Abstract:Aiming at the problem that the wind storage system's power fluctuation suppression effect is unsatisfactory, the wind storage system's operation is optimized. Considering the time series coupling characteristics of wind storage system operation and the influence of future wind power fluctuation on the energy storage system (ESS), an optimization strategy based on stabilizing target optimization method and model predictive control (MPC) is proposed. Firstly, the expected grid?connected power is calculated according to the predicted wind power and the constraints of ESS, and the local prediction accuracy is introduced to correct it. Then, combined with the current state of charge (SOC) of ESS, the optimal stabilization target is obtained by fuzzy control; Lastly, the MPC with particle swarm optimization (MPC?PSO) strategy is used to optimize the ESS power, so as to minimize the difference between the grid?connected power of next time and the optimal target power and minimize the ESS power. The simulation results show that the strategy proposed in this paper has a better wind power fluctuation smoothing effect and can effectively reduce the operation cost of energy storage.

Abstract:The application of energy storage optimization configuration and demand?side response in the integrated energy system can decouple the constraints of "heating to determine power" in the operation of combined heat and power (CHP) units, and improve energy utilization efficiency. Therefore, taking the performance of power generation, heating, electricity consumption and heat consumption on both sides of the source and charge into account, installing the electricity?heat energy storage equipment on the source side to realize the thermoelectric decoupling of CHP, and fully utilizing the dispatching value of the electricity?heat integrated demand response (IDR) resources on the load side, an optimal configuration model for electric and thermal energy storage capacity of regional integrated energy systems considering the integrated demand responses is established. The model considers the constraints of electricity storage and heat storage, power balance constraints, etc., and aims to minimize the total annual cost of the regional comprehensive energy system to solve the problem of optimal economic allocation of electricity and thermal energy storage. The results of case study show that coordinated operation of multiple energy storage and load side can reduce operating cost, and considering the integrated demand response reduces the configuration capacity of energy storage. The energy storage configuration on the source side and the integrated demand response on the load side promote wind power consumption, and improve the economy, precision and flexibility of the integrated energy system operation.

Abstract:The increasing charging load of electric vehicles will have an impact on the economy operation and reliability of the power system. Power system production simulation is an important analysis method of cost and reliability , and it is necessary to consider such charging loads. Firstly, based on the impact of the temperature, the charging load prediction model of electric vehicles is established by the Monte Carlo sampling method. Then, in a continuous load model for electric vehicles to pursuit the lowest annual production and operation cost of the system. The factors such as maintenance plan, start and stop cost, and operation cost constraints of each generator are incorporated, an optimization model of mid/long?term power system production is built. Finally, the CPLEX is used to solve this optimization model to realize the simulation of power system productions, which can modeling the factors like production and operation parameters of each generator, maintenance scheme, total system power generation and total system operation cost. Based on the modified IEEE?RTS 79 test system, a case study is presented to verify the feasibility of the proposed method, and the influence of electric vehicle charging load on power systems is also analyzed.

Abstract:The proportion of renewable energy in the supply side of the electricity market is gradually expanding, whilst its supply uncertainty increases the risk of electricity market trading. Under the background, an integrated trading strategy is proposed considering the day?ahead electricity price and supply uncertainty. Firstly, renewable energy aggregators utilize the demand response to deal with the uncertainty of electricity production. Then, they make a contract with demand response operators by comprehensively considering the contractual settlement price and activation charge of demand response. After that, the power is transferred according to demand. The cost of demand response switching to different times is decided by the contracted power, no need to consider when and where they will be used. Secondly, the operator participates in day?ahead market transactions through non?contractual demand response and further increases its revenue. Finally, the conditional value?at?risk (CVaR) by assessing expected cost volatility. Thereby, the risk aversion is incorporated into the decision model to avoid overly conservative trading scenarios. In the end, the proposed decision method is evaluated in a test system to validate the effectiveness of the method. It is shown that the proposed method can increase the expected returns of different market participants while reducing the associated risks.

