Reactive Power Flow Convergence Adjustment Based on Deep Reinforcement Learning

Power flow calculation is the basis of power grid planning and many system analysis tasks require convergent power flow conditions.To address the unsolvable power flow problem caused by the reactive power imbalance,a method for adjusting reactive power flow convergence based on deep reinforcement learning is proposed.The deep reinforcement learning method takes switching parallel reactive compensation as the action space and sets the reward value based on the power flow convergence and reactive power adjustment.For the non-convergence power flow,the 500 kV nodes with reactive power compensation devices on the low-voltage side are converted into PV nodes by node type switching.And the quantified reactive power non-convergence index is acquired.Then,the action space and reward value of deep reinforcement learning are reasonably designed and the adjustment strategy is obtained by taking the reactive power non-convergence index as the algorithm state space.Finally,the effectiveness of the power flow convergence adjustment algorithm is verified by an actual power grid system in a province.

Analysis of multiple-faults of high-voltage circuit breakers based on non-negative matrix decomposition

High-voltage circuit breakers are the core equipment in power networks,and to a certain extent,are related to the safe and reliable operation of power systems.However,their core components are prone to mechanical faults.This study proposes a component separation method to detect multiple mechanical faults in circuit breakers that can achieve online real-time monitoring.First,a model and strategy are presented for obtaining mechanical voiceprint signals from circuit breakers.Subsequently,the component separation method was used to decompose the voiceprint signals of multiple faults into individual component signals.Based on this,the recognition of the features of a single-fault voiceprint signal can be achieved.Finally,multiple faults in high-voltage circuit breakers were identified through an experimental simulation and verification of the circuit breaker voiceprint signals collected from the substation site.The research results indicate that the proposed method exhibits excellent performance for multiple mechanical faults,such as spring structures and loose internal components of circuit breakers.In addition,it provides a reference method for the real-time online monitoring of high-voltage circuit breakers.

A Novel Defender-Attacker-Defender Model for Resilient Distributed Generator Planning with Network Reconfiguration and Demand Response

To improve the resilience of a distribution system against extreme weather,a fuel-based distributed generator(DG)allocation model is proposed in this study.In this model,the DGs are placed at the planning stage.When an extreme event occurs,the controllable generators form temporary microgrids(MGs)to restore the load maximally.Simultaneously,a demand response program(DRP)mitigates the imbalance between the power supply and demand during extreme events.To cope with the fault uncertainty,a robust optimization(RO)method is applied to reduce the long-term investment and short-term operation costs.The optimization is formulated as a tri-level defenderattacker-defender(DAD)framework.At the first level,decision-makers work out the DG allocation scheme;at the second level,the attacker finds the optimal attack strategy with maximum damage;and at the third level,restoration measures,namely distribution network reconfiguration(DNR)and demand response are performed.The problem is solved by the nested column and constraint generation(NC&CG)method and the model is validated using an IEEE 33-node system.Case studies validate the effectiveness and superiority of the proposed model according to the enhanced resilience and reduced cost.

Flexible Load Participation in Peaking Shaving and Valley Filling Based on Dynamic Price Incentives

Considering the widening of the peak-valley difference in the power grid and the difficulty of the existing fixed time-of-use electricity price mechanism in meeting the energy demand of heterogeneous users at various moments or motivating users,the design of a reasonable dynamic pricing mechanism to actively engage users in demand response becomes imperative for power grid companies.For this purpose,a power grid-flexible load bilevel model is constructed based on dynamic pricing,where the leader is the dispatching center and the lower-level flexible load acts as the follower.Initially,an upper-level day-ahead dispatching model for the power grid is established,considering the lowest power grid dispatching cost as the objective function and incorporating the power grid-side constraints.Then,the lower level comprehensively considers the load characteristics of industrial load,energy storage,and data centers,and then establishes a lower-level flexible load operation model with the lowest user power-consuming cost as the objective function.Finally,the proposed method is validated using the IEEE-118 system,and the findings indicate that the dynamic pricing mechanism for peaking shaving and valley filling can effectively guide users to respond actively,thereby reducing the peak-valley difference and decreasing users’purchasing costs.

