocomposites with different SiO₂ mass concentrations before and after thermal aging was studied by the pulse electro-acoustic method. The results show that thermal aging reduces the electrode injection barrier and generates deep traps and impurities with random distribution, increasing the accumulation of space charge. Many interface regions introduced by nano-SiO₂ generate deep traps with uniform distribution and form interfacial anti-electric field, inhibiting the transport and injection of carriers. Moreover, nanoparticles can improve the stability of materials, thereby enhancing the electrode injection barrier and delaying the thermal aging process. Compared with pure LDPE, LDPE/SiO₂ has the obvious ability to suppress space charge before and after thermal aging.

Abstract:The secondary resonance elimination devices are widely adopted in distribution network, but some field application effects are not satisfactory, and there are some cases of resonance elimination failure. In this paper, based on the analysis of voltage and flux variation characteristics of PT resonance in distribution network, a new control method of resonance elimination is proposed, which is to launch a device when zero sequence flux passes through zero point.This method avoides the influence of PT non-linear characteristics on the secondary resonance elimination process, consumes the resonant energy rapidly, and makes the network return to normal quickly. For the problem that the zero sequence flux cannot be measured in actual system, a zero-crossing detection method based on zero sequence voltage is proposed. The simulation results show that compared with the conventional resonance elimination method, the proposed method can effectively improve the success rate of the secondary resonance elimination, greatly reduce the working time of the secondary resonance elimination devices, and reduce the influence on the distribution network.

Abstract:Due to its special electromagnetic characteristics, transformers may have various hidden faults that fail to trigger related protection during operation, which will endanger the safe and stable operation of equipment and power grid. This paper proposes a model based on J-A hysteresis. The method of monitoring the online operation status of the transformer is to assist the operation and maintenance personnel to judge various faults that may occur in the actual operation of the transformer. Based on this, this paper firstly establishes a transformer model based on J-A hysteresis theory, and then collects the relevant waveform acquisition of the transformer in actual operation, and uses wavelet transform to analyze the waveform of the transformer under real-time conditions and the error of the predicted waveform of the transformer model established. Verification, through the verification results to determine whether the transformer is working properly, and finally, through the simulation experiments, the reliability of the proposed monitoring method is verified.

Abstract:The negative sequence and the zero sequence component cause three-phase unbalance. IEC proposes to measure the imbalance with the ratio between negative sequence and positive sequence component, ignoring the zero sequence component. Because of the difficulty in obtaining the sequence component, the RMS value of the measured voltage and current is often measured in the project. In contrast, such metrics exist in many forms and do not correspond to the order component metrics. It is urgent to construct a unified unbalance measurement method to considering the zero sequence and the non-correspondence between the quantity measurement and the sequence component. Three-phase unbalance control mainly relies on capacitive compensation, power electronic converter type compensation and commutation. Among them, capacitive compensation can achieve continuous adjustment but can not reduce line loss. Commutation can only be discretely switched and cannot be continuously adjusted. In the future, these two methods should be combined to manage the three-phase unbalance, in order to optimize the compensation effect.