Abstract:After the receiving?end AC fault occurs in the multi?infeed HVDC system, it is easy to cause simultaneous commutation failure of multiple DCs, resulting in a large quantity of power shortage and active power fluctuation. In order to achieve effective suppression of grid power fluctuation after the receiving?end AC faults in the multi?infeed HVDC transmission system, from the perspective of AC/DC interaction strength and the critical nodes and lines of active power, a siting method of the energy storage system is proposed for suppressing power fluctuation of the AC system after commutation failure is proposed. Firstly, the candidate areas are determined according to the multi?infeed interaction factor(MIIF), and then the candidate sites are determined by the multi?infeed effective short circuit ratio(MIESCR). Finally, the specific location of the energy storage system is determined by the active power fluctuation rate of the whole network(APFRWN). The simulation results are verified in PSCAD based on the modified IEEE?39 bus system. The location method of the energy storage system proposed in this paper can effectively reduce the power fluctuation caused by DC commutation failure caused by AC fault at the receiving end and improve the stability of the AC system.

Abstract:This paper proposes an optimal siting and sizing method for distributed energy storage in distribution networks considering islanding duration uncertainty. The uncertainty of the islanding duration is described based on robust optimization, which improves the ability of islanding operation and guarantees the un interrupted power supply of important loads. The planning model aims to minimize the sum of the annual investment cost of energy storage, the annual power purchase cost of the distribution network, and the operation and maintenance cost of energy storage, the configuration and operation constraints of distributed energy storage, photovoltaic output constraints, distribution network flow constraints, grid operation safety constraints, power purchase constraints and load power supply constraints and the uncertainty of islanding duration is considered in the planning model. The column?and?constraint generation algorithm (C&CG) is adopted to solve the mixed?integer second?order cone robust programming model. The planning model is carried out under an actual distribution network. The results show that the established model can optimize the planning solution of distributed energy storage while guarantee the power supply of important loads considering the islanding duration uncertainty, and improves the ability of distribution network islanding operation and economy

Abstract:With the strategic goal of developing "double carbon", it is required to further promote the large?scale integration of wind and solar power as the main source of clean energy. In order to improve the wind power consumption capability and reduce the wind power curtailment, a interactive optimization operation model of regenerative electric heating and wind power considering user satisfaction is proposed. Firstly, the operation principle of regenerative electric heating equipment participating with wind power is analyzed; then, a multi?objective optimization model for the combined operation of regenerative electric heating and wind power is established considering wind power consumption, economical property and user satisfaction. A grey relational analysis based improved chaotic particle swarm optimization algorithm is adopted for solving the proposed model; finally, an operation scheme which satisfying all the operation requirements is proposed based on the simulation data of a real grid. The simulation results show that the model can effectively increase the wind power consumption capability, reduce the operating cost, satisfy the user's thermal comfortable level, and provide decision support for future development for the regenerative electric heating and wind power.

Abstract:The planning of electric vehicle charging stations needs to consider not only the cost of connecting to the power grid but also the convenience of electric vehicle users, as well as the impact of traffic congestion on user travel. Based on the spatial and temporal distribution characteristics of electric vehicles, an electric vehicle spatial?temporal distribution model is established. The impact of traffic congestion on the travel characteristics of electric vehicle users and the selection of charging stations is fully considered, and the electric vehicle entry capacity is calculated at different times. A mixed?integer second?order cone programming model is established to minimize the total investment cost, total operating cost, and total user travel cost, taking into account the impact of traffic congestion on electric vehicle charging station planning. The effectiveness and feasibility of the proposed method for planning electric vehicle charging stations considering traffic congestion are verified through hypothetical planning areas.

Abstract:As a high?quality demand response resource, electric heating load groups can improve the operation of distribution network through optimization dispatch. In order to achieve reasonable dispatching of electric heating loads, an optimal strategy for electric heating load groups that takes into account user behavior difference and distribution network flow is proposed. At first, the prediction mechanism of the controllable capacity of electric heating load is analyzed. Secondly, according to the influencing factors of user demand response behavior, the users are integrated into multiple electric heating load aggregates, and the thermal comfort design and controllable capacity solution of each group are carried out. Considering the power flow constraints of the distribution network, an optimal dispatch model for electric heating load groups with the goal of minimizing load fluctuations is established. The example analysis shows that the optimal scheduling model can improve the accuracy of electric heating load response prediction, improve the effect of peak?cutting and valley filling in the distribution network, and is conducive to the economic and safe operation of the system.