Power Quality Assessment Based on Rough AHP and Extension Analysis

Due to the increasing power consumption of whole society and widely using of new non-linear and asymmetric electrical equipment,power quality assessment problem in the new period has attracted more and more attention.The mathematical essence of comprehensive assessment of power quality is a multiattribute optimal decision-making problem.In order to solve the key problem of determining the indicator weight in the process of power quality assessment,a rough analytic hierarchy process(AHP)is proposed to aggregate the judgment opinions of multiple experts and eliminate the subjective effects of single expert judgment.Based on the advantage of extension analysis for solving the incompatibility problem,extension analysis method is adopted to assess the power quality.The assessment grades of both total power quality and each assessment indicator are obtained by correlation function.Through a case of 110 kV bus of a converting station in a wind farm of China,the feasibility and effectiveness of the propose method are demonstrated.The result shows that the proposed method can determine the overall power quality of power grid,as well as compare the differences among the performance of assessment indicators and provide the basis for further improving of power quality.

Distributionally Robust Optimal Dispatch of Virtual Power Plant Based on Moment of Renewable Energy Resource

Virtual power plants can effectively integrate different types of distributed energy resources,which have become a new operation mode with substantial advantages such as high flexibility,adaptability,and economy.This paper proposes a distributionally robust optimal dispatch approach for virtual power plants to determine an optimal day-ahead dispatch under uncertainties of renewable energy sources.The proposed distributionally robust approach characterizes probability distributions of renewable power output by moments.In this regard,the faults of stochastic optimization and traditional robust optimization can be overcome.Firstly,a second-order cone-based ambiguity set that incorporates the first and second moments of renewable power output is constructed,and a day-ahead two-stage distributionally robust optimization model is proposed for virtual power plants participating in day-ahead electricity markets.Then,an effective solution method based on the affine policy and second-order cone duality theory is employed to reformulate the proposed model into a deterministic mixed-integer second-order cone programming problem,which improves the computational efficiency of the model.Finally,the numerical results demonstrate that the proposed method achieves a better balance between robustness and economy.They also validate that the dispatch strategy of virtual power plants can be adjusted to reduce costs according to the moment information of renewable power output.

A Weighted Combination Forecasting Model for Power Load Based on Forecasting Model Selection and Fuzzy Scale Joint Evaluation

To solve the medium and long term power load forecasting problem,the combination forecasting method is further expanded and a weighted combination forecasting model for power load is put forward.This model is divided into two stages which are forecasting model selection and weighted combination forecasting.Based on Markov chain conversion and cloud model,the forecasting model selection is implanted and several outstanding models are selected for the combination forecasting.For the weighted combination forecasting,a fuzzy scale joint evaluation method is proposed to determine the weight of selected forecasting model.The percentage error and mean absolute percentage error of weighted combination forecasting result of the power consumption in a certain area of China are 0.7439%and 0.3198%,respectively,while the maximum values of these two indexes of single forecasting models are 5.2278%and 1.9497%.It shows that the forecasting indexes of proposed model are improved significantly compared with the single forecasting models.

Integrated Generative Adversarial Network and XGBoost for Anomaly Processing of Massive Data Flow in Dispatch Automation Systems

Existing power anomaly detection is mainly based on a pattern matching algorithm.However,this method requires a lot of manual work,is time-consuming,and cannot detect unknown anomalies.Moreover,a large amount of labeled anomaly data is required in machine learning-based anomaly detection.Therefore,this paper proposes the application of a generative adversarial network(GAN)to massive data stream anomaly identification,diagnosis,and prediction in power dispatching automation systems.Firstly,to address the problem of the small amount of anomaly data,a GAN is used to obtain reliable labeled datasets for fault diagnosis model training based on a few labeled data points.Then,a two-step detection process is designed for the characteristics of grid anomalies,where the generated samples are first input to the XGBoost recognition system to identify the large class of anomalies in the first step.Thereafter,the data processed in the first step are input to the joint model of Convolutional Neural Networks(CNN)and Long short-term memory(LSTM)for fine-grained analysis to detect the small class of anomalies in the second step.Extensive experiments show that our work can reduce a lot of manual work and outperform the state-of-art anomalies classification algorithms for power dispatching data network.

