Search

With the widespread use of new energy sources and Internet of things, the power market landscape has become complex. In particular, new energy is more stochastic and volatile; it is prone to the problems of inaccurate forecasting on longer time scales, affecting electricity trading. This study proposes a new method for predicting medium-term load series data based on the transformer-lightGBM. The method first preprocesses electricity market data, including missing value processing, outlier processing, overall analysis, and correlation analysis, to extract features with a strong correlation to medium-term electricity consumption forecasts. Then, a transformer neural network is used to learn the complex patterns and dynamic time scales of the load series data to predict the day-ahead market series. Finally, lightGBM is used to combine power characteristics and time characteristics to forecast power consumption. The effectiveness of the proposed method is proved using the ISO-NE dataset. Experimental results indicate that the present method verified more accurate prediction than LSTM-based methods.

With the development of the Internet of things and smart grid technologies, modern electricity markets seamlessly connect demand response to the spot market through price-responsive loads, in which the trading strategy of load aggregators plays a crucial role in profit capture. In this study, we propose a deep reinforcement learning-based strategy for purchasing and selling electricity based on real-time electricity prices and real-time demand data in the spot market, which maximizes the revenue of load aggregators. The deep deterministic policy gradient (DDPG) is applied through a bidirectional long- and short-term memory (BiLSTM) network to extract the market state features that are used to make trading decisions. The effectiveness of the method is validated using datasets from the New England electricity market and Australian electricity market by introducing a bidirectional LSTM structure into the actor-critic network structure to learn hidden states in partially observable Markov states through memory inference. Comparative experiments of the method show that the method can provide greater yield results.

With the consumption of new energy and the variability of user activity, accurate and fast demand forecasting plays a crucial role in modern power markets. This paper considers the correlation between temperature, wind speed, and real-time electricity demand and proposes a novel method for forecasting short-term demand in the power market. Kernel Support Vector Machine is first used to classify real-time demand in combination with temperature and wind speed, and then the temporal convolutional network (TCN) is used to extract the temporal relationships and implied information of day-ahead demand. Finally, the Gradient Boosting Regression Tree is used to forecast daily and weekly real-time demand based on electrical, meteorological, and data characteristics. The validity of the method was verified using a dataset from the ISO-NE (New England Electricity Market). Comparative experiments with existing methods showed that the method could provide more accurate demand forecasting results.

Accurately identifying the key links in active distribution network (ADN) that affect the system's safety and stability is of great significance to improve the efficiency of operation and maintenance and risk early warning for distribution network (DN). In this study, the node resilience (NR) is first proposed as an indicator for structural stability identification of DN. Then, a critical node identification method considering the security, economy and structure stability in ADN is proposed, which uses the NR, the utility risks of branch load change, node voltage deviation and line loss variation as the identification indicators. Also, combining with subjective experience and objective data, the comprehensive evaluation method of critical nodes is improved, and the multilevel and multidimensional evaluation model of critical nodes is constructed. Furthermore, the time sequence of load change and photovoltaic output is considered. Finally, the rationality and effectiveness of the proposed method are proved by taking IEEE 33-bus system and an actual ADN as examples.

Electric shock current identification is essential for the safety in power distribution network. Moreover, as different categories of object have different electric shock current characteristic, a classification model for shock current is essential to be proposed before identification. Therefore, the authors proposed a two-stage framework, including the AdaBoost for the classification and an improved support vector machine (SVM) method for the identification. In the classification stage, the AdaBoost learns the hidden pattern of different electric shock current and generates a predictive model for current classification. Based on the classification results, a fusion method called SVM–NN is proposed in the identification stage, which is based on SVM and neural network (NN) to make fusion determination. The SVM–NN takes advantages of SVM and NN for integration analysis. Based on real data, these classification and identification methods are evaluated. Results show that the proposed method can significantly improve the identification accuracy of electric shock current signal comparing to traditional methods.