Abstract:The characteristics of AC/DC hybrid system faults are different from those of AC systems, which will affect the reaction of traditional AC protections. For the power frequency variation distance protection, firstly, the variation range of equivalent impedance angle of protection back system is analyzed from the perspective of quadrants, and then its action characteristics are further assessed. The results show that the protection range of power frequency variation distance protection in AC/DC hybrid system will be reduced, the anti?transition resistance capacity will be weakened, and the protection range may be completely lost in some extreme cases. To solve this, the reliability coefficient is diminished, the cosine value of phase difference of zero sequence current on both sides is used to avoid misoperation outside the area, and thus an improved measure of power frequency variation distance protection based on zero sequence current phase compensation can be proposed. An case study platform is built on PSCAD/EMTDC to verify the correctness of the theoretical analysis, and the effectiveness of the proposed improvement measures.

Abstract:With the grid connection of large scale distributed generations, the regulation and control of distribution network becomes more diversified. Under the circumstance, the key influence factors of the decoupling optimization and coordinated optimization of active and reactive power is investigated. Firstly, an analysis method considering the influence of the changing of power source, network, and load on the power optimization of distribution network is proposed for problem of the power optimization for distribution network containing distributed generation. Then, based on the two power optimization models, the interaction mechanisms between power optimization and grid price, network parameters and load are analyzed. And the mathematical expressions of line impedance ratio and load power factor to transmission power are given. In addition, the boundary conditions of two optimization modes are given qualitatively. At last, an improved IEEE?33 bus system is included to verify the impact of core influence factors for two power optimization modes on the total power generation cost and calculation time, including the feed?in tariff, the impedance rate and length of transmission line, the power factors and load.

Abstract:A reasonable comprehensive evaluation of distribution network resilience can reflect the ability of distribution network to restore power supplies when random faults occur, which is also helpful to diagnose the performances of different countermeasures for a higher resilience and lower losses. According to the mutual impacts by network nodes under different random faults, one comprehensive evaluation method for the resilience of active distribution networks (ADN) is presented to handle such an uncertainty. Firstly, three resilience indexes: fault recovery time, outrage time and energy loss percentage is established to comprehensively evaluate the resilience of node. Aiming at quantitatively rating the interactive influences of nodes caused by the network connectivity and distribution energy resource (DER) location in a fault event, this paper retains the data information characteristics of multiple random fault resilience, and proposes a node weight calculation method, to comprehensively evaluate the overall resilience of ADN. In a case study, IEEE 33 bus model is taken as an example medium voltage distribution network, and Monte Carlo simulation is used to simulate random faults in that system. From the results, the proposed method can reflect the improvement of resilience after the implementation of different countermeasures, thus the effectiveness of this method can be verified.

Abstract:Due to bus bar function of single?ended injection for 10 kV main line, the end?terminal is prone to suffering low voltage problems. Transforming the main line voltage drop constraint into the outlet load constraint can provide great convenience for the safe operation and planning of power grid. Therefore, an estimation method of the first?end load security region boundary is proposed in this paper. Aiming at the numerous station loads along the line, a torque method for rapid estimation of the voltage drop across the main line basing on segmented voltage drop characteristics is proposed. Then the actual station load distribution is equivalent to the characteristic form of balanced distribution along the line with several heavy load points, which is more consistent with the actual situation. Based on the moment method, the relationship between the load at the head end of the line and the voltage drop is obtained, and finally the linearized equation of the safety zone boundary of the head load is obtained. The results of the calculation examples show that the voltage drop analysis and safety zone boundary analysis of the moment method satisfy the engineering accuracy requirements. Based on the boundary of the first?end load safety zone, the load margin of the 10 kV line can be obtained, which is convenient for load controlling and rolling distribution network planning, and has good engineering application value.

Abstract:The clock?inaccuracy at the load measuring point on the 10 kV line leads to an abnormal line loss rate, while the existing manual methods have the problems of low efficiency and low intelligence. Therefore, based on the fluctuation characteristics of line loss rate curve, a new method for identifying the load types of clock?inaccuracy metering points is proposed to fit the mapping relationship between load type and the clock?inaccuracy line loss rate by Bayesian network (BN). In order to solve the problem of lack of clock?inaccuracy samples, the metering clock deviation modules are respectively set for the load metering points to generate a sample set of clock?inaccuracy line loss rate in the simulation model based on the actual operation data of the line. The fuzzy C?means clustering is then introduced to classify the load according to the shape similarity of the load curve, and the data dimensionality reduction is realized in scenarios with heavy load. Relying on research data from the synchronous line loss management system, the calculation example verifies the feasibility and accuracy of the proposed method. It is shown that the method can realize load type identification of clock?inaccuracy, and provide a reference for quickly locating the abnormal energy meters.

