In this paper, concerning wind power generating system in which batteries are connected to D.C. generator for load, development of control method and investigation on output characteristic by simulation were carried out. This control method is possible to take out wind energy effectively, using fuzzy reasoning by measurement of wind speed and windmill speed. Firstly, the stationary characteristics of the system were investigated in order to decide membership function. The windmill speed at stationary condition and batteries voltage which gives output (of generator) maximum were decided according to the windmill speed which is divided 5 classes according to struggle. Fuzzy rules were experiential decided as follows ; when there is a windmill speed to an acceleration state, windmill seed takes priority to control rules. On the contrary, when there is a windmill speed to slowdown state, wind speed takes it. As the results of investigation, effectiveness of output control by fuzzy reasining was confirmed for both cases of wind speed is fluctuating sinusoidally or for natural wind or wind speed is varying like natural wind. Especially, at the time of fluctuating wind speed, when frequency of wind speed becomes lower, effectiveness of output control becomes larger. In addition, for natural wind which was measured practically, it was made clear that larger energy generation is possible by practicing output control compared to constant load battery voltage.

View full abstract

This paper presents a neural network based real time maximum power tracking controller for interconnected photovoltaic(PV) systems to commercial power sources through inverters. The neural network is utilized to identify the optimal operating voltage of the PV array to achieve maximum output efficiency. The proposed controller generates the control signal in real time, and the control signal is fed back to the voltage control loop of the inverter to shift the terminal voltage of the PV array to the identified optimal one, which yields the maximum power generation. The controller is simply configured by using proportional and mtegral(PI) control loops. The proportional and the integral gains are tuned to their optimal values to achieve the quick response and also to prevent the overshoot and the undershoot in the transient state. The continuous measurement is required for the open circuit voltage on the monitoring cell, and also the terminal voltage of the PV array. The measurements are quite easy and straightforward. Because of the accurate identification of the optimal operating voltage through the neural network, more than 99% power is drawn from the actual maximum power.

View full abstract

The authors proposed a nonlinear adaptive generator control system with neural networks for improving damp-ing of power systems and showed its effectiveness in a one-machine infinite bus test power system in the previous paper. The proposed neuro-control system adaptively generates appropriate supplementary control signals to the conventional controllers such as automatic voltage regulator and speed governor so as to enhance transient stability and damping of the power system.
In this paper, the applicability of the proposed neuro-control system to multi-machine power systems is dis-cussed. Digital time simulations are carried out for a 4-machine test power system, where one or several synchronous generators are equipped with the neuro-control system. As a result, also in the multi-machine power system, the proposed adaptive neuro-control systems improve the system damping effectively and they work adaptively against the wide changes of the operating conditions and the network configuration.

View full abstract

Decentralized control is generally preferable to centralized one for large scale system. For the decentralized control of multi-machine electric power system, some decentralized systems are constructed by a model equation based on one-machine infinite-bus model system, and the stability analysis of each decentralized system guarantees the stability of the whole system. In this case, the stability analysis of one-machine infinite-bus model system becomes one of basic subjects for multi-machine system.
This paper treats a one-machine infinite-bus model system with governor as a simplest one including controller. First, the swing equation is rewritten into a fuzzy system, and the stability theorem is applied to the system. Next, the P-region which is necessary for the proof is discussed, and feedback gains for the fuzzy control input of the governor are determined. Finally, the control responses and the stability regions are considered. As a result, it is summarized that the stability of the model system can be rigorously analyzed by the proposed method, and the stability region can be evaluated by the fuzzy-type Lyapunov function.

View full abstract

This paper presents a new approach to identify a fault point in a power system on real time. Based on the topology theory, the characteristics of a fault in the power system are treated as a pattern of the fault. Therefore, the calculation complexity of traditional approaches can be avoided. The aim of proposed method is to identify fault points through directly analyzing the type of fault. Since each type of fault has each characteristic pattern of power flow, fault points can be identified by abstracting characteristics of power flow at each node of the power system. In order to abstract fault characteristics, Neocognitron in which symmetrical three phase decomposition and data normalization are calculated using power flow at each node is introduced. Thus, the impacts of voltage grade and unbalanced load can be removed. Since each Neocognitron corresponds to one node of the power system, hierachical autonomous decentralization can be realized. Therefore, the proposed approach can be applied to a large power system. Fault point location is done by BP network. Since the neuro of BP network only corresponds to the nodes of the topologized power system, the training of the neural network can be performed independently. From this point of view, the applicability and flexibility of proposed approach are high. The effectiveness and applicability of the proposed approach are demonstrated on a simple power system model.