Multi-Day Load Forecasting is the basis of Electricity Spot Markets Analysis and operation. At present, the Research on Electricity Spot Markets Load Forecasting has not considered the influence of Day-ahead Market on Real-Time Market, which leads to deviation and consumes a lot of time and energy. Therefore, this paper presents a Multi-Day Load Forecasting model based on Multiple Long Short Term Memory(M-LSTM), Consider the effects of Day-Ahead Locational Marginal Price(LMP) and Day-Ahead Cleared. The proposed method solves the problem that the single deep learning network structure is difficult to keep the time sequence characteristics between samples in the training process. In this paper, Case studies on the New England Electricity Market (ISO-NE) show that our proposed method is superior to S-LSTM in the forecasting. The MAPE of the model is 14.01%, compared with Single Variable LSTM(S-LSTM), it decreased by 3.33%. The RMSE of the model is 457.22 MW, compared with Single Variable LSTM(S-LSTM), it decreased by 98.09 MW. Experiments show that the model has higher prediction accuracy and generalization ability.

In this paper, a method for modeling and simulation of pneumatic transients of natural gas pipelines is proposed. The partial differential equations (PDEs) that describe the dynamics of the gas pipeline are converted to the ordinary differential equations (ODEs) based on the difference approximation. Electrical analogies for pneumatic and thermal quantities are exploited, and the natural gas pipelines are modeled using basic electric circuit elements. The pipeline model can be simulated with any available power-system and circuit simulation tools. Detailed studies are performed to demonstrate the accuracy and effectiveness of the proposed equivalent circuit model.

Solid oxide fuel cell (SOFC) has received considerable attention nowadays. It can be used in the integrated energy system (IES) to provide both thermal power and electric power. For purposes of analysis of multi-physics phenomena, a SOFC model that considers electric, pneumatic, and thermal transients is developed in this paper. Electrical analogies for pneumatic and thermal quantities are exploited. Equivalent circuits are used to represent the dynamic performance of electric, pneumatic, and thermal transients inside the SOFC. The equivalent circuit model is implemented in Matlab/Simulink/SimPowerSystems. Case studies involving3 W and 5 kW SOFCs are conducted to validate the proposed model.

Based on consumer psychology, a price-based demand response load model is established, which divides the load into two parts: the basic load and the response load. The basic load uses the distance space method to filter the load of the five characteristic quantities of the week, the holiday, the season, the date difference, and the weather from the historical load. The response load is modeled based on user psychology, and the sigmoid function is used to describe the psychological change process of the user in the response process, and the load transfer rate is represented by a triangular fuzzy number. The simulation of the actual data in the United States shows the effectiveness of the proposed method.

In order to diagnose the electric shock fault types of new-generation adaptive residual current operated protective devices (RCDs), a combination method based on local mean decomposition (LMD) amount of energy and support vector machine (SVM) is proposed in this study. First, LMD is applied to decompose the original electric shock signals into a finite number of product functions (PFs); then the former PFs containing the main features are selected to calculate the total energy. Second, SVM is used to solve the multi-class classification problem. The amount of energy of the PF component is considered as the input vector of the SVM, and particle swarm optimization (PSO) is used to optimize the model parameters. Experimental results show that the proposed method is effective in diagnosing animal or plant electric shock fault type, which provides a reliable theoretical basis for developing a new generation of adaptive RCDs. © 2017 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

This study proposes a novel protection scheme for residual current devices (RCDs) based on the theory of skewness and the least squares-support vector machine (LS-SVM). The skewness of the residual current is applied to detect the electric shock fault time, and the LS-SVM regression technique is used to perform touch current identification. The experimental test results indicate that the protection scheme can not only detect the electric shock fault time correctly and rapidly but also effectively recognise the circulation through the animal's current. Moreover, the proposed protection technique can provide valuable information for developing new generations of RCDs.

The development of rural energy is directly related to the process of China's energy reform. At present, the energy structure of villages and towns in China is still unreasonable. It is urgent to build a rural energy station integrating the characteristics of rural industrial development to meet the demand of production and living load. In this paper, considering that the rural areas mainly rely on natural production economy and have the advantages of clean energy endowment, through the establishment of a rural comprehensive energy system containing biogas and photovoltaic, the model seeks to maximize the comprehensive benefits including investment cost and expected income, optimize the allocation of energy device capacity, and verify its feasibility with practical cases, so as to provide reference for future rural comprehensive energy planning and green industry Economic upgrading provides theoretical and technical support.