Abstract:Load node classification based on daily load curve is an important part of load modeling. The detailed and appropriate classification results retain the internal characteristics of load nodes and can improve the efficiency of power system simulation calculation. At present, the node clustering method based on artificial intelligence has made rapid progress. However, the overall adaptability to data deep feature extraction is still insufficient. This paper presents the daily load curve clustering method based on the improved deep embedded algorithm, which uses the ability of neural network to effectively extract the deep features of the data. Then, an improved method of increasing the dimension first and then clustering is proposed. Through the comparative analysis of numerical examples, the feasibility of the proposed algorithm and the correctness of the improved dimension reconstruction clustering method are verified.

Abstract:Load identification is an important part of power system simulation. In order to obtain accurate load identification dynamic parameters, scholars at home and abroad have done a lot of in-depth research on the reasonable construction of identification models. Firstly, on the basis of the traditional load model, two real-time dynamic parameter identification load models based on ambient noise are deduced, which realizes the decoupling of the input data sequence of the measurement and identification data in the calculation of the identification model. At the same time, it avoids the negative impact of the identification of the initial physical quantity redundancy and the iterative amplification of the identification quantity error on the parameter power response capability. Then, the data power response capability of the two measurement and identification models is verified and analyzed from the ambient noise simulation data and the actual measurement data. The results show that the load dynamic parameters measured and identified are applicable to the current power simulation system, indicating that the research can provide a new direction for further research on the identification model and provide a data basis for load controllability.

Abstract:Non?intrusive load disaggregation technology can effectively mine the appliance information of customers, which is the basis to carry out interactive customer load response by the grid company. The conventional non?intrusive load disaggregation technology has several drawbacks, such as limited scope of application and low accuracy. In this paper, a non?intrusive load disaggregation model with multiple optimization selection of appliance characteristics is proposed. First, an adaptive sliding data window is designed for appliance operation characteristics to obtain a more complete power segment and to adjust the network input and output dimensions. Second, the appliance features can be extracted and deepened by fusing shallow convolutional neural networks (CNN) with two?layer nested long and short?term memory networks (NLSTM), which is further fed into an improved attention mechanism to obtain the optimum appliance feature sequence by adjusting the feature weights. Finally, experimental analysis on the REDD dataset shows that the multiple selection, deepening and reusing of appliance features can significantly improve the accuracy of load decomposition while reducing training time.

Abstract:In modern DC microgrids, distributed power sources are connected to the common DC bus through power electronic converters. Although the control loop of the converter has good stability margin, the interconnection of multiple converters may affect the microgrid dynamic performance and stability. Therefore, in order to ensure the required dynamic performance of the multi?converter system, a DC bus impedance peak estimation method based on the phase margin of the voltage loop is proposed in this paper. On this basis, an optimized control scheme for the DC bus impedance is further proposed. And through experiments to verify the effectiveness of the proposed control method. First, derive the expression of DC bus impedance based on the voltage control loop gain of the source?side converter; then, through reasonable assumptions, the DC bus impedance peak value is estimated based on the phase margin of the voltage control loop; finally, through the voltage control A sinusoidal signal is injected into the loop to continuously monitor the peak value of the bus impedance. By optimizing the control parameters of the voltage regulator, the peak value of the DC bus impedance is effectively reduced. The experimental results show that the proposed monitoring scheme could reduce the measurement and the calculation burden, and the optimized control scheme improves the stability and dynamic performance of the DC microgrid.

