Gerilim Çentiği Tespiti İçin Yeni Bir Algoritma Tasarımı

Abstract

Günümüzde enerji ihtiyacı her geçen gün artmaktadır. Artan ihtiyaçla birlikte enerjinin üretilmesi ve tüketilmesi konuları da önem kazanmaktadır. Teknolojinin ilerlemesiyle, geçmişte çok fazla dikkat edilmeyen ve ihmal edilebilecek boyutlarda olan bazı olaylar, bugün dikkat edilmesi ve önlem alınması gereken durumlar halini almıştır. Bu durumlar neticesinde, zaman içerisinde bir takım yeni kavramlar ortaya çıkmıştır. Bu kavramlardan biri de “Güç Kalitesi”dir. Güç kalitesi kavramı enerjinin üretim evresinden başlayarak iletimi, dağıtımı ve son kullanıcı tarafından tüketilmesine kadar geniş bir yelpazede ele alınmaktadır. Enerjinin yol aldığı her bir aşama için dünya genelinde güç kalitesini düzenleyen bazı standartlar ortaya konulmuştur. Bu standartlar enerji üretimi ve tüketimi arasında kalite bakımından bir denge kurmak üzere hazırlanmışlardır. Bu tezde güç kalitesini olumsuz yönde etkileyen olaylardan biri olan “Gerilim Çentiği” etkisini tespit etmek için bir algoritma tasarlanmıştır. Birinci bölümde tezin amacına ve neden gerilim çentiği algılama algoritması geliştirilmek istendiğine yer verilmiştir. Gerilim çentiğinin sebep olduğu olaylardan bahsedildikten sonra literatür araştırması yapılmıştır. Literatürde gerilim çentiği tespiti yapılmak üzere hangi yöntemlerin kullanıldığı ve bu yöntemlerin pozitif ve negatif yönlerinden bahsedilmiştir. Ayrıca literatürdeki çalışmaların eksik yönlerine atıfta bulunularak çalışma sonunda bu eksikler kapatılmıştır. İkinci bölümde öncelikle güç kalitesi kavramı açıklanmıştır. Ardından IEEE Std 1159-2009 ve IEEE Std 519-1992 standartları temel alınarak güç kalitesini bozan olaylar sırasıyla açıklanmıştır. Geçici olaylar, kısa süreli gerilim değişimleri, uzun süreli gerilim değişimleri, dengesizlikler, dalga şekli bozuklukları, gerilim dalgalanmaları ve frekans değişimleri tek tek ele alınmıştır. Üçüncü bölümde standartlar çerçevesinde gerilim çentiği tanımına yer verilmiştir. Bununla birlikte gerilim çentiği tespitinde kullanılacak olan kriterler irdelenerek bu kriterler hakkında açıklamalarda bulunulmuştur. Aynı zamanda kabul edilebilir sınır değerler tablo halinde verilmiştir. Daha sonra gerilim çentiği tespiti için tasarlanan algoritmanın akış diyagramı çizilmiştir. Akış diyagramında numaralandırılan işlem basamakları teker teker açıklanmıştır. Dördüncü bölümde geliştirilen algoritmanın performans analizine yönelik çalışmalar yapılmıştır. Öncelikle standartta gerilim çentiğinin oluşmasına sebep olarak gösterilen üç fazlı kontrolsüz bir doğrultucu modeli oluşturulmuştur. Ardından bu modelden elde edilen gerilim dalgası algoritmayı test etmek için kullanılmıştır. Test işleminin ardından elde edilen sonuçlar değerlendirilmiştir. Bir sonraki test işleminde karşılaştırmalı performans analizi yapılmıştır. İlk olarak standartta belirtilen kriterlerden yola çıkılıp bir algoritma oluşturularak, performans analizi harmonikli bir dalga ile gerçekleştirilmiştir. Sonucunda standarttaki kriterlerin yeterli seçiciliğe sahip olmadığı tespit edilmiştir. Ardından standarttaki kriterlere ek olarak yeni bir kriter ortaya konulup işlem basamakları algoritmaya eklenmiştir. Daha sonra aynı harmonikli dalga kullanılarak performans analizi işlemi tekrar edilmiştir. Sonucunda çentik etkili dalga ile harmonikli dalga birbirinden başarı ile ayırt edilebilmiştir. Bir sonraki kısımda grafiksel olarak çenik etkisine çok benzeyen bir dalga ile algoritma performansı karşılaştırmalı olarak test edilmiştir. Test işleminin sonucunda çentik etkisine benzer özellikler gösteren harmonikli dalga çentik etkisinden başarılı bir şekilde ayırt edilmiştir. Bir sonraki performans analizinde darbe işareti kullanılarak algoritma test işlemi gerçekleştirilmiştir. Bu test işlemi sonucunda da hedeflenen sonuç elde edilmiştir. Ardından algoritma şebekenin sinüzoidal formunu bozmayan gerilim düşmesi (voltage sag) olayı ile test edilmiştir. Sonuç olarak algoritma bu testten de başarılı bir şekilde geçmiştir. Bir sonraki performans analizi güç kalitesini bozucu etkilerden biri olan gerilim şişmesi (voltage swell) olayı ile gerçekleştirilmiştir. Algoritma bu performans analizinden de başarılı bir şekilde geçerek ele alınan bütün bozucu etkileri çentik etkisinden ayırt etmede olumlu sonuçlar vermiştir.