In order to solve the problem of “disposal of light” in distributed photovoltaic power plants,this paper proposed that the energy supply of photovoltaic power plants could achieve local consumption in agricultural greenhouses, and predict the photovoltaic power generation in advance according to the load demand of greenhouses. The historical power data and meteorological data of the input variables were normalized, and the PV power prediction was performed at a time interval of 1 h for a time comparison. The prediction method based on the iterative improved BP neural network model was proposed. The prediction results showed that the prediction error was not More than 2%,could meet the energy supply stability requirements of greenhouse loads.

Aiming at the problem of unstable grid connection of photovoltaic power source, an in-situ consumption strategy was proposed, and a photovoltaic energy and phase change energy storage method was proposed to provide energy for greenhouse loading. Design an agricultural greenhouse power supply system with complementary solar energy and phase change thermal storage materials, including constant temperature control center, detection circuit, input circuit and display circuit. It could be seen through software programming and simulation analysis that phase change and photovoltaic combined energy supply could to meet the load demand of agricultural greenhouses, it provided a reference for the local consumption of photovoltaic energy.

In view of the fact that current economic management of distribution network is mostly confined to the study of economic evaluation index and lacks some means to carry out in-depth analysis and excavation of the key influencing factors of distribution network economic operation, this paper presents a multidimensional data analysis distribution network economic operation practical method. By establishing a multidimensional analysis model of economic operation of distribution network, this paper comprehensively analyzes historic economic data of distribution network and various factors from multi-aspects, multi-perspectives and multi-levels to deeply mining the potential of energy-saving and loss reduction of distribution network.

High proportion of renewable energy was increased in the distribution network, which will face complex grid complex test and show the scale the diversity and complexity of the operating state. So that the evolution of Wind-photovoltaic-energy storage(Wind-PV-ES) hybrid network transient process is not clear. In the other hand, Transient simulation tool plays a significant role in obtaining a full and detailed response under various operation conditions. However, the simulation capability and speed become limited with increased system size. In order to study the background of high proportion of Wind-PV-ES in the distribution network, this paper model Wind-PV-ES and its control system in the multi-CPU cluster fine-grained real-time simulation platform. The control system was further decoupled according to its topology. The multi-CPU can be easily realized, and the computation loads are balanced by DG modeling. The model is used to calculate the load balancing as the optimal condition to be decoupled by the distribution parameters. Parallel computing to ensure the accuracy of the transient simulation results and convergence effect and improve the simulation speed. Then this paper used two typical examples to realize the verification of Multi-CPU simulation results. The simulation results show that the simulation accuracy decreases with the increase of model computation, but the simulation speed is greatly improved. The result reflects the characteristics and transient characteristics of grid -connected transient operation. The effectiveness of the proposed method is validated in integrated simulation of distributed generation system.

For the development of residual current protection device in low voltage network, aiming at complexity and uncertainty of residual current spectrum in the transient process of biological electric shock faults, the Hilbert-Huang transform is applied to analyze spectrum characteristics of each intrinsic mode function (IMF) component in residual current and determine the distribution of each IMF at different time and frequency. Based on Hilbert — Huang transform, a method for characterizing the effective information of residual current signal performance is proposed. The study results show that the effective frequency in residual current is mainly below 1 KHZ and the correlation of low-frequency IMF component amplitude can be up to 0.99 above which can provide theoretical basis and technical support for the development of residual current protective device acting based on branch current signal when the human body electric shock occurs.

Modeling of a permanent magnet synchronous generator (PMSG)-based wind energy conversion system is presented for the simulation of diverse transients. In order to perform multi-scale transients, the back-to-back voltage source converter (VSC) is modeled using three different forms including the detailed, switched and average models. The PMSG-based WECS is implemented in PSCAD/EMTDC. The simulation results show that the detailed and switched model of VSC give a detailed and accurate representation, while the average model provides an efficient simulation.

The electrical shock fault signal is very important to the protection and control of power system. However, in practical cases, the signal is usually corrupted by artifacts through the recording process. Thus, denoising of this type of electric shock fault signals seems necessary. In this paper, a novel method for denoising electriccal shock fault signals is proposed based on Empirical mode Decomposition-Thresholding (EMD-T). With the new method, the nonlinear and non-stationary fault signal is decomposed into intrinsic mode functions (IMFs) via EMD, then, mixed modes will be processed by threshold denoising approach. The efficiency of the method is applied on real electric shock fault signals and white Gaussian noise added electrical shock fault signals obtained from Residual Current Operated Protective Device experiment platform. Signal to Noise Ratio (SNR) and Root Mean Square Error (MSE) are used to measure and compare the performance of proposed method with traditional FIR based method. Results show that EMD-T algorithm performs better than conventional FIR filter method.