Abstract:To deal with the power fluctuation and the consumption of photovoltaic power generation with the premise of the demand of power supply and heating in northwest China, this paper studies the collaborative control of multi?types of energy storage based on fluctuation parameters by taking the electric?thermal joint micro?network composed of photovoltaic, heat pump and hybrid energy storage as the research object. Firstly, the characteristics and energy conversion mode of electro?thermal joint microgrid are analyzed. Then, a two?layer collaborative control strategy of multi?type energy storage based on fluctuation parameters is designed for the electro?thermal joint micro?grid. In the upper layer, a strategy of multi?type energy storage collaborative power fluctuation suppression is proposed based on the time?scale of power fluctuation and the low?pass filtering method with variable parameters. In the lower layer, an adaptive control strategy for energy transfer of hybrid energy storage is proposed based on the requirement of the power fluctuation suppression and the mechanism of voltage and frequency in microgrid. Finally, the proposed algorithm is verified by simulation. The simulation results show the proposed method can suppress the power fluctuation of renewable energy and can extend the battery life on the premise of reducing the investment cost of electric energy storage effectively.

Abstract:Aiming at the problem that the recognition accuracy of multiple disturbances is not high under noise interference, a new classification method of power quality disturbances based on deep belief network is proposed. Firstly, the stationary wavelet multi?scale transformation is performed on the power quality disturbance signal, and then the soft threshold function is used to process the estimated wavelet coefficients to reconstruct the original signal, thereby realizing the denoising of the power quality disturbance signal. Moreover, it is further proposed to use the deep belief network to classify and identify the reconstructed single disturbance signal and multiple disturbance signals. The calculation example shows that even under the interference of 20 dB noise, the classification accuracy rate is as high as 93%. The results show that the recognition accuracy of the method is high for 7 kinds of single disturbance and 13 kinds of multiple disturbance signals, which verifies that the method has strong anti?noise interference ability.

Abstract:One of the advantages of the phasor measurement unit (PMU) is to provide synchronous phase angle data. However, due to factors such as abnormal time synchronization, device failures and other factors, some of the measured PMU phase angle data are abnormal, which could affect the PMU application. This paper proposes a method for correcting the time?varying deviation of PMU phase angle difference (PAD) at both ends of the line based on line reactive power loss. Firstly, based on the measured value and estimated value of the line reactive power loss, a single?time PAD deviation estimation model is constructed; Secondly, aiming at the problem that the single?time estimation model is greatly affected by noise, based on the relationship between the voltage PAD and the line parameter, the time?varying deviation estimation model by single?snapshot data is transformed into a constant line parameter estimation model by multi?snapshots data; after estimating the line parameters through the golden section search algorithm, the PAD deviation can be calculated by the PMU amplitude and power data, the estimation and correction of the time?varying PAD deviation at multiple times are realized. The test results of simulated and measured PMU data from a province show that the method could accurately estimate the PAD deviation and has strong anti?noise ability. Compared with the existing methods, this method only needs data under single power flow condition to realize the estimation and correction, and is suitable for the correction of different types of time?varying deviation.

Abstract:The uncertainty of new energy power generation poses a great challenge to the transmission system. Since the transmission cable has a certain overload capacity, it can be overloaded for a short time under the premise that the temperature does not exceed the limit value. A method of making full use of the overload level of the cable is proposed to improve the transmission capacity of the cable to new energy power generation in a short time. This method firstly uses the finite element method to calculate the cable temperature field distribution, establishes the allowable overload running time of the line and analyzes its influencing factors. Then the improved BP neural network algorithm is introduced to predict the overload time in combination with the actual power generation output curve. The results show that the BP neural network model has high accuracy and can be applied to evaluate the cable limit transmission capacity and provide rapid support for dispatching decisions.

Abstract:With the increase of electricity demand, the heavy overload of distribution network lines during the peak period of electricity consumption becomes more serious, which increases the potential threats on the safety of grid operation. The short?term forecast of the heavy overload state of distribution lines is of great significance for rationally arranging the operation mode, for dispatch management, and for the safety operation of the line during peak load periods. This paper proposes a short?term forecast method for the heavy overload state of lines and a prediction model that CNN?GRU hybrid neural network with Attention mechanism. The historical load rate of lines with high auto?correlation and meteorological factors are combined as the input features, which is further used to extract the valid features by the CNN. The GRU neural network is utilized to analyze and predict time series data. By using the Attention mechanism to reassign corresponding weights, the load rate regression prediction result can be outputed，which can be finally converted into the load level according to the load level division standard. The method in this paper is performed on a 10kV line in a certain district of Shanghai. The experimental results show that this prediction method is more suitable for line heavy overload prediction than the method using the classification prediction model with the same model structure but with load level as input.

