**5. References**

174 Advances in Wavelet Theory and Their Applications in Engineering, Physics and Technology

obtained from DWT signal filters with the frequency range of 1 – 2 Hz. Frequency range of 1 – 2 Hz is the space of local oscillations in power system and by a simple comparison of power values of signals in this frequency range, analysed from multiple geographically distant locations , it is easy to establish the location of disturbance. From the power point of view, power values of local oscillations of signals measured/simulated closer to the disturbance will have higher energy power values compared to those distant from the location of disturbance. Furthermore, as we proceed to the higher levels of decomposition (or lower frequency ranges of filters) of chosen signals with sampling frequency of 0.1 sec, we enter the intra-area and inter-area of oscillations which can represent a real danger for electric power system, and should it be that they are not muted, can lead in a black-out. These signals make it possible to identify intra-area and inter-area oscillations, their character and how to mute them. Furthermore, by comparing these signals it is possible to obtain more information on the system's operation as a whole after disturbance (Avdakovic & Nuhanovic, 2009). In line with what has been demonstrated in the example, low frequency component of signal angle or frequency serves to estimate values *df/dt,* that is, to

Power system is a complex dynamic system exposed to constant disturbances of varying intensity. Most of these disturbances are common operator's activities, for example, swich turning on or off system elements, and such disturbances do not have a major influence on the system. However, some disturbances can cause major problems in the system, and the subsequent development of events and cascading tripping of system elements can lead to a system's collapse. One of the most severe disturbances is the outage/failure of one or more major production units, resulting in unbalance of active power in the system, that is, frequency decrease. Many factors influence whether or not the severity of frequency decrease will trigger under-frequency protection. Today, under-frequency protection is based on local measurements of state variables and provides only limited results. Their

This chapter illustrated the estimate of unbalance of active power in the power system with DTW application, provided WAMS is available. Estimate of *df/dt* value is a genuine indicator of active power unbalance, and given the oscillatory nature of signal frequency, its estimate is rather difficult. Taking into account its advantages in signal processing when compared to other techniques, WT enables direct estimate of medium value of the change of frequency of the centre of inertia, providing a complete picture about the system's operation as a whole. In this way, and provided with the complete inertia of the system, we obtain very important information about a complete unbalance of active power in the system, in a rather simple manner. In addition to this particularly important piece of information obtained from the low-frequency component of the signal angle or frequency, other levels of signal decomposition in frequency range encompassing low-frequency electromechanic oscillations provide information about the onset of some dynamic occurrence in the system, localize system disturbance, identify and define the character of intra-area and inter-area oscillations and provide insight into the system's operation after the disturbance. All of this points to a possible development of such under-frequency protective measures which will operate locally, that is, whose operation will be at (or in the vicinity of) the disturbance, in

define total forced unbalance of active power in power system.

operation is frequently unselective and affects the whole system.

**4. Conclusion** 


**9** 

 *Brazil* 

**Application of Wavelet Transform and** 

**Signals in Electric Power Systems** 

M. E. L Tostes2, S. C. F. Freire1 and L. A. Meneses1 *1Federal Institute of Technological Education, Belém, Pará* 

R. N. M. Machado1, U. H. Bezerra2,

*2Federal University of Pará, Belém, Pará* 

**Artificial Neural Network to Extract Power** 

**Quality Information from Voltage Oscillographic** 

Post-operation contingencies analysis in electrical power systems is of fundamental importance for the system secure operation, and also to maintain the quality of the electrical energy supplied to consumers. The electrical utilities use equipments as Digital Disturbance Registers (DDR), and Intelligent Electronics Devices (IED) for faults monitoring, and diagnosis about the electrical power systems operation and protection. In general, the DDR and IED are intended to monitor the protection system performance and detect failures in equipments and transmission lines, and also generate analog and digital oscillographic

The oscillographic signals often analyzed in the post-operation centers are those generated by events that typically cause the opening of transmission lines due to the action of protective relays. So, these records are analyzed in detail to determine the causes and consequences of these occurrences within the electrical system. Although the software used in the post-operation centers presents numerous features for the evaluation of the recorded signals, the selection of the signals to be analyzed is done in a manual way, which leads to an analysis in an individual basis, and many of the oscillographic records that could help analyzing the occurrences are not evaluated due to the long time that would be spent to

Another aspect to be noted is that the oscillographic records remain stored in the post-operation centers for time periods ranging from months to years. These records contain signals acquired in different parts of the electrical system, and the vast majority of them are no longer being considered in the analysis. These data, however, may contain important information about the behavior and performance of the electrical system that may precisely characterize the power quality problem due to a failure or

registers that better characterize the disturbing events.

**1. Introduction** 

select them manually.

disturbance.