View full abstract

In the operation planning and system planning tasks of a power system, various analysis programs must be used to examine the validity of assumed plans. As you well know, these analysis programs have been supplied by different vendors together with the related computer systems and hence they cannot be used in a particular computer system without adding significant modifications. Although it would be possible to use them on a so-called open system, an extensive amount of works is again needed to realize it.
In this paper, the authors propose a new supporting system in which a variety of analysis programs will be executed on a common platform irrespective of differences in data formats or handling procedures. In the first place, we propose to establish a unified database of a power system, from which a set of data necessary for each analysis program will be extracted. An object-oriented database has been chosen to express the hierarchical nature of a power system. To overcome differences in data formats and handling procedures, API (application program interface) is provided for each analysis program. A user-friendly graphical user interface (GUI) is fully exploited to assist a user to correctly handle analysis programs and to understand results properly. The validity and effectiveness of the proposed approach have been demonstrated by a prototype system.

View full abstract

Distribution system loss minimization re-configuration is a 0-1 planning problem, and the number of combinations requiring searches is extremely large when dealing with typical system scales. For this reason, the application of a genetic algorithm (GA) seems attractive to solve this problem. Although Genetic algorithms are a type of random number search method, they incorporate a multi-point search feature and are therefore superior to one-point search techniques. The efficiency of GAs for solving large combinational problems has received wide attention. Further, parallel searching can be performed and the optimal solution is more easily reached.
In this paper, for improving GA convergence characteristics in the distribution system loss minimization re-configeration problem, a chromosome “Limited Life” concept is introduced. Briefly, considering the population homogenization and genetic drift problems, natural selection is achieved by providing this new concept, in addition to natural selection by fitness. This is possible because individuals in a population have an age value.
Simulations were carried out using a model system to check this method's validity.

View full abstract

The high-voltage distribution system is connected in a tree branch configuration so as to meet the need for high reliability feeding. When a service interruption accident occurs, the faulty section is automatically separated and the healthy section carries out power flexibility from an adjacent system, thereby reducing the service interruption area. The faulty section in the distribution system is quickly located by means of time relays installed in some switchgears in the distribution system. However, the setting points of these time relay must be constantly monitored.
As such, we have proposed a “section management method” which can supervise each section unit of the distribution system. Based on this method, we then developed a prototype automatic distribution system for the overall management of time relays using internal software, facilities and customers feedback.
This paper describes operation of the distribution system in regard to the “section management method” was used as.

View full abstract

Recently utilities promote various researches to use existing equipments more effectively. Until now JCS (Japan Cable Maker's Association Standard ) method has been used to calculate the thermal capacity of power cables. However, in order to use existing power cables more effectively, detail calculations considering time varying thermal resistance and capacitance of surrounding soil are necessary. Regarding duct-installed power cables, duct temperatures can be measured by optical fiber distribution sensors which have been developed recently.
In this paper we propose a method to calculate the short-time thermal capacity of duct-installed power cables. The proposed algorithm is based on the estimation of the thermal resistance and capacitance of surrounding soil using continuously measured duct temperatures. In this algorithm an observer is used for the estimation of conductor temperatures, and also Genetic Algorithm is employed to estimate the thermal resistance and capacitance of surrounding soil accurately. The effectiveness of the proposed algorithm is demonstrated using the experiment data. It is shown that the application of proposed method improves the short-time thermal capacity of duct-installed power cables considerably.

View full abstract

Recently, active power filter has been studied as it enables the harmonic compensation according as the fluctuation of harmonic patterns. The compensation of active filter is conducted by detecting harmonic component from load current, then inject the equal-but-opposite harmonic current into power line by inverter. Therefore the accuracy and the speed of harmonic detection is important for the performance of active filter. In case that it has to specify the frequency of harmonic for high performance compensation, it needs to analyze load current into each kind of harmonic by using such the Fourier transformation. Fast Fourier transformation(FFT) enables high speed analysis, however, over one cycle of wave form is required, then the compensation is delayed more than two cycles of current wave.
The purpose of this paper is to improve the harmonic detection based on the frcquency decomposition by using artificial neural network. A half cycle of current wave form is used for harmonic analysis, the amplitudes at time series is the input signals of neural network. Each node in output-layer is corresponded for each kind of harmonic, the content of each harmonic is detected as the signals of the corresponding node. As neural network enables parallel processing, the content of each harmonic component is speedy detected within a half cycle. That means it is possible to compensate the harmonic in each half cycle of current wave. Therefore, the response time of active filter is improved more than 4 times comparing with the usual method. As the simulation results, we conclude that neural network is available to be used as the harmonic detector. By the proposed algorithm, it is possible to improve the performance of active power filter.