Abstract:The safe operation of transmission lines is an important basis for ensuring stable transmission of electric energy, and the safety is greatly affected by meteorological factors. In order to evaluate and improve the risk early warning capability in the transmission system, this paper proposes a transmission line safety evaluation model that takes into account meteorological factors. Firstly, according to the characteristics of the transmission line architecture, an evaluation index system reflecting the operation status of the transmission line is constructed. At the same time, a new graph model weighting method is proposed based on the principle of graph theory to overcome the human subjectivity factor disadvantages in the weighting of the traditional matter?element extension model. Secondly, based on the matter?element extension theory, the comprehensive risk matter?element and safety rating of transmission lines are constructed, and the risk safety assessment and dynamic management model of transmission lines considering meteorological factors are established. Finally, taking a certain region as an example to study the operation risk of its transmission lines, the analysis results verify the scientificity and practicability of the transmission line safety assessment model.

Abstract:Reliability evaluation based on the failure rate data is an important basis for the health status management and maintenance of smart meters. However, the small sample characteristics of outliers and failure rates limit the evaluation performance of traditional smart energy meter reliability prediction models. Therefore, a prediction model of smart meter failure rate under multi?environment stress based on weighted local outlier factor and Gaussian process regression is proposed in this paper. Firstly, a weighted local outlier factor is employed with the model to identify and then delete potential outliers in failure rate data sets; then, different kernel functions are selected to match the characteristics of multiple stress inputs in typical environments, and choose the best one. Finally, the interval change of the 95% confidence level of the failure rate is predicted by the posterior distribution of the Gaussian process, and the interval reliability is obtained based on this. Case analysis of fault samples of smart meters in two typical environmental areas shows that the proposed model could effectively predict the trend of failure rate of smart meters under multi?environmental stress, and could accurately solve its reliability.

Abstract:This paper addresses the difficulty of the accurately detecting electricity theft in complex grid environment and proposes a multi?category electricity theft detection method based on CNN?LSTM hybrid model. Firstly, the excellent feature abstraction ability of convolutional neural networks (CNN) is utilized to extract the non?periodic local features of one?dimensional electricity consumption data. Then, the long short?term memory (LSTM) is adopted to capture the correlation between daily power consumption data and extract periodic power consumption features to establish feature fusion layer network. After that, the feature vectors extracted by CNN and LSTM are horizontally splicing to obtain a new fusion vector. Based on this, the accurate detection of multiple types of electric theft behavior are realized. Experimental results show that the proposed method can accurately identify multiple types of electric theft behavior, and the detection results are more comprehensive and accurate than the existing detection methods.

Abstract:The complex electromagnetic environment around the high-voltage transmission line affects the steady operation of the patrol UAV directly, and further affects the evaluation results of the patrol inspection by ground operators. In order to ensure the reliability of UAV in the harsh electromagnetic environment, the electromagnetic environment around UAV is simulated. Based on the coupling path of interference electromagnetic wave, the electric field coupling model and the magnetic field coupling model are constructed respectively. The coupling between the internal electrical structure of UAV and the high-voltage transmission line is analyzed to determine the disturbed target. Based on the generation of the interference source and the electromagnetic coupling path, the electromagnetic compatibility optimization of the UAV is carried out. The electromagnetic shielding is used to weaken the electromagnetic interference to the internal electrical structure and sensitive components of the UAV, so that the electromagnetic compatibility performance of the UAV is improved, and the operation stability of the patrol UAV is guaranteed.

Abstract:In order to analyze the impulse characteristics of tower grounding device under continuous lightning impulse, continuous impact tests of typical grounding bodies with different soils, materials and shapes are carried out. Under continuous pulse impact, with the increase of impulse time interval, the secondary impact grounding resistance increases from the lower value at soil breakdown to the single pulse impact grounding resistance. However, within a certain time interval, the secondary impulse grounding resistance of grounding device in soil with less water content is significantly greater than that of single pulse impulse. The secondary impulse grounding resistances of grounding bodies with different shapes and materials are different. In the tower grounding design, the increase of impulse grounding resistance under continuous lightning impulse needs to be considered, and the material and shape of grounding body should be comprehensively considered. Based on the experimental results, the ATP Draw simulation modeling method of tower grounding system under continuous lightning impulse is proposed. This method achieves good effect when simulating the spark effect of impulse grounding and the recovery process of soil resistivity after soil breakdown.