When analyzing dynamical stability in power systems, there exist various errors. First, there are analyzing errors which occur in linearlizing non-linear systems on a particular operating condition. Secondly, there are adjusting errors of the parameter values which are the time constants and the gains of the control systems. However, the conventional representative techniques, such as the eigen value method, the Routh-Hurwitz method and the Nyquist method, cannot deal with these kinds of errors. Using these representative methods, if the parameters vary, we have to recalculate the whole system stability. In order to solve this kind of problem, this paper proposes a use of the Kharitonov theorem which is one of the robust control theories.
This paper proposes a new DC load flow method for power system contingency analysis. In recent years, more reliable operation is required to deal with complicated power systems. As a framework of security control, it is important to investigate the effect of potential faults to power systems in advance. Contingency analysis requires computational efficiency in evaluating data for controlling generating units or load flows. So far, several approaches have been developed to the contingency analysis. Among them, the DC load flow is widely used due to the computational efficiency. This paper aims at improving the computational efficiency furthermore. In particular, the indirect methods of a new approach have been applied to the process of solving a set of linear equations in the DC load flow. The conventional approaches are based on the direct method such as the LU decomposition or the Gauss elimination methods. However they have a drawback in terms of the computational time and storage. That is because fill-in elements appear in the computational process. Therefore, the indirect methods such as the CR and CGS algorithms are presented for the DC load flow of the contingency analysis. In addition, the precondition technique is introduced to the indirect methods. The proposed methods have been successfully applied to larger sample systems such as 118, 235, 352, 469 and 586 node systems.
Today's power systems are characterized in that more power demand is concentrated in urban areas, increasingly centralized power facilities are getting larger in size and located in remoter places, and power condensers tend to be upgraded in order to improve power factor. These movements each reduce the stability power limit, permitting the effect of a power failure to spread much further. In such an interconnecting power system, when power is transmitted from a power source (a small system) in a remote place to a big city (a large system) with great power demand, a zero-power flow phenomenon, where power transmission is impossible when power flow in the interconnecting power system is zero, is expected to occur, depending on a combination of the size of branch load and its power factor. In future power systems that will extend further, have a larger capacity, and be more capacitive, such a phenomenon woud be more liable to take place. In this thesis, a new formula to analyze a zero-power flow phenomenon is derived, and using the formula, areas in which a zero-power flow phenomenon takes place are also discussed. With an experimental system simulating a practical 500kV power system, the occurrence of a zero-power flow phenomenon could be verified.
According to the recent analysis results of temporary AC overvoltage in the AC system connected with a frequency converter station, large magnitude of overvoltages were confirmed to occur under some special system conditions. Most of the station insulators currently used cannot withstand such overvoltages according to our evaluation based on the data obtained before. We recognized the necessity of tests to be done in order to evaluate such performance more accurately. We made both power frequency and switching impulse overvoltage flashover tests on contaminated insulators by the method well simulating the natural wetting condition. Switching impulse flashover voltage with the wave shape having a long wave-front time of 2ms can be well correlated with the flashover voltage characteristics of temporary AC overvoltage. Higher flashover voltage characteristics were obtained by clean fog test method compared with those obtained by equivalent fog test method.
This paper describes a method of designing decentralized load frequency regulators considering the generation rate constraints for interconnected power systems. An interconnected power system is composed of a number of areas. Each area is controlled separately by a decentralized regulator. The background theory in this paper is based on servo-type optimal control theory. The design of the proposed method is consisted of the following two stages. The first stage is to get the optimal feedback gains for the decomposed area using servo-type optimal control theory. Depending on the decomposed areas, the subsystem has a possibility of uncontrollability. Then, it is necessary to decompose the power system not so as to loose the controllability. The second stage is to add the obtained optimal feedback gains to the interconnected power system. Then, the adjacent area's observed data are neglected. Because, the feedback gains belonging to the adjacent areas are generally small compared with its own area's those. The effectiveness of this proposed method is illustrated by the numerical simulations. The results on a 3-area power system show that the method is encouraging for random load disturbances.
As electric energy demand has been increasing year after year, a power system tends to increase its capacity and become complicated. This tendency has made it difficult to properly design a protection system for power system equipment. Especially, since transmission lines spread over a vast area, most of faults which exist in a power system occur on them. Therefore, transmission line protection is necessary to deploy protective relays and determine its setting values corresponding to the configuration of a power system. This paper is to propose an expert system for transmission line protection whose knowledge base is extracted from two criteria for the purpose of guiding protection engineers and their knowledge. The main object of this expert system is to fully assist protection engineers in the transmission line protection. It involves the determination of protective relaying systems, deployment of relays, the determination of CT turns ratios and relay setting calculations. The proposed system is applied to a part of the Chugoku Electric Power Company System to verify its advantages.
This paper presents logical representations for causality and temporal reasoning by using these representations. First, a causal relation is interpreted by using two logical relations. One relation represents that if a cause state is occurred, an effect state will be occurred. Another relation represents that if an effect state doesn't exist and all of causes doesn't exist, an effect state will not exist. By using these two basic relations, three more logical relations are given. First one is that if an effect doesn't exist, causes didn't exist. Second one is that if an effect exists and causes didn't exist, an effect existed. Last one is that if an effect didn't exist and an effect exist at the next time, one of causes existed at least. And also, definite causality and possible causality are considered in these logical relations. These logical representations give us various interpretations about causality. These representations are applied to a protection system of a substation. Finally, program was developed based on these ideas. Results of temporal reasoning about states which have causal relations to another states are shown.
To study the long-term characteristics of XLPE cables installed in free air and in water, aging tests were conducted under various testing conditions using XLPE cables with both 3.5mm and 6mm insulation. From the Weibull plots of lifetime distribution under the voltage stress EL being the minimum breakdown strength, the minimum value of time to breakdown tL under the constant voltage was estimated. The results of accelerated aging tests of XLPE cables installed in free air demonstrated that the V-t characteristics of XLPE cables could not be described by the conventional inverse power's law (t∝ V-n) with a single constant life exponent n. Based on the microscopic observation of sliced insulation of removed XLPE cables, it was concluded that bow-tie trees with longer tree length, observed in cables tested in water, were caused by the moisture from outside, whereas the trees in cables tested in free air were caused by the residual moisure originally existing in the insulation. The breakdown strength of the aged cables tested in water increases through cable drying. It doesn't however recover to the original values.
Flat beam intensity at a target surface is required to study high heat flux materials. An optimum beam steering for this purpose is formulated and then numerically solved for the case that a multi-aperture ion source is used. It is found that the optimum distribution of the beam axes at the target surface is discrete when the beam divergence is comparable with the required flat width. The distribution gradually becomes continuous as the required width is enlarged.
The spectral responses in the single crystal silicon solar cell with bias light under the various forward bias voltages were investigated, and it was found that the response around the operating vlotages was lower than that in short-circuit condition. This phenomenon is not mainly due to the change of collection mechanism of photo-generated carriers, but can be explained by amount of (1+ARs)-1, where, Rs is the value of series resistance and factor A is derivated from the current-voltage characteristics and other equivalent circuit parameters.