A QoS-Aware Packet Scheduling Mechanism in Cognitive Radio Networks for Smart Grid Applications

Cognitive radio sensor network is applied to facilitate network monitoring and management, and achieves high spectrum efficiencies in smart grid. However, the conventional traffic scheduling mechanisms are hard to provide guaranteed quality of service for the secondary users. It is because that they ignore the influence of diverse transition requirements in heterogeneous traffi c. Therefore, a novel Qo S-aware packet scheduling mechanism is proposed to improve transmission quality for secondary users. In this mechanism, a Qo S-based prioritization model is established to address data classification firstly. And then, channel quality and the effect of channel switch are integrated into priority-based packet scheduling mechanism. At last, the simulation is implemented with MATLAB and OPNET. The results show that the proposed scheduling mechanism improves the transmission quality of high-priority secondary users and increase the whole system utilization by 10%.

Wind Speed Forecasting Based on ARMA-ARCH Model in Wind Farms

Wind speed forecasting is signif icant for wind farm planning and power grid operation. The research in this paper uses Eviews software to build the ARMA (autoregressive moving average) model of wind speed time series, and employs Lagrange multipliers to test the ARCH (autoregressive conditional heteroscedasticity) effects of the residuals of the ARMA model. Also, the corresponding ARMA-ARCH models are established, and the wind speed series are forecasted by using the ARMA model and ARMA-ARCH model respectively. The comparison of the forecasting accuracy of the above two models shows that the ARMA-ARCH model possesses higher forecasting accuracy than the ARMA model and has certain practical value.

Distributed Real-time Optimal Power Flow Strategy for DC Microgrid Under Stochastic Communication Networks

This paper addresses a distributed real-time optimal power flow(RTOPF) strategy for DC microgrids. In this paper, we focus on the scenarios where local information sharing is conducted in stochastic communication networks subject to random failures. Most existing real-time optimal power flow(OPF) algorithms for the DC microgrid require all controllers to work in concert with a fixed network topology to maintain a zero gap between estimated global constraint violations. Thus, the high reliability of communication is required to ensure their convergence. To address this issue, the proposed RTOPF strategy tolerates stochastic communication failures and can seek the optimum with a constant step size considering the operation limitations of the microgrid. These aspects make the strategy suitable for real-time optimization, particularly when the communication is not reliable. In addition, this strategy does not require information from non-dispatchable devices, thereby reducing the number of sensors and controllers in the system. The convergence of the proposed strategy and the optimal equilibrium points are theoretically proven. Finally, simulations of a 30-bus DC microgrid are performed to validate the effectiveness of the proposed designs.

Single-phase-to-ground Fault Detection with Distributed Parameters Analysis in Non-direct Grounded Systems

The fast and accurate detection of the single-phaseto-ground fault is of great significance for the reliability and safety of the power supply.In this paper,novel algorithms for distribution network protection were proposed with distributed parameters analysis in non-direct grounded systems.At first,novel generating mechanisms of zero-sequence voltage and residual current were proposed.Then the compositions of residue parameters,including residual current and residual admittances,were decomposed in detail.After that,an improved algorithm for a fault resistance calculation of a single phase-to-earth fault was also proposed,and the algorithm is much more convenient as it only needs to measure the variation of the zero-sequence voltage and does not need the prerequisites of the faulty feeder selection.Furthermore,the fault feeder can also be selected by an improved calculation algorithm of zero-sequence admittance of the faulty feeder,which cannot be affected by the asymmetry of the network.Theoretical analysis and the MATALB/Simulink simulation results demonstrate the effectiveness of the proposed algorithms.

Comprehensive Control Scheme for an Interlinking Converter in a Hybrid AC/DC Microgrid

This paper presents a comprehensive control scheme for the interlinking converter(ILC)in a hybrid AC/DC microgrid consisting of the outer loop flexible power sharing control and the improved robust inner loop control.The outer loop power control of ILC is presented to achieve flexible power sharing of distributed generations(DGs)in the hybrid microgrid,depending on different power management objectives,which is realized based on the deduced balance state equation,and regulating the frequency and DC voltage at the same time.The improved robust inner loop control of ILC is also presented to suppress external disturbance and system model uncertainties with the improved dynamic response.This improved inner loop control which includes a disturbance observer link,can force the converter current to track the reference value with no steady error and improve the dynamic stability of the microgrid・With the proposed outer loop power sharing control and improved inner loop control,the comprehensive control scheme for the ILC is presented・Simulations cases show the effectiveness and superiority of the proposed comprehensive control scheme.