In order to solve the refusing action and maloperation problem of electric shock of living organisms of Residual Current operated protective Device (RCD) in low voltage distribution network, an electric shock impedance modeling method is proposed in this paper. Trough building up the experiment test bed of electric shock of living organisms, electric shock voltage and current data are collected. After filtering the original electric shock signal, the electric shock voltage and current signal are transformed from time domain to frequency domain by using Fast Fourier Transform. Electric shock voltage and current power frequency component can be extracted. The total electric impedance amplitude and phase values of the living organisms are from Ohm law. Based on the group of 30 electric shock impedance values, using data fitting method to calculate the value of the living organisms electric impedance amplitude and phase. Then, the digital simulation model is set up by using Simulink of MATLAB. Through validating of the impedance modeling, the results show that the model can reproduce the transient process of electric shock organism well. Experiments and results show that the proposed modeling method can improve the operation of RCD with better accuracy and reliability.

In allusion to the randomness of output power of distributed generation (DG), a reliability evaluation model based on sequential Monte Carlo simulation (SMCS) for distribution system with DG is proposed. Operating states of the distribution system can be sampled by SMCS in chronological order thus the corresponding output power of DG can be generated. The proposed method has been tested on feeder F4 of IEEE-RBTS Bus 6. The results show that reliability evaluation of distribution system considering the uncertainty of output power of DG can be effectively implemented by SMCS.

Based on equivalent simplification of distribution network, an effective method was described to identify the faults' effect on load points' reliability accounting for islanded operation in this paper. According to the characteristics of distributed generators (DGs) and annual average load curve, the corresponding probability models can be formulated. On this basis, an improved sequential Monte Carlo simulation method was proposed to make the reliability evaluation of distribution network considering intentional islanding. The proposed method has been tested on feeder F4 of IEEE-RBTS Bus 6. The results show that the proposed method has better efficiency in reliability evaluation of distribution network.

With the development of distribution network, distributed generation such as wind and photovoltaic (PV) power will become increasingly prominent in the near future. PV is widely constructed because of advantages it has. However, the volatility and randomness of PV makes it more complex than traditional energy in the security assessment of distribution network. Based on risk theory, considering the randomness of PV, node low voltage risk index and line overload risk index are established in this paper. Also, K (N - 1 + 1) principle for distribution network which is developed from traditional (N-1) deterministic principle is applied to reflect the flexible structure of distribution network. IEEE three-feeder example system is utilized to investigate the influence of PV power on the security assessment of distribution network.

To solve the problems of the incorrect operation, the failure action and the low operational percentage on Residual Current Operated Devices (RCDs) in the domestic rural power distribution network, an amount of electric shock signals were demanded to study identifying and extracting method of electrical shock current. In this paper an equivalent circuit model of the human body against electric shock was studied. The frequency characteristic of let-go current according to proposed equivalent circuit model of the human body tallied with the testing results of Dalziel [1–2]. Considering the skin was in the dry condition, wet condition and broken-down condition, touch voltage at 106 V rms, electrical contact making at the peak of touch voltage respectively, the simulation results in low-voltage distribution system based on Matlab indicated the feasibility of explaining certain electrical shock phenomenon by the transient response of the equivalent circuit model. The individual responses against electric shocks were simulated by selecting parameters of the human body circuit model correctly. The results showed great potential value for the application of residual current protected devices (RCDs) of different action time as well as the research and development of new type and high reliability RCDs.

An identifying method of electric shock current amplitude based on Independent Component Analysis is proposed in this paper. The physical testing data of electrical shock current and summation leakage current, which were obtained from the physical electric shock experiment platform, were analyzed by the fast independent component analysis (ICA) model. The mean square error and correlation analysis between the extracted signal and actual results were studied, and the results showed that this method can effectively separate the electric shock current and provide technical support for the development of the residual current protective device.