Search

This paper presents the optimal scheduling of active distribution network (ADN) with linear and non‐linear loads in the presence of an integrated unit of electric spring and electric vehicle parking lot (IUEE). The proposed scheme is formulated in the form of three‐objective optimization, which minimizes the expected operating cost of the ADN, voltage deviation function (VDF), and the total harmonic distortion (THD) of the voltage in separate objective functions. Constraints include harmonic power flow (HPA) equations, operation and power quality limits, and operation model of the IUEE. This scheme has uncertainties of load, energy price, and energy demand of EVs, which are modelled here using stochastic programming. Next, the sum of the weighted functions method as a subset of the Pareto optimization technique is adopted to extract the integrated single‐objective problem. Then, Sine‐Cosine Combined Algorithm (SCA) and Gray Wolf Optimization (GWO) are used to achieve an optimal solution. In the end, by examining the numerical results obtained from the implementation of the proposed scheme on a 69‐bus ADN, the capability of the design in enhancing the economic situation, operation, and power quality of the ADN compared to network power flow studies obtained by optimal scheduling of the IUEE is confirmed. [ABSTRACT FROM AUTHOR]

With the emergence of smart grids, accurate very short-term load forecasting (VSTLF) has become a crucial tool for competitive energy markets. The number of behind-the-meter photovoltaic solar panels, which usually are not monitored are increasing. This could reduce the load visibility and also affects the VSTLF accuracy. While most of the research works focus on demand forecasting of large areas, the performance of VSTLF methods for the future scenario of high solar power penetration in the microgrid environment is unclear. This paper investigates the impact of high solar power integration on the forecasting accuracy of machine learning VSTLF models for microgrids. The performance of Neural Network, Support Vector Regression are evaluated for high penetration scenario and no solar penetration. The accuracy of these models is examined for three different forecasting horizons through various comparative experiments for two real-world microgrid datasets. According to simulation results, using the most relevant variables is highly recommended. Furthermore, the results demonstrate that machine-learning methods can tackle the nonlinear characteristics of net load forecasting as well as total load forecasting. These simulations show that high penetration of solar power generation could not significantly affect the accuracy of machine learning VSTLF if most relevant variables are selected and applied as model inputs. [ABSTRACT FROM AUTHOR]

This paper presents a fault detection and faulty area identification algorithm in double-circuit transmission lines (DCTLs) compensated by the generalized unified power flow controller (GUPFC) for the first time. The topology of the GUPFC-compensated DCTL is threatened by three types of faults: single-circuit faults, inter-circuit faults, and grounded double-circuit cross-country faults. In the first step, the suggested scheme can determine the time of fault in this type of topology using a very fast method that is based on the analysis of a current index called the current jump index. In the second step, the fault type and area are identified with the help of two other indices, i.e., voltage difference index and current difference index. The test network presented in this paper, which includes a DCTL and a GUPFC, along with the proposed algorithm, is implemented in the MATLAB software package. The results of the simulation and sensitivity analysis of the algorithm confirm the correct performance of the algorithm. [ABSTRACT FROM AUTHOR]

There are many types of distributed-generation (DG) sources that cannot be connected to the network directly and use an inverter-based interface. In harmonic compensators [like active power filters (APFs)], the coupling transformer disrupts the compensation capability, therefore the transformerless inverter schemes have grown up in recent years. A new cascaded multilevel transformerless inverter topology is introduced in this study to connect the DG sources to power utility. The proposed circuit of this study can compensate the current harmonic of the non-linear load (as in APF) and inject maximum possible active power as a DG source interface, simultaneously. A fixed gain called KC is obtained based on the load nature, and is used to adjust the set point of the inverter between two operating modes. Proposed scheme has been simulated in Matlab/Simulink to evaluate the circuit performance both in the maximum active power injection mode and the load harmonic compensation mode. Then a 2.2 kW single-phase prototype of the circuit is used for experimental evaluation of the study. Both simulative and experimental results prove that such a circuit minimises the total harmonic distortion of the source side current to an acceptable margin, while injecting the maximum possible active power. [ABSTRACT FROM AUTHOR]

Highlights: [•] Introduce the εv-MOGA to solve emission, reserve and ELD problem. [•] Form the εv-MOGA based on the idea of ε-dominance. [•] Consider the valve-point, prohibited zones, frequency, and practical constraints. [•] Increase the social welfare for consumers and GENCOs by frequency constraints. [•] Illustrate the more speed and effectiveness of the εv-MOGA than other algorithm. [Copyright &y& Elsevier]

Deregulation of electric power market and aggregation of renewable resources raise the need for new hour-ahead load forecasting models. This paper proposes a new hybrid data-driven method for hour-ahead electrical load forecasting based on innovative features that represents the nonlinear and dynamic characteristics of electrical load. These features predict hourly load changes and improve the accuracy and performance of STLF. These innovative features first construct the pool of features along with historical load variables. Then, a feature selection method called RReliefF is used for choosing most relevant features and finally, a multi-layer perceptron neural network is employed as a forecasting engine—due to its advantages such as self-organization, fault tolerance and ease of integration in existing technologies. The efficiency of the proposed model is evaluated through various comparative experiments and compared with benchmark models using the three years' real energy market data from New England ISO by four evaluation criteria. The results demonstrate the superiority of proposed method in forecasting performance for the period of analysis including 12 test months as well as special days. • Proposes a STLF method based on innovative features, feature selection and NN. • Innovative features efficiently identify the dynamic characteristics of load data. • The model is also suitable for STLF in holidays and special days. • The results clearly confirm the accuracy of method using four evaluation criteria. [ABSTRACT FROM AUTHOR]