View full abstract

When a fault occurs in power transmission lines, we have to take proper measures depending on the cause and degree promptly. Now many excellent men have to be reared and engaged in the main-tenance work. A system to make the work efficient and speedy is thus expected. To achieve above, it is necessary to estimate the causes and places quickly and precisely. We studied on the fea-sibility of an expert system to assist the maintenance work. First, we surveyed the relationship between causes of faults in the past and their corresponding information such as substation's records, meteorological conditions, geographical features and so on. Then we built up an expert system using the past 5 years' data and evaluated it using the other past 2 years' data. As a result, correct answer rate with the first candidates of fault causes was 96% for lightning, and more than 80% for snow or wind. Also, when the second candidates are included, the correct answer rate turned to be more than 80% for birds, nests, snakes or monkeys. These evaluation expects the effectiveness of our system in actual power transmission lines.

View full abstract

Recently the task of planning in power systems is becoming a very complicated process for utility planners. This kind of planning has a lot of objectives to accomplish. In this paper we employ Negotiation Algorithm, which was proposed by the group of Chen-Ching Liu, for subtransmission power system planning to create a plan having many objectives. First, the Goal-Decision-Network(GDN) is constructed to model this planning problem, and Negotiation Algorithm is applied to utilize both subtrasmission system planning GDN, which attempts to capture its knowledge, and negotiation operators, which search for the most feasible and promising decisions in this planning GDN. Finally the negotiation expert system is demonstrated for the subtransmission system planning using a real system.

View full abstract

In Hokkaido Electric Power Co., we had developed some de-snowing wires to cope up with wet-snow accretion. And also preventive patrole is important to reduce power failure. We usualy conduct as early stage patrole and an assembling of circuits to melt snow by load dispatching operation. To support decisions of those measures, we developed a system on a CRT screen of engineering work station to forcast the wet-snow accretion. This system is very practical and useful, because we use transmission line data (loss levels of power line carrier and signals from snow accretion detecting devices), as well as meteorological meshed data which have fuzzy characteristics itself.

View full abstract

The authors have proposed a new framework of automatic knowledge acquision method and its prototype system has been reported at⁽⁶⁾, which has generated switching sequences for substations. This paper will present its application to instruction table generation for local load dispatching offices. In the case of dispatching office the huge number and individual particularity are the points at issue. The office has to deal with hundreds of substations and transmission lines, and make a few hundreds of switching instruction tables a month. Generating tables is one of the most difficult tasks because the uniqueness of substations such as structural differences and particular management rules have prevented us from completing knowledge base which will contain thousands of rules.
Our approach concerns the point for easy knowledge acquisition of the expert system, using an original learning algorithm named HCL (Hierarchical Concept based Learning). The HCL generates operational knowledge from cases in the past. This paper will discuss the fundamental problems of instruction table generation for dispatching offices and the solution to it by our approach. The method will be shown to construct and update its knowledge base automatically, using an actual network and operational cases of a dispatching office.

View full abstract

A stable supply of power is required from power systems as the importance of electrical energy increases. Particularly in the case of high voltage systems (500, 275, 154 kV bulk power systems), this requirement is high. There are, however, cases when power system faults cannot be avoided. Accordingly, it is very important to restore power systems quickly and safely if these failures happen. The characteristics of the power system restoration (its combinatorial aspects, use of knowlege from a wide variety of origins and of different types, number of criteria to satisfy) make it a difficult problem, for which the use of expert systems to generate restoration plans is being considered by many reserch teams, and promising results have already been obtained. But yet most of the systems stay in a prototype stage⁽¹⁾. One of the reasens seemes the studies of support system for knowledge-based behavior (unable to describe with any rule) are insufficient.
This paper begins by analyzing the power system restoration based on the human performance model⁽⁶⁾ and discuss the knowledge-based behaviour that is a high conceptual level human performance to solve the problem with combinatorial aspects. Then the application for the trunk line dispatching center is reported. Finally the relations between knowledge-based behavior and the designed human interfaces are verified with power networks restoration case-study.

View full abstract

For the design of a grounding electrode, the ground configuration at an intended site for the electrode needs to be clear. For this purpose, in many cases, a ρ-a curve from the measured results with Wenner's four electrodes arrangement is analysed by Sundberg's standard curves and Hummel's auxiliary curves. This paper is concerned with the analysis of the ρ-a curve for three-layer ground.
First, for deduction of three parameters of ρ₁, h₁, and ρ₂, the best analysed range of ρ-a curves is searched from many model data, and it is found that the best range is from the beginning point to 0.8 L_i1, , where L_i1 is the value of a at the first inflection point of the ρ-a curve. Next, for deduction of two parameters of h₂ and ρ₃, usual Hummel's and other four auxiliary curves are descussed from the view point of two layers displace-ment to one layer. For deduction of ρ₃, usual Hummel's curves are good enough for all types of ρ-a curves. But, for deduction of h₂, it is pointed out that, series-displacement curves are excellent for UD-type of ρ-a curves and Ono-displacement curves for DD-type, meanwhile usual Hummel's curves are good enough for DU-and UU-type. Then the best analysed range for deduction of h₂ and ρ₃ is also searched, and it is found that the best range is estimated from L_i1 or L_i2, where L_i2 is the value of a at the second inflection point of the ρ-a curve.

