COMPUTATIONAL INTELLIGENCE IN ELECTRICAL ENGINEERING (INTELLIGENT SYSTEMS IN ELECTRICAL ENGINEERING)
Year:2018 | Volume:8 | Issue:4 |Papers Count:6

In this paper the modified grey models are proposed for short-term load forecasting in presence of demand response. The demand response is a valuable element of the smart grids. On the other hand, short-term load forecasting is very important for energy purchase planning and optimal operating of restructured power systems. Since the consumer participation is undetermined and inherently uncertain, the load forecasting will be a difficult task in presence of demand response. Numerous methods have been proposed to the load forecasting, which they have little ability to track demand side reaction. Therefore, requirement of a high precision method to model and predict the electric load in presence of demand response is appreciable. In this paper, grey models, which utilize low number of data to high precision prediction, have been modified by an iterative strategy to short-term load forecasting in presence of demand response. Since grey models are local predictors, they show better ability in modeling and forecasting of the load profiles with the unexpected and sudden changes. After applying demand response scenarios on Iran consumption load data, they have been utilized to verify the proposed method. Simulation results show high performance and accuracy of the proposed methods.

The production of electric energy for power systems with the goal of minimizing the total production cost for existing active units in the power network is one of the most important issues of modern systems. In other words, the purpose of economic dispatching is proper and optimized planning for production units, by taking into account the existing nonlinear factors and limitations in the power network and manufacturing units. The issue of economic dispatch is a challenging, non-linear, and non-convex optimization problem, which due to its intricate characteristics, heuristic algorithms are utilized as the resolution. In this paper, the issue of economic dispatch has become an optimization issue considering non-linear constraints and it has been solved using learning backtracking search algorithm (LBSA). The proposed algorithm is hybrid of backtracking search algorithm (BSA) and teaching-learning based optimization (TLBO). In order to evaluate the efficiency of the proposed algorithm, two test systems are used as case studies and the obtained results are compared to that of other algorithms in the literature. Based on numerical results, the LBSA algorithm is capable of offering better solutions and, in some cases, solutions identical to other reported methods regarding the fuel cost.

High voltage direct current (HVDC) transmission lines can be used to transfer bulk power over long distances. Accurate estimation of fault location in these transmission lines is very essential to speed up the maintenance operations. This paper presents a new approach for intelligent fault locating in bipolar HVDC transmission lines using the pattern recognition techniques and the machine learning algorithms. In the proposed approach, using a combination of Prony analysis (PA) and the singular value decomposition (SVD), some useful features are extracted from the post-fault voltage signals measured at one the line terminals. Then, a pre-trained generalized regression neural network (GRNN) receives the extracted features and estimates the corresponding fault location. Tests conducted on a sample bipolar system show that the proposed fault locator has accurate and appropriate performance despite changes in fault location, fault resistance, and pre-fault current. The obtained average percentage fault location errors for the positive-pole-to-ground (PG), positive-poleto-negative-pole (PN), and positive-pole-to-negative-pole-to-ground (PNG) faults in the sample system under study are 0.264%, 0.287%, and 0.225%, respectively.

In this paper, a multi-loop hierarchical controller for a renewable DG based microgrid system is proposed in order to control the voltage of point of common coupling and also to perform accurate active and reactive power sharing. Inner current and voltage loops that are designed based on the dynamic model of microgrid system are used to obtain appropriate switching functions for interfaced inverters and the reference values for currents of DG units. Also, a modified droop controller is introduced to enhance the performance of designed voltage control loop. In order to improve the dynamic operation of the proposed controller in supplying demanded power for harmonically distorted loads, the extraction of harmonic components of PCC voltages is accomplished. Upper and lower limits of voltage amplitude and frequency droops are analyzed using capacity curves of each DG unit. The effect of renewable power generation uncertainty on DC link voltage variations are also compensated using an additional control loop in hierarchical structure. The main contributions of this paper are the above-mentioned multi-objective control system along with a droop control based on capacity curves. The MATLAB/SIMULINK simulation shows a proper steady state and transient performance during changes in generation and nonlinear harmonically distorted loads.

For the most relay networks with fading channels, it is practicable to devise communication schemes that are optimal in terms of diversity multiplexing trade-off (DMT). Characterizing the DMT of a general n-relay is an ongoing challenging problem. In this paper, we show that to compute DMT of a diamond relay channel one has to solve an optimization problem. Direct computation of DMT of a diamond network with n relays requires solving 2n optimization problems (an optimization for each cut in the network). Solving that many optimization problems in networks with many relays is not practical. We demonstrate that for any n -relay full-duplex diamond channel, if all the exponential orders of the average signal to ratios of links in the network are greater than the multiplexing gain, then computing DMT of such network is equivalent to minimizing a sub modular function and can be carried out in polynomial time.

Reactive power management is essential for transferring real power and supporting power network security. Therefore, it is important to present a correct and possible method for pricing of reactive power in electricity markets. Optimization and reactive power dispatch is performed by solving continuous variables problems (i.e., generator busbars voltages), discrete variables such as transformer tap-changers, and the size of the parallel switching capacitors in power systems. In this paper, optimization of reactive power dispatch has been performed with combined method of average fuzzy of clustering algorithm and opposition based gravitational search algorithm. Moreover, in line with the aim of minimizing the power loss, total voltage deviation and improving voltage stability criteria have been carried out.Also, the potential of the proposed method and their effectiveness for solving reactive power dispatch optimization problems in power systems have shown in this study. Optimization results show that the combined method of average fuzzy clustering algorithm and opposition based gravitational search algorithm have improved parameters such as convergence time, voltage stability criteria, absolute value of total voltage deviation, and the total loss of the transmission lines.