مجله مهندسی برق و الکترونیک ایران
سال چهاردهم شماره 1 (بهار 1396)

A new PMU-based fault detection/location algorithm for multi-terminal transmission lines is proposed in this paper, which works on the basis of synchronized voltage and current phasors received from PMUs installed in various terminals. The Clark transform (for transposed transmission lines), Eigen-values and eigenvectors theory (for un-transposed ones) are used to decouple 3-phase differential equations and to calculate the line parameters, by applying aerial modes (including α and β) to this issue. Distributed-parameters frequency-dependent transmission line model (J-Marti) is used to implement and simulate this algorithm.
The proposed algorithm consists of three main steps, as: 1) Fault detection, 2) Faulty section identification and 3) Fault location determination. The fault detection is done using an index based on differential equations response. The faulty section is identified by comparing the calculated teed point voltages.
Multi-Terminal Transmission Line (MTTL) is reduced to a simple two-terminal faulty line and the fault location is determined by four methods, by considering differential equation response dependency to time and distance. In the first and second methods, two indices are used, which their values are equal to the fault location during faulty condition. In the third and fourth methods, two new indices are defined based on the voltage and current phasors (∆V and ∑I), which the fault location is monitored by using these indices variations. The value of ∆V is minimized in the faulty point, while ∑I is maximized at that point.
To evaluate the proposed algorithm, extensive simulations are performed on a transposed three-terminal double circuit line and an un-transposed five-terminal double circuit line (a section of Iran power transmission grid), while the maximum error of less than 1% (and less than 3% for specific conditions) is achieved. The MTTL cases are simulated in EMTP software and the obtained data are processed in MATLAB software. All effective factors including power sources with different phase angles and various impedances, zero sequence mutual coupling, un-transposed construction, unbalanced loads are regarded to simulate phase to phase, phase to ground, cross-country and simultaneous faults. Also (n-1) PMU contingency is considered in these studies. Finally, the results and the proper indices are introduced by regarding a series of simulations and comparisons.

Power distribution feeders contains many laterals, sub-laterals, different number of circuit line, load taps, balanced and unbalanced load, different type of lines (cable or overhead), different cross-sections and numerous distribution transformers with different capacities. Moreover, the voltage and current are measured only at the beginning of the feeder. Thus, fault location is very complex. Because of lack of public electrical MV line corridors and load increasing cause to increase the double circuit lines. Consequently, locating faults in a double circuit line need to special fault-location algorithms. In this article a new method for fault location in double circuited MV power distribution lines is proposed. In this method the impedance based method is improved for locating fault in double circuit distribution networks using the precise line model. This method is tested in a 13 bus IEEE network in different conditions such as various fault resistances and different fault start angles in various distances. The results from the Matlab’s simulation show high accuracy and validity of the proposed method.

Integration of distributed generation (DG) into distribution networks causes some protection challenges, which can be solved by appropriate coordination of directional overcurrent relays. The optimal coordination of directional overcurrent relays is an important task to fulfill the protection requirements of the network. This paper presents an optimal coordination method for directional overcurrent relays in the presence of synchronous DGs. The optimization problem is solved by using Harmony Search algorithm to obtain optimal protection settings. Then, the transient stability of synchronous DGs is studied by modeling DGs excitation system according to the clearance time of the network faults. Instantaneous setting of the relays is used in case of the generators instability. The simulation results on an MV network using DIgSILENT Power Factory show that the transient stability of synchronous DGs plays an important role in the coordination of overcurrent relays. The results validate the effectiveness of the proposed method for coordination of overcurrent relays when considering the transient stability of synchronous DGs.

In the recent years use of Special Protection Schemes due to impact of notable these schemes in improving the performance of power systems has found increasing usage. The purpose of this paper is design a suitable and fast special protection schemes to deal with extreme contingency that will led to collapse of power network if they occurred. So at the first, for a clear form of power transmission paths a graph model of Fars region was found. Then using propose algorithm in this paper extreme contingency in power transmission network by DigSILENT program analysis and identified. Finally after identifying these events, generation rejection and load rejection schemes as the most popular and most common schemes in power networks around the world with transmission switching as new special protection scheme proposed in this paper to deal with these contingencies is presented. The results of implementation these schemes show acceptable performance of them to deal with events and contingencies identified.

In this paper, a new method is proposed for fault location estimation in overhead-cable multi-section transmission lines. This method takes advantage of synchronized samples gathered from both terminals of the line in the time domain. Because of this, it does not need a filter to extract fundamental frequency of the voltages and currents. Thus, it shows a higher accuracy, since it is not influenced by frequency response of the filter and decaying DC component. On the other hand, this algorithm uses distributed parameter line model and consists of K stages for fault location determination (K: the number of the line sections), and unlike the other methods, it does not require a recognizer to determine the faulty section of the line. The accuracy of the algorithm has been investigated by various tests, which show the high accuracy of the proposed method while different fault types occur on various locations with a variety of incidence angles.

In this paper, a new method has been presented for differential protection of power transformers. In this approach, chaotic method has been used for distinguishing of inrush current from internal fault current of the transformer. The proposed algorithm has been designed based on analysis of harmonic components of the differential current of the transformer due to internal faults and inrush currents in a specific frequency band. To verify the capability of the proposed algorithm a 230/63 kV power transformer connected to the 230 kV power system has been simulated at EMTP/ATP software under various internal faults and switching situations. The obtained information has been used in MATLAB/SIMULINK In order to apply the chaotic theory for recognizing inrush current from internal fault current. Simulation results show successful operation of the proposed algorithm.