At the present time, the need of electrical energy has been increasing. With this rising need, the subjects of being produced and being consumed of energy has gained importance. By development of technology; some events like transients, RMS variations, unbalance conditions, waveform distortions, voltage fluctuations and power frequency variations that were unimportantly observed and could be accepted as neglected events in the past have today become the situations to be considered and taken precautions on. As a result of those cases mentioned above, in time, some new concepts have come out and one of these concepts is called ‘’Power Quality’’. The concept of power quality has been handled within the large spectrum of starting from the production stage of electricity to its transmission, distribution and also being consumed by last user. For each process of progressing energy, some worldwide standards which regulate the power quality have been presented. Those standards have been prepared to harden between the producing and consuming electrical energy in terms of power quality. In this thesis, an algorithm has been proposed in order to detect the ‘’Voltage Notch’’ which is one of the events that affects power quality. In the first section of the thesis included objectives of the thesis, a literature review is carried out for power quality and voltage notch detection. The literature survey covers present techniques used to detect voltage notch. The advantages and disadvantages of the present techniques in detecting voltage notch are also discussed. Notching is a periodic voltage disturbance caused by the normal operation of power electronics devices when current is commutated from one phase to another. During this period, there is a momentary short circuit between two phases. Voltage notching represents a special case that is periodic, yet has frequency content that is quite high. Thus it has attributes that could be considered both transients and harmonic distortion. Since notching occurs continuously (steady-state), it can be characterized through the harmonic spectrum of the affected voltage. However, the frequency components associated with notching can be quite high and may not be readily characterized with measurement equipment normally used for harmonic analysis. Three-phase converters are the most important case of voltage notching. In the second section, primarily the concept of “Power Quality” was explained. Then, the events distorting Power Quality and included in IEEE Std 1159-2009 and IEEE Std 519-1999 standards were discussed in detail. These events are; transients, short-duration RMS variations, long-duration RMS variations, imbalanced, waveform distortions, voltage fluctuations and power frequency variations. Transient simply states that it is the “part of the change in a variable that disappears during transition from one steady-state operating condition to another. Broadly speaking, transients can be classified into two categories, impulsive and oscillatory. These terms reflect the wave-shape of a current or voltage transient. An impulsive transient is a sudden, non-power frequency change from the nominal condition of voltage, current, or both, that is unidirectional in polarity. An oscillatory transient is a sudden, non-power frequency change in the steady-state condition of voltage, current, or both, that includes both positive and negative polarity values. Short-duration RMS variations category encompasses the IEC category of voltage dips and short interruptions as well as the antithesis of dip: the swell. Short-duration voltage variations are almost always caused by fault conditions, the energization of large loads which require high starting currents, or intermittent loose connections in power wiring. Long-duration variations encompass rms deviations at power frequencies for longer than 1 min. Long-duration variations can be either overvoltages or undervoltages, depending on the cause of the variation. Overvoltages and undervoltages generally are not the result of system faults. They are caused by load variations on the system and system switching operations. Imbalance (sometimes called unbalance) in a three-phase system is defined as the ratio of the magnitude of the negative sequence component to the magnitude of the positive sequence component, expressed as a percentage. This definition can be applied for either voltage or current. Typically, the voltage imbalance of a three-phase service is less than 3%. The current imbalance can be considerably higher, especially when single-phase loads are present. Voltage fluctuations are systematic variations of the voltage envelope or a series of random voltage changes. Such