Abstract Artificial intelligence (AI) and Deep learning (DL) methods in power systems are being tested and prepared for practical use in many applications. In this work an artificial neural network models for fault identification and classification and switching logic control in middle voltage (MV) power electricity network is presented. Models are implemented in Google’s Python based tool Tensorflow with belonging program libraries. For fault detection and classification example a few thousand simulations are conducted in order to obtain enough fault current and voltage samples for high accuracy artificial neural network (ANN) with backpropagation model. Achieved accuracy and speed of presented deep learning model, open a possibility for application in digital relay protection devices. Second example is implementation of switching control rules in HV/MV substations. Presented models are patterns for power system controlling centres as part of broader controlling and protection logic.

Treatment of transformer neutral point in middle-voltage (MV) networks become an important issue with increasing proportion of MV cables in power networks. As consequence, overall capacitance of MV network is increased and moreover earth fault currents magnitudes. In MV networks with feeding transformer winding in delta connection (isolated networks), that earth fault current increase requires forming of artificial ground point – a neutral connection point on a three-phase ungrounded power system. Grounding transformer use, in zigzag or delty-wye connection, is common, well-known solution for constructing neutral connection in power systems. Physical characteristics of grounding transformers, protection principles, short-circuit calculations with symmetrical components and simulation techniques are presented in this paper. Characteristical operational modalities of MV power networks are also revieved on practical examples.

This paper presents the finite element (FE) time-domain model for numerical solution of frequency-dependent transmission lines (TLs) problem. Based on the finite element method (FEM) and the generalized trapezoidal rule applied to frequency-dependent TL equations, a novel numerical procedure for a solution of a system of the telegraph equations in time domain is presented. The principles of the proposed method, for the sake of simplicity, have been shown to the case of two-conductor TL where frequency-dependent per-unit length TL parameters have been incorporated into the FE local system of equations by the direct convolution algorithm. The proposed numerical method has been tested and compared to the solutions available in the literature, and a good correlation has been demonstrated.

SummaryThis paper presents a finite element time domain model for a numerical solution of a coupled non‐uniform transmission line problem. On the basis of the finite element method, a novel numerical procedure for the solution of a system of the non‐uniform multi‐conductor transmission line equations in the time domain is presented. The results obtained by the proposed method have been compared with the solution obtained using the finite difference time domain method, and an excellent correlation has been demonstrated. Copyright © 2014 John Wiley & Sons, Ltd.

Using of mathematical transformations and digital signal processing (DSP) methods of recorded voltage and currents in power systems in order to extract important informations about physical processes is already known. Higher measuring resolutions and massive introduction of PMU devices in the power systems allows appliance of DSP - based fault location methods in complex power networks. High availability of power systems is required in the modern societies. Fast fault location algorithms and consequent fast repairing of faults is necessary condition for high availability of power networks. Here represented numerical procedure for fault location is testing of method already introduced in scientific papers on complex configuration of power network consists of XLPE underground submarine and land cables between land and offshore substation connecting large wind park with main power system.

In the procedure of connecting large wind parks (WP) to the external power grid, TSO's or DSO's request is that WPs in operation do not excessively affect the power quality in the connection point. In the process of determining WP operating procedures it is necessary to include switching operations in the WP main feeder. The transformers connected near the wind generator units are then energized and a transformer inrush current appears, followed usually by voltage distortions. Those currents may cause nuisance tripping of protection devices and power quality problems in the WP connection point. For illustration, the connection to a real power grid of a large WP with twelve generators and block transformers is simulated. Considered network short circuit power is comparable to typical parameters of the transmission or distribution medium-voltage networks. Simulated inrush currents depend on the characteristics of connected devices in the network. Results of our study can be useful for correct protection relay device setting and determining optimal WP exploitation procedures.

The paper introduces a practical approach to power system fault location in power networks using advanced fault signal processing. The three-phase fault voltages are converted to the vector of absolute values of its complex space-phasor. This vector represents fault traveling wave and it is further processed for fault location finding with the Hilbert–Huang transform. The simulation results, including single line to ground faults, faults in mixed feeders and high-impedance arcing faults, confirm the accuracy and practical applicability of the proposed approach.

The most important parameter in estimations and calculations of technical adequacy of wind park (WP) connection is short-circuit power in the Point of Common Connection (PCC). Design and planning of wind park substations can be important factor in fulfillment of transmission or distribution system operator (TSO, DSO) grid code. Likewise, selection of high-voltage equipment such as power transformer, power underground cables or overhead lines can have a considerable effect on connection terms. Typical large WP collector substation is considered. Short-circuit power dependence from high-voltage equipment characteristics in Connection Points (CP) of WP is especially examined. Detail computer simulations are elaborated for chosen substation design. The graphs quantifying the most important wind park connection parameters in relation with chosen high voltage equipment and overhead/underground option for interconnection feeder are presented.

In overall expansion of power distribution and sub-transmission (10 kV - 110 kV) networks, conditioned with growing consumption of electrical power, there is an increasing of possibility for crossing underground power cables with home electrical installation cables, telecommunication equipment cables and other equipment. Furthermore, surge arresters on middle-voltage levels usually are not applied along overhead lines. Above mentioned can affect on equipment or even human life safety. It is known that interference of the different electrical installations and equipment is the main theme of researching in Electromagnetic Compatibility (EMC). In this article, a possibility of using the Finite Element Method (FEM) for EMC calculations in practice, is explored and presented. For that purpose, a real problem from distribution power network will be used for an illustrative example. It is worth of mentioning that FEM technique is quite a new tool in power networks calculations area.

The aim of this paper is to present a new numerical model for multi-conductor transmission line (MTL) analysis in frequency-domain. The proposed model based on the finite element method (FEM) is compared with traditional technique such as the modal method. Applying analytical solution of the telegrapher's equations we can construct the finite element local system of equations and apply standard assembly procedure. Capability of modern computational software tools with built-in numeric functions allows us to simplify necessary calculations in numerical procedure.

Uz izgradnju novih dionica autocesta u Republici Hrvatskoj izgrađeni su i novi elektroenergetski objekti (EEO) koji služe za napajanje trosila u funkciji sigurnosti prometa. Iako se ti objekti Hrvatskih autocesta d.o.o. (HAC) inkorporiraju u domicilne distribucijske mreže HEP-a i postaju njihov sastavni dio, s aspekta optimizacije, mogu se promatrati kao specifican elektroenergetski sustav (EES). Desetgodisnje iskustvo projektiranja, građenja, pogona i održavanja EEO bilo je dovoljno da se nastojanja i realizirani zahvati na optimizaciji elektroenergetike usmjere u razvoj ekspertnog sustava (ES), cija je primjena u optimizaciji elektroenergetskog sustava HAC-a. Optimizacija ukljucuje elektromagnetsku kompatibilnost, racionalizaciju i pouzdanost. ES kao dio umjetne inteligencije potpomognut racunalom obuhvaca specijalisticko znanje iz pojedinog podrucja, te modelira inteligentne elemente covjekovog rjesavanja problema: zakljucivanje, prosudbu, odlucivanje na osnovi nekada cak nepouzdanih i nepotpunih informacija, te kontrolu i tumacenje odluka. U radu su prezentirani strateski pravci razvoja ekspertnog sustava pod nazivom ExSyS-HAC.