View full abstract

Studies are being made on the fault prevention by detecting discharge Phenomena with the optical sensors before the breakdown in switchgears. Monte Carlo Simulation has been applied to determine the illuminance distribution on the inside surface of wall of a model cylindrical tank, which is newly devised for investigating the relation between illuminance distribution and light sources caused by many types of partial discharge.
This paper introduces an outline and calculation example of a program to determine illuminance distribution in the gas-insulated switchgear (GIS). We regard discharge light as a collection of particles in Monte Carlo Simulation and pursue particle loci using random numbers ; thus the light distribution simulation on the inside wall of the equipment is obtained. The paper also describes the pertinence by comparing measurement and calculation results with a mimic discharge light in the cylindrical model tank and then shows shadow effects of the bus bar connectors on the illuminance distribution in GIS.
As a result, it is considered that Monte Carlo Simulation has the great advantage of wide applicabilty without any limitation to the illuminance calculation in complicated lighting systems having bus bars, isolators, etc.

View full abstract

A wide band magnetic direction finder is an element of the magnetic direction finding system, which is now widely used for lightning location. A magnetic direction finder exhibits systematic errors in the measurement of the bearing and the intensity of the incoming electromagnetic wave, and they are often called “site errors”. Two magnetic direction finding systems, having three direction finders each, were installed in Niigata area. The differences between two direction finders installed within several tens of meters, in the observed bearings and the amplitudes of electromagnetic waves associated with lightning return strokes, were measured. The results are explained by the theory of the quadrantal error, which is caused by nearby objects which scatter electromagnetic fields. A method to estimate site errors is proposed, and its application to the observed data draws the same conclusion, regarding the causes of the site errors at all of the three sites, as that based on the direct measurement of the differences of site errors.

View full abstract

Transient induced voltages due to lightning in an intelligent building are simulated using the generalized circuit analysis program EMTP based on a conventional circuit theory. A calculated result agrees satisfactorily with an experimental result, and thus the calculation method is proved to be applicable to an analysis of transient induced voltages. Using the method, transient voltages and currents in an intelligent building, which behaves as an inducing circuit are evaluated. It is observed that the size of the steel frame, a grounding resistance and a lightning current waveform affect significantly the voltages and currents. Also, it is found that a lightning current flowing into the building structure produces a voltage difference between different points on the same floor. A transient induced voltage across a resistance of a square-shape closed loop is basically determined by the rise time of a lightning current, the size of the loop and a separation distance from the vertical conductor which carries the largest lightning current. The maximum induced voltage to an 1×1m loop is observed to be 11V across 50 Ω resistance when the applied lightning current is of 1kA amplitude and 1/40μs waveshape. If a 10kA lightning, the average current measured, hits a building, the induced voltage reaches 100V which may cause a trouble on digital/electronic equipments.

View full abstract

This paper describes a new flashover model for lightning surge analysis. It is shown, at first, that starting conditions of leader propagation and leader propagation processes with the short tail light-ning impulse voltage waveform, which appears across the insulator strings on the transmission line by a lightning stroke, are clarified by experimental results. The new flashover model based on the experimental results is proposed in this paper. The proposed model consists of simple two equations, that give expression to the leader propagation processes, and can be realized very easily on the EMTP and other transient analysis programs. A comparison of results calculated by the proposed model with the experimental results shows good agreements.

View full abstract

Stability of an open-cycle supersonic disk MHD generator is studied, and effects of load conditions on stabilizing the generator are investigated. The growth rate of inlet disturbances during the propagation to the exit is evaluated by means of a sensitivity analysis. The analysis shows that the growth rate (sensitivity) is greatly affected by load conditions, and that constant load current operations bring very high sensitivity. A method of linear stability analysis is proposed for supersonic disk MHD generators considering their load conditions in order to analyze insta-bility induced by load current disturbances. Whether the disturbances grow or decay is judged in a way similar to the Nyqu fist's method. The sensitivity analysis, the linear stability analysis and tine-dependent calculations show that constant load voltage operations with multiple-loading schemes axe effective to suppress instability and high sensitivity, and that series inductors do not always stabilize the generator.

View full abstract