Transformer is one of the most important components of power system, which determining its reliability is very important. To evaluate transformer overall reliability, in addition to probability of failures occur due to chance events, the probability of failures owing to aging should be taken into account. To determine transformer reliability in aging area, estimating its remaining life is of great importance. One of the parameters that affects on transformer mean lifetime is its hot spot temperature, that is a function of transformer loading. Effects of loading on transformer reliability is modeled in this paper. At first transformer reliability in useful lifetime interval is discussed to be used in transformer overall reliability formulation. Using thermal modeling and aging acceleration factor, impacts of transformer loading on variation of transformer mean lifetime is formulated. Then using calculated mean lifetime and by using normal distribution function, transformer reliability in aging area is modeled. Electric and thermal characteristics data and also hourly load data of a real transformer in one year are applied to the model. Using them, impacts of various loading strategy and age increasing on transformer reliability in wearout area as well as transformer overall reliability are simulated and analyzed. The results show the significant impact of loading on transformer reliability and can be used for on-line monitoring of transformers.

Parameters estimation of dynamic model of synchronous generators is an essential prerequisite for viable and reliable power system operation and offline studies. Advent of phasor measurement units (PMU) and its growing applications in power systems have led to an evolution in the electric power network wide-area monitoring, protection, and control. Alongside, these devices have revealed unprecedented potentials in offline studies such as parameter estimation of various equipment and event analysis. This paper focuses on parameters estimation of classical model of synchronous generators using PMU data. Accordingly, a new method for estimation of transient reactance of classical model of generator is proposed and electromechanical parameters of this generator are estimated with two techniques of recursive least square and Kalman filter. Electromechanical parameters include inertia constant, damping coefficient and mechanical power. A single generator connected to an infinite bus is simulated in PSAT and parameters of generator are estimated employing the proposed method. The impact of noise-polluted input data on the effectiveness of the new techniques is examined. The results obtained from two mentioned methods are compared under various circumstances.

In this paper, frequency response analysis (FRA) of transformer based on the detailed model of the transformer windings, is investigated. The determination of the detailed model parameters, which can be considered as a large scale model, is a complex task. To overcome this problem, a heuristic method is presented in this paper. This method is applied to a large scale detailed model with nineteen sections.The state equation model is used, to extract the transfer function of the large scale detail model of the transformer. In this paper, all parameters of the model are assumed to be frequency dependent. Also, a new improved detailed model is presented. It is shown that this model can represent the transient behavior of the transformer winding better than the well-known detailed model. In the laboratory, all new ideas presented in this paper, are applied to a 20/0.4-kV and 1600-kVA transformer. The comparison of simulation and measurement results show the privilege of the improved detailed model in comparison with the conventional ones.

Induction motors are so important in industry, so their protection and maintenance seem a vital issue. Continuous control of structural parameter of such motors is the way that can protect them. Appearance of a small problem in motor can change the value of structural parameter of squirrel cage induction motor, such as; resistances of stator and rotor, inductances of stator and rotor and mutual inductance. Therefore, precise estimation of these structural parameters with acceptable reliability can help to control condition of motor. In this paper, in order to estimate the structural parameters of under studied induction motor, gray box identification procedure has been used, such that using extracted data from motor (namely; values of root mean square of stator current and power factor) and meta-heuristic algorithms, which are Particle Swarm Optimization (PSO), Improved version of Particle Swarm Optimization (IPSO), Gravitational Search Algorithm (GSA), Improved version of Gravitational Search Algorithm (IGSA), Harmony Search (HS) and Simulated Annealing (SA), a model of under studied induction motor is identified. Obtained results show that meta-heuristic optimization algorithms can be a good chose for estimating the parameters of induction motor. Assessments show that implementing tradeoff between speed, accuracy and reliability based on the user’s requirement best meta-heuristic algorithm can be selected.

In the structure of direct vector control of an induction motor, there are three main sections that have important role in improvement of the induction motor drive operation. This three sections, including: 1- the estimation block of the feedback signals, 2- the block of the required PWM pulses generation for the inverter unit, and 3- control blocks (usually PI) for generating the required reference signals. This paper continues the recent researches of same authors to improve sections 1, 2 and 3. In this paper, the induction motor speed control drive is designed using intelligent controllers based on the emotional learning. By changing the rotor and stator resistances, the coefficients of the intelligent controllers are tuned automatically and therefore, the problem of tuning PI controller coefficients has been solved by the trial and error method. Consequently, the induction motor speed control drive will be robustness against the common variations of motor parameters meanwhile motor work.

Brushless doubly fed machine (BDFM) is indeed an induction machine which has two windings in the stator and a cage rotor with a special design. Due to absence of brushes and slip rings, this machine is considered for the applications needs high reliability, such as wind turbines. In this paper, stead state performance of BDFM in synchronous mode of operation, the most interesting mode of operation, is studied in more detail and, in addition to theoretical analysis, the required experimental results are presented. A new relationship is developed for calculation of the electromagnetic torque, which based on that, the share of synchronous and induction torques are determined in the synchronous mode. It is shown that in this mode, the synchronous mode has the dominant effect and hence, the induction torques could be neglected in steady state calculations, reducing calculations and facilitating analysis of results.

This paper presents a new approach for single-phase shunt active power filter with LCL output filter based on the two-loop control theory. In this method unlike the conventional two loop control, internal loop stability theorem with considering delay effect is studied. Moreover, an observer based on the Kalman-filter is combined with the proposed control scheme, to reducing the numbers of power sensors in the overall system. Finally, the performance of the proposed method is investigated by extensive simulation tests. Simulation results confirm the excellent performance of the proposed control scheme in steady state and transient conditions.