Showing papers on "Blackout published in 1998"

Early prediction of electric power system blackouts by temporal machine learning

Abstract: This paper discusses the application of machine learning to the design of power system blackout prediction criteria, using a large data base of random power system scenarios generated by Monte-Carlo simulation. Each scenario is described by temporal variables and sequences of events describing the dynamics of the system as it might be observed from realtime measurements. The aim is to exploit the data base in order to derive as simple as possible rules which would allow to detect an incipient blackout early enough to prevent or mitigate it. We propose a novel “temporal tree induction” algorithm in order to exploit temporal attributes and reach a compromise between degree of anticipation and selectivity of detection rules. Tests are carried out on a data base related to voltage collapse of an existing large scale power system.

Computer experiments on Radio Blackout of a Reentry Vehicle

Abstract: We examined the radio blackout of a reentry vehicle by performing electromagnetic PIC (Particle-In-Cell) simulations. We focused on (1) blackout phenomenon and its avoidance with static magnetic field applied to the plasma, and (2) measurement of the reentry plasma layer with radio waves. Regarding (1), we could clearly show that the radio blackout occurs when the plasma frequency of reentry plasma exceeds the frequency of radio wave emitted from the vehicle. To avoid the radio blackout, we applied static magnetic field to the reentry plasma, which modifies the wave dispersion relation. We found that the radio waves can penetrate the reentry plasma mainly by whistler mode. Regarding (2), we found that the reflectometry method is applicable to the estimate of the spatial profile of the plasma layer, which utilizes the information of phase difference between the waves emitted from the antenna and reflected against the plasma layer. We can also obtain the maximum plasma density by searching for abrupt decrease in the field spectra of the reflected waves. 19 18 / 10 10 m − and it causes so called radio blackout (MaCabe and Stolwyk, 1962; Huber and Sims, 1964; Rybak and Churchill, 1971). During the blackout, radio waves emitted from the reentry vehicle cannot reach ground stations because they are reflected against the plasma layer. The theoretical analysis of the radio blackout is difficult because the reentry plasma has very steep gradient in space in a short distance compared to the wave length. Meanwhile active experiments in space provide significant data regarding the radio blackout. However, we cannot perform the experiments so often because they take a long time for the preparation and cost very much. In the present study, we analyzed the radio blackout by performing computer experiments with an electromagnetic PIC (Particle-In-Cell) code which solves the equation of motion for the plasma dynamics and Maxwell's equations for the associated electromagnetic fields at each time step. We also examined a method of blackout avoidance. Various methods concerning the blackout avoidance have been proposed mainly in the U.S. Although most of them are classified in the military and not open to the public, the simplest method is to apply magnetic field to the reentry plasma and locally modify the wave dispersion relation. We will examine the capability of this method by computer experiments. In the research of the radio blackout, we need to understand the parameters of the reentry plasma such as the density profile and the maximum density. Conventionally the density of the reentry plasma has been measured by using a probe with a help of CDF (Computational Fluid Dynamics) analysis. However, the probe data obtained in the temperature more than 500 degree is not reliable due to the breakdown of insulation. Instead of using a probe, we use radio waves to measure the reentry plasma. In the present study, we adopted the reflectometry method which is used in the fusion plasma measurement (Simonet, 1985). In the reflctometry we actively emit radio waves to plasma and obtain the reflected waves for the density measurement. The advantage is that the plasma environment is not perturbed by probe insertion. In the HYFLEX experiment (Itoh et al., 1996) this method was utilized to measure the density of the reentry plasma by using two different antennas. To examine the capability of this method in detail, we perform one-dimensional computer experiments with PIC (Particle In Cell) model. Radio Blackout and its avoidance In the computer experiments, we use an electromagnetic particle code called KEMPO which has been developed in the space plasma group of Kyoto university (Matsumoto et al., 1985). This simulation code basically solve the Maxwell equations for the fields and the equation of motion for the particles in the model region and enables us to trace not only linear but also nonlinear evolution of phenomena of interest. The model region for one-dimensional computer experiments is schematically illustrated in Figure 1. We focus on the vicinity of a reentry vehicle including a dense plasma layer. The y-z plane at x=0 corresponds to the vehicle surface. The reentry plasma is placed in front of the vehicle surface with a Gaussian distribution in space. By current oscillation at the vehicle surface, we radiate radio wave which propagates along the x direction. The electric and magnetic field components of the radiated wave oscillate in the y and z directions, respectively. In this model, we neglects the effects of the geomagnetic field and the collision frequency against neutral particles.

Influence of power system protection on system blackouts

Abstract: The important part that power system protection may play in a major power system blackout is well known. Many papers have been written to describe particular events and to analyse the causes of failure. The following paper is based on an actual protection failure on a major transmission system. The main purpose of the paper is not to describe the event in detail, but to describe how the occurrence of this event initiated a detailed review into the transmission system protection. This review was initiated to identify any limitations of protection systems or their application and maintenance which might result in a similar event occurring in the future and to recommend actions to correct any perceived limitations. The paper provides a general overview of the significance of protection operation on power system performance, provides a brief description of an actual event and gives details of the protection review which followed. The need to record both correct and incorrect protection operation events in order to be able to assess current and potential and future protection performance is also discussed.

An intelligent training agent for power system restoration

Abstract: Total blackouts, though rare in a modern power system, could cause huge direct and indirect financial losses. Power system operators should be trained to restore their system from a total or a partial blackout condition in a relatively short period of time. A case-based reasoning approach has been utilized to develop a training simulator for the SaskPower network. System operators can use the simulator in an interactive manner to simulate a restoration process. The simulator can guide system operators through the steps of a restoration process and displays the outcome(s) of a restoration action in terms of system states. The interaction between the simulator and an operator has been achieved through an object-oriented graphical interface.

Implementation of an on-line intelligent system to prevent a system blackout

Abstract: In this paper, an approach for assisting operators to prevent power system blackouts during emergency conditions is presented. It is based on expert system and fuzzy logic techniques to summarize the operating experience of power system dispatchers. The Fault Diagnosis and Restoration Control (FDARC) program has been developed using the C++ language and object-oriented programming technology, on a PC-486 which is integrated with a VAX computer via a serial link. It has been implemented in the East China Power System Control Center for trial operation since March 1998.