voltage fluctuations can be perceived by humans by changes in lamp illumination intensity. To measure these voltage fluctuations, the functional and design specifications of flickermeter are given in IEC 61000-4-15. Power frequency variations are the deviation of the power system fundamental frequency from its specified nominal value (e.g., 50 Hz, 60 Hz). The steady-state power system frequency is directly related to the rotational speed of the generators on the system. At any instant, the frequency depends on the balance between the load and the capacity of the available generation. When this dynamic balance changes, small changes in frequency occur. Waveform distortion is defined as a steady-state deviation from an ideal power frequency sinusoid principally characterized by the spectral content of the deviation. There are five primary types of waveform distortion: DC offset, harmonics, interharmonics, notching and noise. In the third section, the definition of ‘’Voltage Notch’’ took place firstly as in the related standards. Secondly the weakness of the notch detection algorithm included in the standard is discussed. The present algorithm consists of three criteria as follows. Criteria 1: The average depth of the line voltage notch from the sine wave of voltage. Criteria 2: The area of the line voltage notch. It is the product of the notch depth, in volts, times the width of the notch measured in microseconds. Criteria 3: The ratio of the root-mean-square of the harmonic content to the root-mean-square value of the fundamental quantity, expressed as a percent of the fundamental.  The weakness in the present algorithm stated in the standard was about the discrimination between detection of the harmonic related events and voltage notches. A new criterion was added to present algorithm to overcome the weakness. The new criterion use the derivative of the voltage signal, calculates the RMS value of the derivative and compare the obtained with a pre-defined value. This pre-defined value was obtained using several experimental tests. The new criterion provides an ability to distinguish between voltage notch and harmonic related events. Thus the proposed algorithm consists of four different criteria as stated follows, while the standard has a technique with three criteria. Criteria 0: The derivative of the distorting signal which is obtained by superposition principle will be calculated. Then RMS of this signal will be calculated and normalized by the mains voltage. Lastly, the result will be compared with the setting value of 1.5 which is obtained by previous experiences. With this comparison, application of the rest of the criteria will be determined. Criteria 1: The average depth of the line voltage notch from the sine wave of voltage. Criteria 2: The area of the line voltage notch. It is the product of the notch depth, in volts, times the width of the notch measured in microseconds. Criteria 3: The ratio of the root-mean-square of the harmonic content to the root-mean-square value of the fundamental quantity, expressed as a percent of the fundamental. Finally, the flow chart of the proposed algorithm was also given step by step in section 3. In the fourth section, the computer simulation studies required to analyze the performance of the new algorithm in detecting voltage notch and distinguishing between voltage notch and other power quality events. For this purposes a three phase uncontrolled rectifier model was developed using Matlab-Simulink to produce voltage waveform corrupted with voltage notches. Then these types of waveforms were used to test the performance of the new algorithm. Computer simulation studies have shown that the new method detects the voltage notch correctly as stated in the related standards. Computer simulation studies were also extended to include the comparative performance analysis of the developed algorithm with the algorithm defined in the related standard. A voltage distortion similar to the voltage waveform distorted by a notch was produced using Matlab. This distortion was produced using different harmonic waveforms rather than using a switching event. The new algorithm did not detect this switching event as a voltage notch as it is expected. However the method in the standard detects this event as a voltage notch. The new algorithm performance was also tested using different voltage signals as corrupted by transient, sag and swell. As it is expected the new algorithm did not detect these events as a voltage notch.